Hybrid robust beamforming for enhanced multiple moving object detection with phased array scanning radar Authors: Sheikh Shafaet Islam, Md. Selim Hossain, Shariful Islam, Rakibul Hasan Published: 2025 ABSTRACT The deficiencies of conventional radar systems can be addressed by utilizing a phased array scanning radar with robust beamforming capable of detecting and tracking multiple moving objects while rejecting interferences under noisier environments. This paper offers the idea of a 30 × 30 Uniform Rectangular Array (URA)-based radar system with the employment of a hybrid robust beamformer, a combination of Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV) beamformer with diagonal loading. Initially, a simulation work with experimental validation of a 1 × 16 Uniform Linear Array (ULA)-based radar system with a conventional beamformer is performed under a well-conditioned scenario, and it shows an overall accuracy of 89.80% for object detection; however, it is not feasible for the noisier condition. In order to enhance radar performance in challenging situations, a new hybrid robust beamforming model on a 30 × 30 URA-based phased array scanning radar system is used. Comparing the proposed radar system to conventional radar systems, the proposed phased array radar shows an overall accuracy of 99.29% for multiple object detection and interference rejection, according to the simulation results.Supplementary InformationThe online version contains supplementary material available at 10.1038/s41598-025-26071-4. Introduction Radar technology has long been an indispensable tool in various applications, ranging from weather monitoring and border security to marine surveillance, air traffic control, and military operations. Its ability to detect and track objects accurately over long distances and in challenging environments makes radar an essential component of modern surveillance systems. Among the many advancements in radar technology, phased array radar systems have emerged as a powerful solution due to their ability to electronically steer the beam without mechanically moving the antenna. This capability allows for improved gain, directivity, and performance in specific directions, making phased array radar particularly well-suited for scenarios requiring rapid and precise detection of multiple moving objects1. In recent years, the need for robust and efficient radar systems capable of detecting multiple moving objects has become increasingly critical, especially in areas such as national security, environmental monitoring, and disaster response. Traditional radar systems—commonly referred to as non-adaptive, mechanically steered radars—are effective for detecting single targets but often struggle to identify and track multiple targets in environments with high interference due to their limited target discrimination capabilities. These challenges are further exacerbated in environments with high levels of noise and interference, where distinguishing between signals of interest and unwanted signals becomes more complex2,3. To address these challenges, beamforming techniques have been developed and refined to enhance the detection performance of phased array radar systems. It employs an array of antennas to optimize signal reception from specific directions while minimizing noise and interference4. By focusing on the primary lobes and nulls towards the interference signal of no interest (SONI), beamforming can selectively capture signals from desired directions and reject others5. Phased array radar systems present a promising solution for overcoming these challenges by scanning the horizon in an angular pattern without mechanical rotation, using voltage-dependent phase shifts in the antenna elements. While single target detection by radar is well-established, detecting multiple targets remains challenging. Phased array radar, combined with appropriate algorithms, holds potential for addressing this challenge more effectively6,7. Despite substantial advancements in radar detection performance, there remains considerable scope for improvement. Various methods have been explored to enhance multiple target detection. For instance, the outlier rejection method was introduced in8 to handle multivariable data for target detection but struggled with moving targets and undefined azimuths. Improvements were made with the application of the weighted k-nearest neighbors (KNN) classifier proposed in9, but moving targets remained problematic. The antenna scanning method reported in10 could detect multiple targets in the same range-azimuth but faced limitations due to range azimuth resolution cells. The low-frequency ultra-wideband array-antenna synthetic aperture radar (SAR) was mentioned in11 for imaging both fixed and moving targets. Still, the moving target remained unobtainable to their system. The use of bistatic radar systems for the detection and positioning of several targets was proposed in12,13. Both literature12,13 prioritized multiple-input multiple-output (MIMO) radar, whereas literature13 focused on synthetic aperture radar (SAR) that was forward-looking. By successfully detecting and recognizing numerous objects, these methods resolved the spatial synchronization problem in traditional bistatic radars. In literature14, the authors provided a revision of that concept by introducing a bistatic linear array synthetic aperture radar (BLA-SAR) system for imaging and localizing moving targets. However, the detection capabilities for vast amounts of objects degraded as a result of their methodologies, which increased economic costs and processing complexity. The narrowband radar method in15 detected high-speed targets but failed to correct range migration adequately. The ultra-wideband (UWB) antenna method as in16 offered high gain for multiple target detection but was unsuitable without a UWB directional antenna. The low-rank and sparsity-constrained method introduced in17 provided a real beam image for multiple target detection but suffered from poor azimuth resolution due to antenna size limitations. Additionally, techniques requiring multiple radars as suggested in18 were impractical due to their feasibility constraints. MIMO radar reported in19 examined target correlation and detection sensitivity but struggled with new or nearby tracks deviating from assumptions of targets and clutter statistics. Cognitive beamforming techniques in20 for co-located MIMO radars based on Reinforcement Learning (RL) faced performance declines with rising correlation levels. Radar systems with antenna allocation methods in21 and two-element rotating antenna strategies in22 showed potential but had limitations in angle estimation and resolution. In literature23, the Rotman lens beamforming technique was utilized to ascertain the position of an object by analyzing the phasor disparity. However, it was able to differentiate between targets that were equidistant but not at a different angle. Recent studies have explored diverse strategies to enhance radar performance using intelligent and adaptive technologies. In literature24, reconfigurable intelligent surfaces (RIS) were used to improve radar detection, though the analysis was limited to basic setups, lacking multi-target or complex interference considerations. The displaced phase center antenna (DPCA)-based pulse-to-pulse subtraction method proposed in25 effectively suppressed clutter from platform motion, enhancing slow-moving target detection; however, the optimal space-time adaptive processing (STAP)-based techniques required intensive computation and reliable training data, which is scarce in airborne settings. A reduced-dimension STAP framework introduced in26 lowered computational complexity while preserving detection accuracy, though its success hinged on accurate subspace estimation under ideal assumptions. The intelligent reflecting surface (IRS)-assisted MIMO radar proposed in27 showed performance gains in non-line-of-sight conditions, but the semidefinite relaxation and alternating optimization techniques posed real-time implementation challenges. The decomposition-based resource and waveform management scheme in28 achieved efficient multi-node coordination yet relied on perfect synchronization and target state accuracy, which are difficult in practice. In literature29, an intelligent metasurface antenna with reconfigurable beams using hybrid optimization was presented, although its control complexity and circuit demands could hinder scalability for larger systems. The tapered or banded shrinkage covariance (TABASCO) matrix, introduced in30, combined shrinkage towards a scaled identity matrix with tapering to enhance covariance estimation in high-dimensional settings with low sample-to-dimension ratios. Optimal regularization parameter derivation and empirical validation demonstrated its effectiveness, including in STAP for radar systems, though it remains computationally intensive, assumes elliptically symmetric data, and is sensitive to bandwidth selection. In literature31, Constant False Alarm Rate (CFAR) detection methods based on banded sample autocovariance matrices (SACM) achieved superior detection performance and computational efficiency over conventional approaches, with the Maximum Eigenvalue Trimmed (MET) detector offering adaptability without requiring non-target reference data. However, the theoretical complexity of deriving SACM eigenvalue distributions limits practical implementation. A random slow-time coding method based on Signal-to-Interference-Plus-Noise Ratio (SINR) maximization was proposed in32 to mitigate ghost target interference in autonomous driving, demonstrating improved detection but at the cost of a higher noise floor, potentially obscuring weak targets. In literature33, a high-resolution imaging method for high-speed maneuvering targets employed fast sparse Bayesian learning (SBL) and minimum entropy algorithms, leveraging fast Fourier transform (FFT)-based decomposition to reduce matrix inversion costs. While computationally efficient, its performance degraded at low SNRs. Furthermore, literature34 proposed a Frequency Diverse Array–Multiple-Input Multiple-Output (FDA-MIMO) detection technique integrating training data and prior covariance information via Bayesian methods, reducing dependency between transmission snapshots and signal dimensions. This approach, however, relied heavily on accurate prior information and adequate training data. Despite significant progress in this area, existing methods often struggle with issues such as poor resolution, high computational complexity, and limited effectiveness in real-world scenarios involving multiple moving targets. The key contributions of this paper are as follows. This paper: Proposes an enhanced beamforming approach tailored for accurate detection and tracking of multiple moving objects using phased array radar in challenging environments.Develops modified radar equations for rectangular arrays to improve modeling accuracy and system representation.Designs a hybrid robust beamformer that combines Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV) beamforming techniques, offering improved interference rejection and detection performance.Integrates Uniform Rectangular Array (URA) with the hybrid beamforming strategy to achieve higher spatial resolution and robustness.Utilizes a condition number-based optimization technique to reduce computational burden.Evaluates the proposed system through simulations and experimental validation, demonstrating effective multi-target detection and accurate estimation of range, angle, and velocity. Proposes an enhanced beamforming approach tailored for accurate detection and tracking of multiple moving objects using phased array radar in challenging environments. Proposes an enhanced beamforming approach tailored for accurate detection and tracking of multiple moving objects using phased array radar in challenging environments. Develops modified radar equations for rectangular arrays to improve modeling accuracy and system representation. Develops modified radar equations for rectangular arrays to improve modeling accuracy and system representation. Designs a hybrid robust beamformer that combines Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV) beamforming techniques, offering improved interference rejection and detection performance. Designs a hybrid robust beamformer that combines Minimum Variance Distortionless Response (MVDR) and Linearly Constrained Minimum Variance (LCMV) beamforming techniques, offering improved interference rejection and detection performance. Integrates Uniform Rectangular Array (URA) with the hybrid beamforming strategy to achieve higher spatial resolution and robustness. Integrates Uniform Rectangular Array (URA) with the hybrid beamforming strategy to achieve higher spatial resolution and robustness. Utilizes a condition number-based optimization technique to reduce computational burden. Utilizes a condition number-based optimization technique to reduce computational burden. Evaluates the proposed system through simulations and experimental validation, demonstrating effective multi-target detection and accurate estimation of range, angle, and velocity. Evaluates the proposed system through simulations and experimental validation, demonstrating effective multi-target detection and accurate estimation of range, angle, and velocity. To be more specific, in this research, receiver (Rx) beamforming is proposed. The beamforming operation is performed during signal reception using adaptive weights applied to received signals at the antenna array. Modeling of the phased array system Let us assume a two-dimensional (2D) rectangular array antenna system shown in Fig. 1, which comprises \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:M\times\:N$$\end{document} antenna elements, where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:M$$\end{document} denotes the elements in the horizontal direction and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:N$$\end{document} denotes the elements in the vertical direction. Each element’s signal is multiplied by a complex weight and then added simultaneously to produce the array output. In Fig. 1, the blue lines represent the received signals from each array element. These signals are routed individually from each antenna element to corresponding processing channels. The blue lines visualize the raw input signal flow from the rectangular array to the weighting network. The orange lines denote the feedback connection to the weight estimator, which is responsible for determining the optimal complex weights. These weights are computed based on the desired signal direction and possibly using adaptive or robust algorithms to maximize target gain while suppressing interference. The red lines indicate the application of computed weights to the received signals. This multiplication step forms the core of the beamforming process, aligning the phases and amplitudes of the incoming signals to steer the main lobe in the desired direction and nullify interference sources. Finally, the green lines represent the summation stage, where all weighted signals are coherently combined to produce the beamformer output. This output maximizes gain in the direction of interest while minimizing contributions from noise and undesired directions. Fig. 1Layout of the rectangular array antenna system. Layout of the rectangular array antenna system. The expression for the array output while dealing with noncoherent signal processing is given by35:1\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:y=\sum\:_{m=1}^{M}\sum\:_{n=1}^{N}{w}_{mn}{x}_{mn}$$\end{document} where m and n are the indices representing the horizontal and vertical positions of the antenna element respectively, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:y$$\end{document} represents the array output, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{w}_{mn}$$\end{document} represents the complex weight applied to the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(m,n)$$\end{document}th antenna element, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{x}_{mn}$$\end{document} represents the signal induced on the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:(m,n)$$\end{document}th antenna element. Expressing the weights of the array system utilizing matrix notation as2\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W} = \left[ {\begin{array}{*{20}c} {w_{{11}} } & {w_{{12}} } & \cdots & {w_{{1N}} } \\ {w_{{21}} } & {w_{{22}} } & \cdots & {w_{{2N}} } \\ \vdots & \vdots & \ddots & \vdots \\ {w_{{M1}} } & {w_{{M2}} } & \cdots & {w_{{MN}} } \\ \end{array} } \right]$$\end{document} and signals induced on all elements as3\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X} = \left[ {\begin{array}{*{20}c} {x_{{11}} } & {x_{{12}} } & \cdots & {x_{{1N}} } \\ {x_{{21}} } & {x_{{22}} } & \cdots & {x_{{2N}} } \\ \vdots & \vdots & \ddots & \vdots \\ {x_{{M1}} } & {x_{{M2}} } & \cdots & {x_{{MN}} } \\ \end{array} } \right]$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}\in\:{\mathbb{C}}^{M\times\:N}$$\end{document} represents the weight matrix of the array, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}\in\:{\mathbb{C}}^{M\times\:N}$$\end{document} represents the signal matrix of the array. Thus, the output of the array becomes35:4\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}\in\:{\mathbb{C}}^{N\times\:N}$$\end{document} represents the output matrix of the array, and superscript H denotes the complex conjugate transposition of a vector or matrix. The output power of the array \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}$$\end{document} at any time is given by35:5\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}={\left|\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}\right|}^{2}=\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}^{*}$$\end{document} where * denotes the complex conjugate. Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}$$\end{document} from Eq. (4), the output power becomes35:6\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}{=\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} Suppose the components of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}$$\end{document} can be represented as stationary processes with a mean of zero. In that case, the mean output power of the array system for a given \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} can be determined by calculating the conditional expectation over \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}$$\end{document}35.7\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}\left(\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}\right)=\mathbb{E}\left[\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}^{*}\right]=\mathbb{E}\left[{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}\right]={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\mathbb{E}\left[\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}^{H}\right]\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} where E[❚] denotes the expectation operator, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}\in\:{\mathbb{C}}^{M\times\:M}$$\end{document} represents the array correlation matrix defined by35:8\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}=\mathbb{E}\left[\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}^{H}\right]$$\end{document} Imagine a desired signal source surrounded by undesired interference and random noise. Background noise and electronic noise are both part of the random noise. The signal matrix resulting from the desirable signal source, undesired interference, and random noise are represented as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{s}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{I}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}$$\end{document}, respectively. Then, by taking the inner product of the weight matrix with \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{s}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{I}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}$$\end{document}, the components of signal, interference, and random noise in the output \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{s}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{I}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{v}$$\end{document} are obtained35.9\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{s}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{s}$$\end{document}10\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{I}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{I}$$\end{document}11\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}}_{v}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}$$\end{document} The array correlation matrices caused by the signal source, undesired interference, and random noise can be defined as (in that order)35:12\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{s}=\mathbb{E}\left[{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{s}\:{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{s}}^{H}\right]$$\end{document}13\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}=\mathbb{E}\left[{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{I}\:{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}}_{I}}^{H}\right]$$\end{document}14\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}=\mathbb{E}\left[\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}}^{H}\right]$$\end{document} It is noted that \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}$$\end{document} is the sum of these three matrices, that can be written as35:15\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{s}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}$$\end{document} Let us assume \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{s}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{I}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{v}$$\end{document} represent the mean output power due to the signal source, unwanted interference, and random noise, respectively. The following are given considering Eq. (7)35:16\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{s}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{s}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document}17\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{I}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document}18\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{v}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} Let \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}$$\end{document} stand for the mean power at the array’s output that is attributable to undesired interference and random noise; thus,19\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{I}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{v}$$\end{document} We define \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{{\Lambda\:}}}_{V}$$\end{document} as the mean noise power at the array system’s output. Now, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{I}$$\end{document} from Eq. (17) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{v}$$\end{document} from Eq. (18) in Eq. (19), we obtain35:20\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\left({\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}\right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} Thus,21\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{V}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{I}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{v}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{V}$$\end{document} represents the noise array correlation matrix. Then,22\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{V}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}$$\end{document} represents the mean noise power at the output of the system in terms of weight matrix and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{V}$$\end{document}. Therefore, the output SNR of the array system is given by35:23\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{array}=\frac{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{s}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}}$$\end{document} Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{s}$$\end{document} from Eq. (16) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varLambda\:}}_{V}$$\end{document} from Eq. (22) in Eq. (23), we finally obtain35:24\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{array}=\frac{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{s}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\varGamma\:}}_{V}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}$$\end{document} Signal model It is assumed that a desired source signal (plane wave) is propagating from the azimuth angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document}, in free space at the speed of light, which is 3 × 108 m/s. The signal is impinging on the rectangular array on the same plane as seen in Fig. 2. The steering matrix, denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}$$\end{document}, is derived accordingly36:25\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S} _{\theta } = \left[ {\begin{array}{*{20}c} {b_{1} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A} } & {b_{2} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A} } & { \ldots ~~~b_{N} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A} } \\ \end{array} } \right] = \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B} ^{T} \otimes ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}\in\:{\mathbb{C}}^{M\times\:N}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B}$$\end{document} represent the steering vectors associated with the x-axis and y-axis, respectively, superscript T denotes the transposition of a vector or matrix, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\otimes ~$$\end{document} denotes the Kronecker product. Representing \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B}$$\end{document} in the matrix form:26\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A}={\left[\begin{array}{ccc}1&\:{e}^{-j{\varOmega\:}_{x}}&\:\dots\:\:\:\:{e}^{-j\left(M-1\right){\varOmega\:}_{x}}\end{array}\right]}^{T}$$\end{document}27\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B}={\left[\begin{array}{ccc}{b}_{1}&\:{b}_{2}&\:\dots\:\:\:\:{b}_{N}\end{array}\right]}^{T}={\left[\begin{array}{ccc}1&\:{e}^{-j{\varOmega\:}_{y}}&\:\dots\:\:\:\:{e}^{-j\left(N-1\right){\varOmega\:}_{y}}\end{array}\right]}^{T}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A}\in\:{\mathbb{C}}^{M\times\:1}$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B}\in\:{\mathbb{C}}^{N\times\:1}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:j$$\end{document} represents imaginary unit, and28\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\varOmega\:}_{x}=\frac{2\pi\:f{\delta\:}_{x}\text{cos}\theta\:}{c}$$\end{document}29\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\varOmega\:}_{y}=\frac{2\pi\:f{\delta\:}_{y}\text{sin}\theta\:}{c}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f$$\end{document} represents the operating frequency, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\delta\:}_{x}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\delta\:}_{y}$$\end{document} represent the spacing between adjacent antenna elements in the horizontal and vertical direction, respectively, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{cos}\theta\:$$\end{document} accounts for the angle of arrival in the horizontal direction, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\text{sin}\theta\:$$\end{document} accounts for the angle of arrival in the vertical direction, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:c$$\end{document} is the speed of light. Fig. 2Rectangular array geometry. Rectangular array geometry. Array factor formulation Now, the signal matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}$$\end{document} can be represented in terms of the steering matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}$$\end{document} as35,36:30\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}=\sum\:_{m=1}^{M}\sum\:_{n=1}^{N}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\alpha\:}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}+\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\alpha\:}$$\end{document} represents the complex modulating function that is associated with each element (m, n) in the array. These coefficients model the spatially dependent gain and phase shift introduced by each element due to its physical location, element pattern, and the angle of arrival of the signal, which is essential for accurate beamforming. Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}$$\end{document} from Eq. (30) in Eq. (4), it follows that35,36:31\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{Y}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{X}=\sum\:_{m=1}^{M}\sum\:_{n=1}^{N}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\alpha\:}\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}+{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{v}$$\end{document} The contribution from all directional sources constitutes the first term on the right side of Eq. (31); the second term is the random noise contribution to the array output. Recall that the weighted total of modulating functions of all directional sources included in the first term determines their contribution. The weight assigned to every source represents the complex response of the processor toward the source employing the inner product of the processor weight matrix and steering matrix related to that source. Therefore, the response of a processor with weight matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} toward a source in the azimuth direction \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document} is given by35,36:32\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\mathcal{Y}}_{\theta } = \left\langle {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W} |\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S} _{\theta } } \right\rangle = \left( {vec~\left( {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W} } \right)} \right)^{H} vec~\left( {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S} _{\theta } } \right)$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathcal{Y}}_{\theta\:}$$\end{document} represents the array factor of the system, <❚ |❚> denotes the inner product operator, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$vec$$\end{document}(❚) denotes the vectorization operator. Since \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathcal{Y}}_{\theta\:}$$\end{document} is a scalar quantity, we need to perform column-wise vectorization on \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S}}_{\theta\:}$$\end{document}.33\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}=\text{v}\text{e}\text{c}\:\left(\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{W}\right)={\left[\begin{array}{ccccccccccccc}{w}_{11}&\:{w}_{21}&\:\cdots\:&\:{w}_{M1}&\:{w}_{12}&\:{w}_{22}&\:\cdots\:&\:{w}_{M2}&\:\cdots\:&\:{w}_{1N}&\:{w}_{2N}&\:\cdots\:&\:{w}_{MN}\end{array}\right]}^{T}$$\end{document}34\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } = vec~\left( {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{S} _{\theta } } \right) = vec~\left( {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{B} ^{T} \otimes \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{A} } \right)$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}\in\:{\mathbb{C}}^{MN\times\:1}$$\end{document} represents the vectorized weight matrix or weight vector and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}\in\:{\mathbb{C}}^{MN\times\:1}$$\end{document} represents the vectorized steering matrix or steering vector. The vectorization process simplifies matrix inner products into dot products, enabling easier expression of 2D array operations in one-dimensional (1D) linear algebraic form. This is particularly helpful in deriving the array factor scalar from multidimensional weights and steering matrices. Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}$$\end{document} from Eq. (33) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}$$\end{document} from Eq. (34) in Eq. (32), it follows that35,36:35\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathcal{Y}}_{\theta\:}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}$$\end{document} Beamforming techniques Beamforming is a signal processing technique used in sensor arrays for directional signal transmission or reception. According to the effectiveness, beamforming techniques can be classified into three parts – conventional beamforming, adaptive beamforming, and robust beamforming35. Conventional beamformer In a conventional (or classical) beamformer, also known as a delay-and-sum beamformer, the weight vector is designed to steer the array in the direction of the desired signal without considering interference or noise explicitly. The primary goal is to align the phases of the signals received at different array elements to enhance the desired signal’s direction. For a conventional beamformer, the weight vector is chosen to match the steering vector of the desired signal. This ensures that the array is steered in the direction of the desired signal35. The weight vector for the conventional beamformer is denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{conv}$$\end{document} and is given by35:36\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{conv}} = \frac{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{0}$$\end{document} is the steering vector in the direction of desired signal and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|$$\end{document} denotes the Euclidean norm of the steering vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{0}$$\end{document}. The expression for \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|$$\end{document} is defined as35:37\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\| = \sqrt {\mathop \sum \limits_{{i = 1}}^{{MN}} \left| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{{0_{i} }} } \right|^{2} }$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{0}}_{i}$$\end{document} denotes the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:i$$\end{document}th element of the vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{0}$$\end{document}, |❚| denotes the absolute value, and the summation is over all the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:MN$$\end{document} elements of the vector. Adaptive beamformer An adaptive beamformer is a beamforming approach that dynamically modifies the weights of array elements to focus the beam toward a specified direction while reducing interference from other angles. The adaptation is generally predicated on a cost function, such as minimizing output power or maximizing the SINR37. Minimum variance distortion-less response beamformer The MVDR algorithm, often called the Capon beamformer, guarantees a distortion-free beam pattern in the desired signal direction while minimizing the output power of the beamformed signal. The optimal weight vector for the MVDR beamformer is denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{MVDR}$$\end{document} which is given by (See the full derivation in Appendix A)35,36,38:38\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{MVDR}} = \frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } \in{\mathbb{C}}^{{MN \times MN}}$$\end{document} denotes the transformed correlation matrix and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \in {\mathbb{C}}^{{MN \times MN}}$$\end{document} denotes the inverse transformed correlation matrix. Linearly constrained minimum variance beamformer Utilizing linear constraints to receive the required signal with signal integrity or null steering towards interferers, the LCMV approach optimizes beamformer output power. The optimal weight vector for the LCMV beamformer is denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{LCMV}$$\end{document} which is given by (see the full derivation in Appendix B)39–41:39\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{LCMV}} = \left( {\frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C}$$\end{document} represents the constraint matrix of order \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:MN\times\:L$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g}$$\end{document} represents the response vector of order \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:L\times\:1$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:L$$\end{document} is the number of linear constraints. Robust beamformer A robust beamformer is engineered to sustain its performance despite uncertainties or mismatches in the system model, including direction of arrival (DOA) errors, calibration inaccuracies, or variations in signal statistics. The objective of robust beamforming is to mitigate the impact of these faults to attain enhanced reliability in performance relative to conventional adaptive beamformers. Robust beamforming techniques generally adjust or enhance adaptive beamforming algorithms to address these uncertainties42. Diagonal loading MVDR beamformer In this beamformer, a regularization term (diagonal loading) is added to the array correlation matrix to make the beamformer more resilient to errors in the array correlation matrix estimation. So, the loaded array correlation matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document} can be expressed as42:40\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} = \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} represents the diagonal loading factor and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I}}_{MN}$$\end{document} denotes the identity matrix of order \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:MN\times\:MN$$\end{document}. The loading factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} determines the strength of the regularization. A small \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} improves robustness without significantly degrading resolution, whereas a large \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} overly suppresses adaptive nulling. The choice of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} is therefore critical and is often tuned based on noise level, sample support, or system sensitivity42. According to41, the optimal weight vector for the diagonal loading MVDR beamformer \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_MVDR}$$\end{document} can be formulated as:41\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{DL\_MVDR}} = \frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}$$\end{document} Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document} from Eq. (40) in Eq. (41), we finally obtain41:42\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{DL\_MVDR}} = \frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}$$\end{document} Diagonal loading LCMV beamformer The process of adding a diagonal loading factor to the array correlation matrix to make the LCMV beamformer more resilient to errors in the array correlation matrix estimation is kept same as the diagonal loading MVDR beamformer. According to41, the optimal weight vector for the diagonal loading LCMV beamformer \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_LCMV}$$\end{document} can be formulated as:43\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{DL\_LCMV}} = \left( {\frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} = \left( {\frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g}$$\end{document} Proposed hybrid robust beamformer The proposed hybrid robust beamformer is constructed by integrating the strengths of the diagonal loading MVDR beamformer and diagonal loading LCMV beamformer. Diagonal loading is incorporated into both methods to address mismatches and enhance resilience against faults, including inaccurate DOA data and array deficiencies. The diagonal loading MVDR is employed to obtain the highest gain at the DOA and the diagonal loading LCMV to obtain the deepest null at the interference. The hybrid beamformer can combine the output of the two beamformers in a way that balances their performance under different conditions. A hybrid weighting factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:\in\:\left[\text{0,1}\right]$$\end{document} is used to control the relative contributions of the two beamformers where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_MVDR}$$\end{document} is the weight vector for the diagonal loading MVDR beamformer and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_LCMV}$$\end{document} is the weight vector for the diagonal loading LCMV beamformer. The final weight vector for the proposed hybrid beamformer \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}$$\end{document} can be formulated as:44\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}=\beta\:\bullet\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_MVDR}+\left(1-\beta\:\right)\bullet\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_LCMV}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} can be dynamically adjusted based on the scenario, interference, or user-defined criteria. Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_MVDR}$$\end{document} from Eq. (42) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_LCMV}$$\end{document} from Eq. (43) in Eq. (44), it follows that:45\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w} _{{hybrid}} = \beta \left[ {\frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}} \right] + \left( {1 - \beta } \right)\left[ {\left( {\frac{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} } \right]$$\end{document} The hybrid weighting factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} controls the balance between MVDR and LCMV which can be adapted based on the environment. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=1$$\end{document} gives the pure MVDR solution and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:=0$$\end{document} gives the pure LCMV solution. However, a scenario-based choice of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} can provide a better compromise between interference suppression and robustness to mismatches. Fig. 3 illustrates the flowchart of the proposed hybrid robust beamforming approach. The process begins by loading the required covariance matrices and steering vectors for both MVDR and LCMV beamformers. Each branch applies diagonal loading to enhance robustness, computing the loaded array correlation matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document}. Subsequently, the respective beamformer weights \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_MVDR}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{DL\_LCMV}$$\end{document} are computed. These are then linearly combined using the hybrid weighting factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} to form the hybrid beamformer weight \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}$$\end{document}. Finally, the array output is calculated and evaluated. If the result meets the desired performance criteria, the process ends; otherwise, the weighting factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} is adjusted iteratively. In a stable scenario, conventional and adaptive beamformers can perform efficiently. However, a challenging scenario degrades the performance of both conventional and adaptive beamformers. The proposed hybrid robust beamformer can be a solution to this problem. Thus, the performance of the proposed hybrid model can be evaluated in the challenging scenario. Fig. 3Flowchart for the proposed hybrid robust beamformer. Flowchart for the proposed hybrid robust beamformer. Optimization technique In order to obtain the optimal weights for both diagonal loading MVDR and diagonal loading LCMV beamformer, we must select the optimal diagonal loading factor by which desired robustness can be achieved. In this research, the condition number-based method is employed for the selection of the diagonal loading factor dynamically. The condition number-based method can be defined as43:46\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right) = \frac{{\Xi _{{\max }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)}}{{\Xi _{{\min }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)}} = \frac{{\Xi _{{\max }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)}}{{\Xi _{{\min }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)$$\end{document} represents the condition number of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Xi _{{\max }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\Xi _{{\min }} \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)$$\end{document} represent the largest and smallest eigenvalues of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document}, respectively. Eigenvalues are important because they provide insights into the behavior and stability of the matrix. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\varXi\:}_{max}$$\end{document} corresponds to the largest scaling factor of the matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\varXi\:}_{min}$$\end{document} corresponds to the smallest scaling factor to the matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document}. The eigenvalues give a measure of how the matrix behaves under transformations, and in signal processing, these values help assess the matrix’s sensitivity to noise and interference. The condition number gives a measure of the conditioning of the matrix. A high condition number indicates that the matrix is ill-conditioned, meaning that it is close to being singular or unstable, making it difficult to invert or compute accurately. An ill-conditioned matrix is sensitive to small changes in its inputs and is prone to errors. If \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}} } \right)$$\end{document} is large, it means that the matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } _{{load}}$$\end{document} has a large disparity between its largest and smallest eigenvalues, indicating poor numerical stability. This suggests that the system is sensitive to perturbations, and a larger loading factor γ should be needed to stabilize it. If the condition number is close to unity, the matrix is well-conditioned, meaning it is numerically stable, and the computations involving this matrix will be more reliable42–44. In the context of beamforming, it is aimed to determine the optimal diagonal loading factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{opt}$$\end{document} to attain a desired condition number \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\kappa\:}_{target}$$\end{document} for the loaded array correlation matrix \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)$$\end{document}. The desired condition number \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\kappa\:}_{target}$$\end{document} is initially assumed by the Randomized Numerical Linear Algebra (RNLA). RNLA is a set of probabilistic algorithms aimed at efficiently approximating large-scale matrix operations, including low-rank decompositions and least-squares solutions, by utilizing random sampling and projection methods. This approach decreases computational complexity while maintaining accuracy, making it suitable for scalable solutions to problems like Singular Value Decomposition (SVD) and solving linear systems in high-dimensional spaces, such as beamforming for large antenna arrays. This is an effective way to define \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\kappa\:}_{target}$$\end{document} and faster than SVD. It reduces complexity from \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathcal{O}\left({M}^{3}{N}^{3}\right)$$\end{document} (SVD) to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\mathcal{O}\left({M}^{2}{N}^{2}k\right)$$\end{document} for rank-\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:k$$\end{document} approximations, where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:k\ll\:MN$$\end{document}45. Fig. 4Flowchart for the computational process of the grid search method. Flowchart for the computational process of the grid search method. The objective function in this instance could correspond to the absolute difference between the current condition number \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)$$\end{document} and the desired condition number \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\kappa\:}_{target}$$\end{document}46:47\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$f\left( \gamma \right) = \left| {\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right) - \kappa _{{t\arg et}} } \right|$$\end{document} This involves computational approaches in order to obtain the optimal diagonal loading factor. The traditional approach can be the grid search method which is a brute-force optimization strategy employed to determine the optimal parameter value by exhaustively exploring a predetermined grid of potential values. The grid search method can be mathematically expressed as47:48\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{opt}=\underset{\gamma\:\in\:\left\{{\gamma\:}_{1},{{\:\gamma\:}}_{2},{\:\:}\cdots\:,{\:\:}{\gamma\:}_{n}\right\}}{\text{arg\:min}}f\left(\gamma\:\right)$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{1},{\:\gamma\:}_{2},\:\:\cdots\:,\:\:{\gamma\:}_{n}$$\end{document} are the values of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document} in the predefined grid and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f\left(\gamma\:\right)$$\end{document} denotes the objective function evaluated at each \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document}. The computational process of the grid search method is illustrated in Fig. 4. This method is simple and easy to implement. However, it needs to evaluate the objective function at every point inside the grid. This becomes computationally unfeasible in the case of high-dimensional issues or large grids. The number of evaluations increases exponentially with the number of parameters or dimensions under optimization48. In order to reduce the computational complexity and speed up convergence, the proposed computational approach is the Newton-Raphson method. The Newton-Raphson method is an effective numerical technique for determining the roots of a function. The objective of the condition number-based technique in the hybrid robust beamformer is to optimize the diagonal loading factor to maintain an optimally conditioned array correlation matrix. The Newton-Raphson update rule for finding the optimal diagonal loading factor is as follows46:49\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{n+1}={\gamma\:}_{n}-\frac{f\left({\gamma\:}_{n}\right)}{f{\prime\:}\left({\gamma\:}_{n}\right)}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:n$$\end{document} indicates the number of iterations (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:n=0,\:1,\:2,\:\dots\:$$\end{document}), and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f{\prime\:}\left(\gamma\:\right)$$\end{document} is the derivative of the objective function \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f\left(\gamma\:\right)$$\end{document} with respect to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document}. Following Eq. (47), the derivative of the condition number can be approximated numerically using the change in the condition number with a small perturbation \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\partial\:$$\end{document} in \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document}46:50\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$f'\left( \gamma \right) \approx \frac{{\kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \left( {\gamma + \partial } \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right) - \kappa \left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } + \gamma \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} _{{MN}} } \right)}}{\partial }$$\end{document} The iteration process continues until it reaches the defined maximum iteration \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:max\_iter$$\end{document}, or the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left|f\left({\gamma\:}_{n}\right)\right|$$\end{document} is below a chosen tolerance level \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\epsilon\:$$\end{document}. Fig. 5Flowchart for the computational process of the proposed Newton-Raphson method. Flowchart for the computational process of the proposed Newton-Raphson method. Fig. 5 presents the flowchart of the Newton-Raphson-based computational process used to optimize the beamforming weight parameter \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\gamma\:$$\end{document}. The procedure begins with initializing parameters such as the number of array elements, snapshots, array correlation matrix, desired condition number, tolerance , and maximum iterations. An initial guess \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{0}$$\end{document}​ is assumed, and the iterative process commences by computing the loaded array correlation matrix and condition number. The objective function \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f\left({\gamma\:}_{n}\right)$$\end{document} and its derivative \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f{\prime\:}\left({\gamma\:}_{n}\right)$$\end{document} are evaluated to update \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{n}$$\end{document} using the Newton-Raphson formula depicted in Eq. (49). The convergence data is stored at each iteration, and the loop continues until the convergence criterion \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\left|{\gamma\:}_{n+1}-{\gamma\:}_{n}\right|<\epsilon\:$$\end{document} or the maximum iteration count is met. Upon convergence, the final optimized value \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\gamma\:}_{opt}$$\end{document}​ is obtained. Unlike brute-force or basic grid search, Newton-Raphson converges more quickly by utilizing the derivative information. It can handle real-time updates in dynamic environments. This approach is particularly appropriate when fast modifications to the diagonal loading factor are required due to fluctuating interference and noise conditions46. Modeling of the radar system The schematic diagram integrating transmission and reception of the radar system is demonstrated in Fig. 6. The system’s phased array radar is set as a monostatic radar so that signals can be transmitted and received via the same antenna array. Fig. 6Schematic block of the phased array radar system. Schematic block of the phased array radar system. The waveform generator generates the radar signal first; next, the power amplifier amplifies it in order to assure that the transmitted signal has the required strength to reach the appropriate detection range and resolution. Once this step is completed, the radar signal gets transmitted to the propagation channel using a phased array antenna (PAA) with the requisite gain capability. Upon reaching the cross-section of the target within the propagation channel, the reflected signal is reabsorbed by the antenna array, thereby being collected by it. Moreover, the duplexer enables the receiver to obtain the reflected signal, facilitating further operations. Subsequently, an analog-to-digital converter is employed to sample the analog signal at discrete intervals, converting these sampled signals into a digital representation that can be further processed. The next step is to steer the beam in a given target azimuth and create nulls for interference by applying optimal beamforming techniques to the signals received by each antenna in the array. Improvements in range resolution, target differentiation, and clutter management are achieved by the utilization of pulse synthesis and matched filtering. To make the radar system practical, the pulse integration technique is employed in the matched filtering section to reduce the transmitted power required. Following the modification of the given pulse, a comparison is made with the signal that is being received. A method known as Doppler processing is applied to moving objects in order to provide the main computer with information regarding their velocity and range. After the detector has precisely counted the number of moving objects within the main computer, the tracking and parameter estimate of objects will come to an end. When this stage is completed, threshold selection is carried out in order to reduce the number of false alarms and improve the detection probability, even in an environment that is noisy. A visual representation of the outcome is created, and it is recorded for use in the future as a radar dataset. In order to model the radar system, it is necessary to ascertain the design specifications listed in Table 1. Table 1Design specifications of the radar system.Design parametersDescriptionProbability of detection, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{d}$$\end{document}0.9Probability of false alarm, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{fa}$$\end{document}10− 6Maximum unambiguous range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{max}$$\end{document}5000 mRequired range resolution, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{res}$$\end{document}50 mRequired target radar cross section, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sigma\:$$\end{document}1 m2Operating frequency, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f$$\end{document}10 GHzType of antenna arrayURAScanning range−90° to + 90°Type of targetsNon-fluctuatingTarget detection schemeNoncoherentPropagation channelFree space Table 1Design specifications of the radar system.Design parametersDescriptionProbability of detection, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{d}$$\end{document}0.9Probability of false alarm, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{fa}$$\end{document}10− 6Maximum unambiguous range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{max}$$\end{document}5000 mRequired range resolution, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{res}$$\end{document}50 mRequired target radar cross section, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sigma\:$$\end{document}1 m2Operating frequency, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:f$$\end{document}10 GHzType of antenna arrayURAScanning range−90° to + 90°Type of targetsNon-fluctuatingTarget detection schemeNoncoherentPropagation channelFree space Design specifications of the radar system. A noncoherent detection approach is employed in the system because coherent detection is computationally expensive and requires phase information. Furthermore, the free space environment is assumed in this scenario for simplicity. The proposed model employs a rectangular waveform for its design. The required range resolution subsequently influences the bandwidth of the rectangular waveform, thereby determining its pulse width. The pulse repetition frequency (PRF) is a crucial parameter of a pulse waveform, and it is dictated by the maximum unambiguous range. Finally, the sampling rate is established at double the signal bandwidth of the rectangular waveform, in accordance with Nyquist’s sampling criterion. The pulse bandwidth \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:BW$$\end{document} can be obtained by49:51\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:BW=\frac{c}{2{R}_{res}}$$\end{document} The pulse repetition frequency \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:PRF$$\end{document} can be obtained by49:52\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:PRF=\frac{c}{2{R}_{max}}$$\end{document} It is assumed for simplicity that the receiver experiences only thermal noise throughout the simulation. The receiver’s bandwidth and the level of thermal noise are corresponded. The receiver’s noise bandwidth and waveform bandwidth are set to be equal. The noise bandwidth \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{BW}_{v}$$\end{document} can be expressed as50:53\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{BW}_{v}=BW$$\end{document} Furthermore, it is presumed that the receiver’s noise figure is 0 decibels (dB). Thus, the noise power at the receiver \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{v}$$\end{document} can be expressed as50:54\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{v}={k}_{B}\left({BW}_{v}\right){T}_{v}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{k}_{B}$$\end{document} represents the Boltzmann’s constant and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{v}$$\end{document} represents the system noise temperature. Now, for a single antenna element (isotropic antenna), the received signal power \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{r}$$\end{document} of the scanning radar is given by1:55\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{r}=\left[\frac{{P}_{t}\tau\:{G}_{t}\sigma\:}{{\left(4\pi\:\right)}^{2}{R}^{4}}\right]{A}_{e}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{t}$$\end{document} represents the peak transmit power, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\tau\:$$\end{document} represents the duty cycle of the transmitter, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{t}$$\end{document} represents the transmitter gain, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\sigma\:$$\end{document} represents the radar cross-section of the target, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R$$\end{document} represents the range of the target, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{e}$$\end{document} represents the effective area of the antenna. The effective area of the antenna can be expressed as49:56\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{e}=\frac{{G}_{r}{\lambda\:}^{2}}{4\pi\:}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{r}$$\end{document} denotes the receiver gain, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\lambda\:$$\end{document} denotes the wavelength of the signal. Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{A}_{e}$$\end{document} from Eq. (56) in Eq. (55), it follows that1:57\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{r}=\left[\frac{{P}_{t}\tau\:{G}_{t}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]{G}_{r}$$\end{document} The required SNR for the scanning radar system is obtained by Albersheim’s equation. Hence, the required SNR in dB is expressed as51:58\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{dB}=-5{\text{log}}_{10}\left({N}_{p}\right)+\left(6.2+\frac{4.54}{\sqrt{{N}_{p}+0.44}}\right){\text{log}}_{10}\left(\stackrel{\sim}{A}+0.12\stackrel{\sim}{A}\stackrel{\sim}{B}+1.7\stackrel{\sim}{B}\right)$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{p}$$\end{document} represents the number of pulses that are being integrated and59\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\stackrel{\sim}{A}=\text{ln}\left(\frac{0.62}{{P}_{fa}}\right)$$\end{document}60\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\stackrel{\sim}{B}=\text{ln}\left(\frac{{P}_{d}}{1-{P}_{d}}\right)$$\end{document} Let us consider that \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{dB}$$\end{document} be the minimum SNR required in dB denoted as \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{min\_dB}$$\end{document}. Thus, the required peak transmit power of the radar system for the single antenna element \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{t}$$\end{document} can be formulated as1:61\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{P}_{t}=\frac{{\left(4\pi\:\right)}^{3}{R}_{max}^{4}\left({SNR}_{min}\right){k}_{B}\left({BW}_{v}\right){T}_{v}}{{{G}_{t}^{2}\lambda\:}^{2}\sigma\:}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{SNR}_{min}$$\end{document} is the minimum SNR required in linear form that indicates the threshold above which the radar can confidently distinguish the target signal from noise. It directly influences how much peak power is needed for effective detection. Modified radar equations for the phased antenna array For the rectangular array antenna system, the gain of the receiving antenna \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{r}$$\end{document} can be expressed in terms of the array factor \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathcal{Y}}_{\theta\:}$$\end{document}. The array factor represents the directional gain provided by the array in the specific azimuth direction \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document}. Considering the transmitting and receiving arrays are the same (monostatic radar), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{r}$$\end{document} in Eq. (57) can be replaced by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}$$\end{document}. Therefore, the received signal power expression for the phased array monostatic radar considering rectangular array antenna system \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r}$$\end{document} can be written as:62\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{t}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{a}$$\end{document} represent the peak power and the array gain of the rectangular array antenna system, respectively. In this analysis, for noncoherent signal processing, the array gain (in linear form) is approximated as52:63\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{G}_{a}=M\times\:N$$\end{document} Now, following the process of Eq. (5), we can reform the term \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}$$\end{document} in Eq. (62) as:64\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}={\mathcal{Y}}_{\theta\:}{{\mathcal{Y}}_{\theta\:}}^{*}$$\end{document} Substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mathcal{Y}}_{\theta\:}$$\end{document} from Eq. (35) in Eq. (64), it becomes:65\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}={\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}$$\end{document} Then, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|{\mathcal{Y}}_{\theta\:}\right|}^{2}$$\end{document} from Eq. (65) in Eq. (62), we obtain the modified received power expression in terms of the weight vector and steering vector for the phased array monostatic radar considering rectangular array antenna system as:66\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]\left[{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}\right]$$\end{document} The adaptive weight vector \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}$$\end{document} adjusts the received signal to maximize the power of the desired signal and minimize interference and noise. The term \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}$$\end{document} represents the effective gain in the direction of the desired signal after adaptive beamforming. This formulation reflects how adaptive beamforming enhances the received power from a target direction while mitigating the effects of noise and interference. This time, the required peak transmit power of the radar system for the rectangular array antenna system can be written as1:67\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{t}=\frac{{\left(4\pi\:\right)}^{3}{R}_{max}^{4}\left({SNR}_{min}\right){k}_{B}\left({BW}_{v}\right){T}_{v}}{{{G}_{t}^{2}{G}_{a}^{2}\lambda\:}^{2}\sigma\:}$$\end{document} For the conventional beamformer, the received power expression can be written as:68\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r\_conv}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]\left[{\left({\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{conv}\right)}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{conv}\right]$$\end{document} Thus, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{conv}$$\end{document} from Eq. (36) in Eq. (68), it follows that69\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\begin{aligned} \tilde{P}_{{r\_conv}} & = \left[ {\frac{{\tilde{P}_{t} \tau G_{t} G_{a}^{2} \lambda ^{2} \sigma }}{{\left( {4\pi } \right)^{3} R^{4} }}} \right]\left[ {\left( {\frac{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|}}} \right)^{H} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } ^{H} \left( {\frac{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|}}} \right)} \right] \\ & = \left[ {\frac{{\tilde{P}_{t} \tau G_{t} G_{a}^{2} \lambda ^{2} \sigma }}{{\left( {4\pi } \right)^{3} R^{4} }}} \right]\left[ {\left( {\frac{1}{{\left\| {\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right\|}}} \right)^{2} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } ^{H} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} } \right] \\ \end{aligned}$$\end{document} For the MVDR beamformer, the received power expression can be written as:70\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r\_MVDR}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]\left[{\left({\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{MVDR}\right)}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{MVDR}\right]$$\end{document} Thus, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{MVDR}$$\end{document} from Eq. (38) in Eq. (70), it follows that71\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\tilde{P}_{{r\_MVDR}} = \left[ {\frac{{\tilde{P}_{t} \tau G_{t} G_{a}^{2} \lambda ^{2} \sigma }}{{\left( {4\pi } \right)^{3} R^{4} }}} \right]\left[ {\left( {\frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}} \right)^{H} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } ^{H} \left( {\frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{0} }}} \right)} \right]$$\end{document} For the LCMV beamformer, the received power expression can be written as:72\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r\_LCMV}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]\left[{\left({\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{LCMV}\right)}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{LCMV}\right]$$\end{document} Thus, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{LCMV}$$\end{document} from Eq. (39) in Eq. (72), it follows that73\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\tilde{P}_{{r\_LCMV}} = \left[ {\frac{{\tilde{P}_{t} \tau G_{t} G_{a}^{2} \lambda ^{2} \sigma }}{{\left( {4\pi } \right)^{3} R^{4} }}} \right]\left[ {\left( {\left( {\frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} } \right)^{H} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} _{\theta } ^{H} \left( {\left( {\frac{{\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} ~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} ^{H} \overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } ^{{ - 1}} \underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} } \right)} \right]$$\end{document} For the proposed hybrid robust beamformer, the received power expression can be written as:74\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{P}}_{r\_hybrid}=\left[\frac{{\stackrel{\sim}{P}}_{t}\tau\:{G}_{t}{G}_{a}^{2}{\lambda\:}^{2}\sigma\:}{{\left(4\pi\:\right)}^{3}{R}^{4}}\right]\left[{\left({\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}\right)}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s}}_{\theta\:}}^{H}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}\right]$$\end{document} Thus, substituting \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{w}}_{hybrid}$$\end{document} from Eq. (45) in Eq. (74), it follows that75\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$${\tilde {P}_{r\_hybrid}}=\left[ {\frac{{{{\tilde {P}}_t}\tau {G_t}G_{a}^{2}{\lambda ^2}\sigma }}{{{{\left( {4\pi } \right)}^3}{R^4}}}} \right]\left[ \begin{gathered} {\left( \begin{gathered} \beta \left\{ {\frac{{{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}}}{{{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}^{H}{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}}}} \right\} \hfill \\ +\left( {1 - \beta } \right)\left\{ {\left( {\frac{{{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }^H}{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} } \right\} \hfill \\ \end{gathered} \right)^H} \hfill \\ {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_\theta }{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_\theta }^{H}\left( \begin{gathered} \beta \left\{ {\frac{{{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}}}{{{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}^{H}{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{s} }_0}}}} \right\} \hfill \\ +\left( {1 - \beta } \right)\left\{ {\left( {\frac{{{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}~\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}{{{{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }^H}{{\left( {\overset{\lower0.5em\hbox{$\smash{\scriptscriptstyle\smile}$}}{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{\Gamma } } +\gamma {{\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{I} }_{MN}}} \right)}^{ - 1}}\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{C} }}} \right)\underset{\raise0.3em\hbox{$\smash{\scriptscriptstyle-}$}}{g} } \right\} \hfill \\ \end{gathered} \right) \hfill \\ \end{gathered} \right]$$\end{document} The final received power expression for the proposed hybrid robust beamformer is represented by Eq. (75). It incorporates the optimal weight vector derived from both the diagonal loading MVDR and LCMV beamformers, as given by Eq. (45). This formulation allows the beamformer to adaptively balance the contributions of both methods, providing enhanced robustness against mismatches and interference while maximizing the received power in the direction of the desired signal. By integrating both beamforming techniques, the hybrid system is designed to perform optimally under varying conditions, offering a superior compromise between interference suppression and system reliability. Radar signal processing Radar signal processing involves pulse synthesis, matched filtering, pulse integration, detection threshold selection, and Doppler processing. Pulse synthesis: In order to improve the ability of radar systems to detect targets by optimizing the pulse waveform, pulse synthesis is applied to the radar signal processing for the proposed system. In pulse synthesis, fast time grid is used to resolve the range of the target. The fast time variable is typically associated with the sampling of the radar return signal at regular intervals after each transmitted pulse. The fast time samples are taken within the duration of each pulse. For a pulse width \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{t}_{p}$$\end{document}, the fast time grid can be defined as52:76\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{t}_{fast}\left[n\right]=n\varDelta\:t\:;n=\text{0,1},2,\dots\:,{N}_{t}-1$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{t}_{fast}\left[n\right]$$\end{document} represents the fast time samples (time within a single pulse), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:n$$\end{document} represents the sample index within the pulse, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:t$$\end{document} represents the sampling interval in fast time which is determined by the radar’s sampling rate \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{f}_{s}$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{t}$$\end{document} represents the total number of samples in a single pulse, which depends on the pulse width \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{t}_{p}$$\end{document} and sampling rate \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{f}_{s}$$\end{document}. The sampling interval \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:t$$\end{document} can be defined as50:77\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\varDelta\:t=\frac{1}{{f}_{s}}$$\end{document} The total number of samples in a single pulse \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{t}$$\end{document} can be defined as52:78\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{t}=\frac{{t}_{p}}{\varDelta\:t}$$\end{document} The fast time grid is used for range resolution, as each sample corresponds to a specific range bin which is defined as52:79\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R\left[n\right]=\frac{c\times\:{t}_{fast}\left[n\right]}{2}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R\left[n\right]$$\end{document} denotes the range to the target for the nth sample, and the factor of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:2$$\end{document} accounts for the round-trip travel time of the radar pulse. The free space path loss should be considered in radar signal processing for the time varying gain calculation. The expression for the free space path loss in dB (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{FSPL}_{dB}$$\end{document}) can be derived from the Friis transmission equation which is as follows1:80\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{FSPL}_{dB}=20{\text{log}}_{10}\left(R\right)+20{\text{log}}_{10}\left(f\right)+20{\text{log}}_{10}\left(\frac{4\pi\:}{c}\right)$$\end{document} Therefore, the time varying gain \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:TVG$$\end{document} (in dB) is calculated as follows1:81\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:TVG={L}_{ref}-{L}_{range}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{L}_{ref}$$\end{document} denotes the reference loss and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{L}_{range}$$\end{document} denotes the range loss. They both are the functions of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{FSPL}_{dB}$$\end{document}1. Matched filtering: Matched filtering is a crucial method in radar signal processing as it enhances target identification by optimizing the signal-to-noise ratio (SNR), boosting range resolution via pulse compression, and increasing the radar’s capacity to differentiate between closely situated targets or targets amongst noise. The implementation is essential for providing accurate and reliable target detection in modern radar systems. Matched filtering involves convolving the received signal \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:r\left(t\right)$$\end{document} with the matched filter impulse response \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:h\left(t\right)$$\end{document}. In the discrete domain, this can be represented as52:82\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:y\left[n\right]=\sum\:_{k=0}^{Q-1}r\left[k\right]\bullet\:h\left[n-k\right]$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:y\left[n\right]$$\end{document} is the filtered output at index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:n$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:r\left[k\right]$$\end{document} is the received signal at time index \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:k$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:h\left[n-k\right]$$\end{document} represents the matched filter shifted by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:n-k$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:Q$$\end{document} denotes the number of samples or the length of the signals involved. This equation describes convolution between the received signal and the matched filter. Pulse integration: For the proposed scanning radar system, the received signal is first passed through the matched filter, and then all pulses are integrated for each scan angle in order to process the received signal. The pulse integration technique is applied to improve SNR by summing the returns from multiple pulses, enhancing the radar’s ability to detect distant targets. The expression for pulse integration in noncoherent signal processing is given by52:83\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{y}_{int}=\sum\:_{i=1}^{{N}_{p}}\left|{y}_{i}\right|$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{y}_{int}$$\end{document} represents the integrated output after noncoherent integration, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{y}_{i}$$\end{document} are the individual pulse outputs after matched filtering, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{p}$$\end{document} represents the number of pulses that are being integrated. Detection threshold selection: The time varying gains are applied to the received signal in order to compensate for the signal power loss caused by range. After that the threshold detection selection process is done to obtain an accurate estimation of the target parameters. For noncoherent samples in white Gaussian noise, detection of a nonzero mean leads to a square-law detector. The linear detection threshold \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{T}_{D}$$\end{document} for the noncoherent Neyman-Pearson (NP) detector is given by52:84\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\frac{{T}_{D}}{\eta\:}=\sqrt{{\psi\:}^{-1}\left({N}_{p},1-{P}_{fa}\right)}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\eta\:$$\end{document} denotes the standard deviation the noise, ψ-1(❚,❚) denotes the inverse of the lower incomplete gamma function, and the term \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\frac{{T}_{D}}{\eta\:}$$\end{document} indicates the detection threshold scaled relative to the noise standard deviation. The threshold can be expressed as a signal-to-noise ratio in dB which is as follows52:85\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:10{{log}}_{10}\left(\frac{{{T}_{D}}^{2}}{{\eta\:}^{2}}\right)=10{{log}}_{10}\left\{{\psi\:}^{-1}\left({N}_{p},1-{P}_{fa}\right)\right\}$$\end{document} Doppler processing: Doppler velocity measurement of pulse radars uses the Doppler effect of relative motion of an object to obtain the change rate of range. According to the Doppler theorem, radial velocity hence the component of velocity toward or away from the radar (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D}$$\end{document}) and the maximum unambiguous Doppler velocity (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D\_max}$$\end{document}) are given as follows1:86\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D}={f}_{D}\times\:\frac{\lambda\:}{2}$$\end{document}87\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D\_max}=PRF\times\:\frac{\lambda\:}{4}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{f}_{D}$$\end{document} represents the Doppler frequency shift. Algorithm of the simulation process The algorithm for the simulation process of the proposed hybrid robust beamformer, which is performed in the MATLAB53 software environment, can be structured as follows: Algorithm 1 Hybrid robust beamformer. Hybrid robust beamformer. Overall accuracy measurement parameters The relative error for the range \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{R}$$\end{document} is given by7:88\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{R}=\frac{\left|{R}_{m}-{R}_{t}\right|}{{R}_{t}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}$$\end{document} denotes the range measured and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}$$\end{document} denotes the true range of the target. Thus, the accuracy for the range \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document} can be defined as7:89\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}=1-{e}_{R}$$\end{document} Similarly, the relative error for the angle \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{\theta\:}$$\end{document} is given by7:90\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{\theta\:}=\frac{\left|{\theta\:}_{m}-{\theta\:}_{t}\right|}{{\theta\:}_{t}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}$$\end{document} denotes the angle measured and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}$$\end{document} denotes the true angle of the target. Thus, the accuracy for the angle \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document} can be defined as7:91\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}=1-{e}_{\theta\:}$$\end{document} Similarly, the relative error for the velocity \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{{v}_{D}}$$\end{document} is given by2:92\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{e}_{{v}_{D}}=\frac{\left|{{v}_{D}}_{m}-{{v}_{D}}_{t}\right|}{{{v}_{D}}_{t}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} denotes the velocity measured and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} denotes the true velocity of the target. Thus, the accuracy for the velocity \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document} can be defined as2:93\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}=1-{e}_{{v}_{D}}$$\end{document} Therefore, the overall accuracy \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{overall}$$\end{document} can be measured as follows3:94\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{overall}=\frac{{\mu\:}_{R}+{\mu\:}_{\theta\:}+{\mu\:}_{{v}_{D}}}{3}$$\end{document} In Eq. (94), equal weighting for range (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}), angle (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}), and Doppler velocity (\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}) accuracies is assumed to provide a simple and balanced assessment of the radar system’s performance across its core measurement domains. This assumption is considered appropriate in the context of general-purpose surveillance radars, which often require balanced detection capabilities across spatial and kinematic parameters to maintain comprehensive situational awareness. However, in mission-specific scenarios (e.g., tracking fast-moving targets or identifying stationary objects), the relative importance of these parameters may vary. To address this, a modification of Eq. (94) can be done by introducing weights to the range, angle, and velocity accuracies that demonstrates how the overall accuracy changes when different weights are assigned to the components. Thus, the modified version of Eq. (94) is as follows1,54:95\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{\mu\:}}_{overall}={w}_{R}{\mu\:}_{R}+{w}_{\theta\:}{\mu\:}_{\theta\:}+{w}_{{v}_{D}}{\mu\:}_{{v}_{D}}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\stackrel{\sim}{\mu\:}}_{overall}$$\end{document} denotes the overall accuracy for the mission-specific scenarios, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{w}_{R}$$\end{document} denotes the weight for range accuracy, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{w}_{\theta\:}$$\end{document} denotes the weight for angle accuracy, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{w}_{{v}_{D}}$$\end{document} denotes the weight for velocity accuracy. The relationship among these three weights is given by54:96\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{w}_{R}+{w}_{\theta\:}+{w}_{{v}_{D}}=1$$\end{document} In this paper, a general-purpose surveillance radar is considered and so Eq. (94) is utilized in order to analyze the performance of the proposed radar. Experimentation process For the experimentation, a uniform linear array (ULA)-based phased array scanning radar system with conventional beamforming is employed in RUET Radar & Satellite Laboratory. From Fig. 7, one can visualize and understand the process of working of the phased array radar system in the laboratory. The two targets, half-cylindrical in shape, are placed in a target table and their trajectory pattern is circular. The target table is adjusted perpendicular to the line of sight (LOS) of the phased array antenna (PAA). \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{AT}$$\end{document} denotes the distance between the PAA and the target table. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{max}$$\end{document} denotes the maximum range for the detection. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{x}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{y}$$\end{document} denote the distance between the two targets, and the distance from the PAA to the targets, respectively. From Fig. 7, one can also observe the scanning range and the radiation pattern of the PAA. The beam is steered electronically without the rotation of the PAA mechanically. The PAA can generate a beam (Beam 0 to Beam 15) at different angles within the scanning range. The system parameters for the experimentation are documented in Table 2. Fig. 7Visualization of the experimentation process of the phased array scanning radar system. Visualization of the experimentation process of the phased array scanning radar system. Table 2System parameters for the experimentation.System parametersDescriptionAntenna array typeULANo. of antenna elements16Operating frequency9.4 GHzScanning range−45° to +45°Radiated power density0.8 watts/m2Gain4Maximum range3.6 mTrajectory of targetsCircularBeam scanning mode for range and angle estimationContinuous (270 scan/min)Beam scanning mode for velocity estimationContinuous (54 scan/min)Single PRF288 HzBeam sequenceIncremental (Beam 0 to Beam 15) Table 2System parameters for the experimentation.System parametersDescriptionAntenna array typeULANo. of antenna elements16Operating frequency9.4 GHzScanning range−45° to +45°Radiated power density0.8 watts/m2Gain4Maximum range3.6 mTrajectory of targetsCircularBeam scanning mode for range and angle estimationContinuous (270 scan/min)Beam scanning mode for velocity estimationContinuous (54 scan/min)Single PRF288 HzBeam sequenceIncremental (Beam 0 to Beam 15) System parameters for the experimentation. Experimental setup The experimental setup, shown in Fig. 8, is organized according to the system visualization as discussed elaborately in Section ‘Experimentation Process’. The moving target indication (MTI) processor is used here for the detection of moving targets. The phased array antenna controller is employed here for the beam steering control of the PAA. It enables the three scan modes of manual, continuous, and PRF locked (depending on the radar PRF) to be operated with the PAA. In order to facilitate further processing and display, the Dual-channel sampler expands the time of the radar receiver’s I (in-phase) and Q (quadrature) channel baseband signals. Quadrature modulation and demodulation are performed in it. The radar transmitter with complex modulation can generate waveforms utilizing I/Q signals that are visualized on the oscilloscope display. It is also possible to extract the target’s Doppler information from complex data derived from demodulated I/Q signals. The range, azimuth angle, and speed of the multiple targets can be estimated from the radar display. From Fig. 9, one can observe the configuration of the setup of targets on the target table in front of the PAA. The height \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{h}_{HC}$$\end{document} of the two targets is 10 cm = 0.1 m and the curvature length \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{c}_{HC}$$\end{document} is 5 cm = 0.05 m. The curvature of a half-cylinder is the circumference of a semi-circle. If the radius of a half-cylinder is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{r}_{HC}$$\end{document}, the equation for the curvature of the half-cylinder \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{c}_{HC}$$\end{document} is given by55:97\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{c}_{HC}=\pi\:{r}_{HC}$$\end{document} Thus, the RCS of the half cylindrical targets \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{HC}$$\end{document} (m2) is given by55:98\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{HC}=\pi\:{r}_{HC}\left({r}_{HC}+{h}_{HC}\right)$$\end{document} From Eqs. (97) and (98), the RCS of the half cylindrical targets can be obtained. For experimentation, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{D}_{AT}$$\end{document} is set as 140 cm = 1.4 m. Thus, the targets will be moving back and forth between 1.4 m and 3.6 m (maximum range) and the PAA will detect them by the electronic beam steering. The target controller of the experimental setup is used to control the position, speed, and trajectory of the targets. Fig. 8Full experimental phased array radar setup using FESTO (LabVolt) radar training system in RUET radar & satellite laboratory. Full experimental phased array radar setup using FESTO (LabVolt) radar training system in RUET radar & satellite laboratory. Fig. 9Configuration of the targets for the experimentation. Configuration of the targets for the experimentation. Results and analysis Simulation results of the ULA-based radar with the conventional beamformer In the simulation process, two objects are assumed to be present within the surveillance area of the ULA-based phased array radar system. From Fig. 10, one can observe the scan map of the ULA-based phased array radar system utilizing the conventional beamformer. The scanning range of the radar system is − 45° to + 45° which is noticeable in Fig. 10. In the scanning region, two yellow blips are appeared that indicates the two objects to be identified. The position of the two objects is represented in a three-dimensional (3D) Cartesian coordinate system in which the position of Object-1 and Object-2 is (0.97 m, 2.83 m, 0 m) and (0.36 m, 2.97 m, 0 m) respectively. Likewise, two objects at different positions are detected by the system and the tracking parameters such as range, angle, and velocity are obtained in the MATLAB53 software environment documented in Table 3. All the system parameters in this process are set according to the configuration of the FESTO (LabVolt) Radar Training System. Moreover, the average accuracy for range, angle, and velocity estimation is displayed in Table 4. Fig. 10Scan map of the ULA-based phased array radar system utilizing the conventional beamformer by simulation. Scan map of the ULA-based phased array radar system utilizing the conventional beamformer by simulation. Table 3Tracking parameters (range, angle, and velocity) of two objects by simulation for the conventional beamformer.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Act. velocity (cm/s)Est. velocity (cm/s)Object-110.800.652.232.252.36922.342019.735216.113410.210.920.840.782.342.322.48622.447619.746818.58311.412.030.920.892.392.372.56102.531621.053620.582512.513.240.870.862.462.502.60932.643819.476618.983213.814.550.850.832.492.532.63112.662718.848118.16281515.960.910.702.582.632.73582.721619.428414.904316.717.370.940.732.692.722.84952.816319.261615.023218.619.181.000.972.802.832.97322.991619.653818.919520.421.190.900.892.912.853.04602.985717.185717.342623.524.0100.830.952.992.923.10313.070715.514218.021924.525.3Object-210.10−0.012.212.152.21232.15002.5908−0.266511.612.720.140.122.302.232.30432.23323.48333.080213.214.630.220.202.372.342.38022.34855.30344.885216.417.540.170.152.422.392.42602.39474.01833.591319.420.150.150.132.462.422.46462.42353.48933.074921.322.660.210.182.542.512.54872.51644.72634.101825.625.970.240.212.672.642.68082.64835.13644.54826.727.080.300.363.012.973.02492.99175.69176.911227.227.690.200.162.872.842.87702.84453.98633.224528.128.6100.130.072.962.922.96292.92082.51481.373329.630.0 Table 3Tracking parameters (range, angle, and velocity) of two objects by simulation for the conventional beamformer.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Act. velocity (cm/s)Est. velocity (cm/s)Object-110.800.652.232.252.36922.342019.735216.113410.210.920.840.782.342.322.48622.447619.746818.58311.412.030.920.892.392.372.56102.531621.053620.582512.513.240.870.862.462.502.60932.643819.476618.983213.814.550.850.832.492.532.63112.662718.848118.16281515.960.910.702.582.632.73582.721619.428414.904316.717.370.940.732.692.722.84952.816319.261615.023218.619.181.000.972.802.832.97322.991619.653818.919520.421.190.900.892.912.853.04602.985717.185717.342623.524.0100.830.952.992.923.10313.070715.514218.021924.525.3Object-210.10−0.012.212.152.21232.15002.5908−0.266511.612.720.140.122.302.232.30432.23323.48333.080213.214.630.220.202.372.342.38022.34855.30344.885216.417.540.170.152.422.392.42602.39474.01833.591319.420.150.150.132.462.422.46462.42353.48933.074921.322.660.210.182.542.512.54872.51644.72634.101825.625.970.240.212.672.642.68082.64835.13644.54826.727.080.300.363.012.973.02492.99175.69176.911227.227.690.200.162.872.842.87702.84453.98633.224528.128.6100.130.072.962.922.96292.92082.51481.373329.630.0 Tracking parameters (range, angle, and velocity) of two objects by simulation for the conventional beamformer. From Fig. 11, one can observe the change in ranges by varying positions of the two objects. The actual and estimated ranges are closely valued for Object-1, and no appreciable divergence is seen. However, for Object-2, a slight variance between the actual and estimated values occurs within the positions of 2.3 m and 2.6 m. The change in angle caused by changing the position of the targets is depicted in Fig. 12. Between the actual and estimated angles for Object-2, there is no noticeable distinction as before, but for Object-1, a noticeable difference is observed between 2.7 m and 3 m. The underlying reason for this deviation is that the declining performance at extreme positions is attributed to the falling gain response of the PAA at larger scanning angles (toward ± 45°), which reduces signal strength and hence angular resolution4. Hence, the PAA’s detection accuracy declined as azimuth angles increased. This result emphasizes a limitation of the conventional beamformer when operating near the edges of the array’s scanning region, especially under signal degradation due to increased steering angles. The change in velocities by varying positions of the two objects can be observed from Fig. 13. The velocity range for the two objects is kept 10 cm/s to 30 cm/s which is not so fast in the context of reality. For this reason, the actual and estimated velocities are closely valued for both Object-1 and Object-2, hence, no significant variation is observed. Fig. 11Changes in range by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Changes in range by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Fig. 12Changes in angle by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Changes in angle by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Fig. 13Changes in velocity by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Changes in velocity by varying positions of Object-1 and Object-2 at 0° of elevation in simulation. Table 4Average accuracy of tracking parameters of two objects by simulation for the conventional beamformer.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (cm/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (cm/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-12.36922.34200.988519.735216.11340.816510.210.90.93142.48622.44760.984519.746818.5830.941111.412.00.94742.56102.53160.988521.053620.58250.977612.513.20.9442.60932.64380.986819.476618.98320.974713.814.50.94932.63112.66270.98818.848118.16280.96361515.90.942.73582.72160.994819.428414.90430.767116.717.30.96412.84952.81630.988319.261615.02320.7818.619.10.97312.97322.99160.993819.653818.91950.962620.421.10.96573.04602.98570.980217.185717.34260.990923.524.00.97873.10313.07070.989615.514218.02190.838424.525.30.9673Object-22.21232.15000.97192.5908−0.2665−0.102911.612.70.90522.30432.23320.96923.48333.08020.884313.214.60.89392.38022.34850.98675.30344.88520.921116.417.50.93292.42602.39470.98714.01833.59130.893719.420.10.96392.46462.42350.98333.48933.07490.881221.322.60.9392.54872.51640.98744.72634.10180.867925.625.90.98832.68082.64830.98795.13644.5480.885526.727.00.98883.02492.99170.9895.69176.91120.785727.227.60.98532.87702.84450.98873.98633.22450.808928.128.60.98222.96292.92080.98582.51481.37330.546129.630.00.9865Average accuracy for range estimation0.9860Average accuracy for angle estimation0.8192Average accuracy for velocity estimation0.9564 Table 4Average accuracy of tracking parameters of two objects by simulation for the conventional beamformer.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (cm/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (cm/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-12.36922.34200.988519.735216.11340.816510.210.90.93142.48622.44760.984519.746818.5830.941111.412.00.94742.56102.53160.988521.053620.58250.977612.513.20.9442.60932.64380.986819.476618.98320.974713.814.50.94932.63112.66270.98818.848118.16280.96361515.90.942.73582.72160.994819.428414.90430.767116.717.30.96412.84952.81630.988319.261615.02320.7818.619.10.97312.97322.99160.993819.653818.91950.962620.421.10.96573.04602.98570.980217.185717.34260.990923.524.00.97873.10313.07070.989615.514218.02190.838424.525.30.9673Object-22.21232.15000.97192.5908−0.2665−0.102911.612.70.90522.30432.23320.96923.48333.08020.884313.214.60.89392.38022.34850.98675.30344.88520.921116.417.50.93292.42602.39470.98714.01833.59130.893719.420.10.96392.46462.42350.98333.48933.07490.881221.322.60.9392.54872.51640.98744.72634.10180.867925.625.90.98832.68082.64830.98795.13644.5480.885526.727.00.98883.02492.99170.9895.69176.91120.785727.227.60.98532.87702.84450.98873.98633.22450.808928.128.60.98222.96292.92080.98582.51481.37330.546129.630.00.9865Average accuracy for range estimation0.9860Average accuracy for angle estimation0.8192Average accuracy for velocity estimation0.9564 Average accuracy of tracking parameters of two objects by simulation for the conventional beamformer. RCS validation via physical optics approximation To validate the assumed 1 m² RCS of the half-cylinder target used in this study, a simulation is performed using a Physical Optics (PO)-based approximation in MATLAB53 software. The monostatic RCS is approximated using the PO-based method, mathematically expressed as1,56:99\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\sigma\:}_{HC}=\left(4\pi\:{r}_{HC}^{2}\bullet\:\frac{1}{{N}_{s}}\sum\:_{i=1}^{{N}_{s}}{\left|\widehat{k}\bullet\:{\widehat{n}}_{i}\right|}^{4}\right)\bullet\:\frac{{h}_{HC}}{\lambda\:}$$\end{document} where \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{N}_{s}$$\end{document} represents the number of surface discretization points along the arc and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\widehat{k}$$\end{document} represents the unit vector in the direction of wave incidence. This equation estimates the RCS of a half-cylinder using a simplified PO approximation. The term \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\left|\widehat{k}\bullet\:{\widehat{n}}_{i}\right|}^{2}$$\end{document} represents the cosine of the angle between the incident wave and surface normal, which is a common factor in PO-based RCS models and squaring that again (i.e., raising to 4) emphasizes strong scattering directions at broadside. Fig. 14Angular RCS pattern of the required 1 m2 half-cylinder target at broadside incidence. Angular RCS pattern of the required 1 m2 half-cylinder target at broadside incidence. The half-cylinder is modeled with a radius of 0.08 m and height of 1.0 m, illuminated at broadside under monostatic conditions at 10 GHz. The resulting RCS is found to be approximately 1.014 m² (0.06 dBsm), which supports the modeling assumption and is consistent with canonical target behavior reported in1,55. From Fig. 14, one can observe the angular RCS pattern of a half-cylinder target at broadside incidence. The red curve in Fig. 14 represents the angular variation of the RCS of the half-cylinder target as observed from different angles. Specifically, it shows how the RCS changes as the radar beam scans around the target. The plot typically reflects the RCS at broadside incidence (0° or 180°), where the target presents its maximum reflective area to the radar, and the values decrease as the radar moves away from the broadside position. The curve’s peaks and troughs represent the angular distribution of the target’s RCS, with the highest point usually corresponding to the broadside orientation of the half-cylinder. Experimentation results of the ULA-based radar with the conventional beamformer The experimentation results are required to validate the simulation results utilizing the conventional beamformer. So, in the experimentation process, two objects are assumed to be present within the surveillance area of the ULA-based phased array radar system. The whole experimentation work is accomplished by utilizing the FESTO (LabVolt) Radar Training System in RUET Radar & Satellite Laboratory. Fig. 15 depicts the measurement of range and azimuth angle at 0° of elevation of the moving objects from the radar display of the experimental setup. As seen in Fig. 15, two square-shaped and lawn-green colored blips are an indication of two targets on the radar display. Target-1 and Target-2 are indicated by the left and right blips respectively. The ranges of Target-1 and Target-2 are indicated by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{1}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{2}$$\end{document}, respectively. Likewise, the azimuth angles of Target-1 and Target-2 are indicated by \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{1}$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{2}$$\end{document}, respectively. The values of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{1}$$\end{document} (234 cm), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{2}$$\end{document} (210 cm), \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{1}$$\end{document} (3°), and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{2}$$\end{document} (20°) for moving targets are figured out from the cursor position. Additionally, one can observe in Fig. 15 that there are some fixed blips close to the ring’s center that are considered to be noise due to the PAA’s internal reflection. Furthermore, there are also some random noises noticeable on the radar display which are static and do not create any difficulty during detecting the targets since the targets are moving and continuously changing their positions which are visible on the display. Similarly, other ranges and azimuth angles of the two targets are estimated as the same process from this radar display that are recorded in Table 5. Fig. 15Measurement of range and azimuth angle of the moving objects from the Radar Display during experimentation. Measurement of range and azimuth angle of the moving objects from the Radar Display during experimentation. In order to measure the velocity of a target, let it be considered that the target’s distance from the radar is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R$$\end{document}, and the angular separation between two target echoes from two successive scans is \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document}. The target trajectory is along the side of the right angle if one makes the assumption that the target is moving perpendicular to the radar Line-of-Sight (LOS) demonstrated in Fig. 16. In the right triangle, the tangent equals the length of the side opposite to \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document}, divided by the length of the adjacent side. Since \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document} are known, the distance separating both target echoes within two successive scans, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D$$\end{document} is given by the following equation55:100\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D=2R\text{tan}\left(\frac{\theta\:}{2}\right)$$\end{document} Fig. 16Velocity estimation process for moving targets for the experimentation. Velocity estimation process for moving targets for the experimentation. Then, let it be assumed that \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:t$$\end{document} denotes the interval between successive appearances of the target. The distance equation can be used to estimate the target’s velocity \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D}$$\end{document} if it continues to move at a constant velocity55. Thus, it follows that101\documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{D}=\frac{D}{t}$$\end{document} For experimentation, the velocity is set at 15 cm/s from the target controller of the radar system shown in Fig. 17. The estimation of the velocity of moving targets from the radar display from the experimental setup is depicted in Fig. 17. \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{1}$$\end{document} (345°) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{2}$$\end{document} (349°) denote the left and right blip bearing for Target-1, respectively. Likewise, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{3}$$\end{document} (8°) and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{4}$$\end{document} (12°) denote the left and right blip bearing for Target-2, respectively. Fig. 17Velocity control of targets by the Target Controller of the radar system for the experimentation. Velocity control of targets by the Target Controller of the radar system for the experimentation. Table 5Tracking parameters (range, angle, and velocity) of two objects by experimentation for the conventional beamformer.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Act. velocity (cm/s)Est. velocity (cm/s)Object-110.760.632.212.252.33702.3418.98161110.7020.840.762.362.322.50502.4419.59181211.8030.900.872.422.372.58192.5220.4201412.9040.820.842.492.502.62152.6418.23191514.8550.800.812.542.532.66302.6617.48181615.8460.920.682.672.632.82412.7219.01141817.2270.890.712.782.722.91902.8117.75152018.9380.970.952.862.833.02002.9918.73192221.5490.930.872.972.853.11222.9817.39172524.31100.830.933.102.923.20923.0614.99183028.91Object-210.06−0.022.172.112.17082.111.58−11110.7020.140.112.322.242.32422.243.4531211.8030.200.172.352.272.35852.284.8641412.9040.180.152.462.362.46662.364.1841514.8550.200.102.492.412.49802.414.5921615.8460.220.162.602.532.60932.544.8441817.2270.190.142.712.672.71672.674.0132018.9380.270.292.792.762.80302.785.5362221.5490.230.152.872.842.87922.844.5832524.31100.130.212.962.912.96292.922.5143028.91 Table 5Tracking parameters (range, angle, and velocity) of two objects by experimentation for the conventional beamformer.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Act. velocity (cm/s)Est. velocity (cm/s)Object-110.760.632.212.252.33702.3418.98161110.7020.840.762.362.322.50502.4419.59181211.8030.900.872.422.372.58192.5220.4201412.9040.820.842.492.502.62152.6418.23191514.8550.800.812.542.532.66302.6617.48181615.8460.920.682.672.632.82412.7219.01141817.2270.890.712.782.722.91902.8117.75152018.9380.970.952.862.833.02002.9918.73192221.5490.930.872.972.853.11222.9817.39172524.31100.830.933.102.923.20923.0614.99183028.91Object-210.06−0.022.172.112.17082.111.58−11110.7020.140.112.322.242.32422.243.4531211.8030.200.172.352.272.35852.284.8641412.9040.180.152.462.362.46662.364.1841514.8550.200.102.492.412.49802.414.5921615.8460.220.162.602.532.60932.544.8441817.2270.190.142.712.672.71672.674.0132018.9380.270.292.792.762.80302.785.5362221.5490.230.152.872.842.87922.844.5832524.31100.130.212.962.912.96292.922.5143028.91 Tracking parameters (range, angle, and velocity) of two objects by experimentation for the conventional beamformer. Fig. 18Measurement of velocity of the moving objects from the radar display during experimentation. Measurement of velocity of the moving objects from the radar display during experimentation. From Fig. 18, for Target-1, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R=234\:cm$$\end{document}, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{1}=345^\circ\:=-15^\circ\:$$\end{document}, and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{2}=349^\circ\:=-11^\circ\:$$\end{document}. Thus, the angular separation of the two blips, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:=\left|{\theta\:}_{1}-{\theta\:}_{2}\right|=\left|-15^\circ\:+11^\circ\:\right|=4^\circ\:$$\end{document}. Putting the values of \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:R$$\end{document} and \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\theta\:$$\end{document} in Eq. (100), the distance between the two target blips, which is the distance traveled by the target during one complete scan of the PAA is determined. Thus, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:D=16.3429\:cm.$$\end{document} Since the scan speed is 54 scan/min for velocity estimation, one complete scan is performed every 60/54 = 1.1 s. By using Eq. (101), the velocity of Target-1, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{{D}_{1}}=14.85\:cm/s$$\end{document}. Similarly, the velocity of Target-2 is determined with the help of Eqs. (100) and (101). The velocity of Target-2, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{v}_{{D}_{2}}=14.85\:cm/s$$\end{document}. Since the two targets are moving constantly at the same speed, their velocity is equal. In addition, the estimated velocity is approximately equal to the actual velocity set by the target controller. Similarly, other velocities of the two targets are estimated as the same process from this radar display that are recorded in Table 5. Fig. 19Changes in range by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. Changes in range by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. From Fig. 19, one can observe the change in ranges by varying positions of the two objects. The actual and estimated ranges are closely valued for both Object-1 and Object-2. However, a slight deviation between the actual and estimated values is noticed for both Object-1 and Object-2. The change in angle caused by changing the position of the targets is depicted in Fig. 20. Between the actual and estimated angles for Object-2, there is small distinction is observed, but for Object-1, a noticeable difference is observed between 2.3 m and 2.5 m, and between 2.7 m and 3 m. The underlying reason for this deviation is discussed in the simulation process of the ULA-based radar using conventional beamformer. While the pattern of deviation in Fig. 20 matches the simulation trend from Fig. 12, the magnitude is slightly greater due to unavoidable physical imperfections and environmental noise in the real-world setup. The change in velocities by varying positions of the two objects can be observed from Fig. 21. With the help of target controller of the experimental setup, the velocity range for the two objects is kept 10 cm/s to 30 cm/s which is not so fast in the context of reality. For this reason, the actual and estimated velocities are closely valued for both Object-1 and Object-2, hence, no significant variation is observed. Apart from that, the average accuracy for range, angle, and velocity estimation is obtained in Table 6. Fig. 20Changes in angle by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. Changes in angle by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. Fig. 21Changes in velocity by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. Changes in velocity by varying positions of Object-1 and Object-2 at 0° of elevation in experimentation. Table 6Average accuracy of tracking parameters of two objects by experimentation for the conventional beamformer.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (cm/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (cm/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-12.33702.340.999818.98160.8431110.700.97272.50502.440.974619.59180.91881211.800.98332.58192.520.977820.4200.98041412.900.92142.62152.640.99418.23190.95781514.850.99002.66302.660.997617.48180.97031615.840.99002.82412.720.961919.01140.73651817.220.95672.91902.810.963117.75150.84512018.930.94653.02002.990.988518.73190.98562221.540.97913.11222.980.957517.39170.97762524.310.97243.20923.060.954914.99180.79923028.910.9637Object-22.17082.110.9721.58−1−0.63291110.700.97272.32422.240.96493.4530.86961211.800.98332.35852.280.96524.8640.8231412.900.92142.46662.360.95874.1840.95691514.850.99002.49802.410.96564.5920.43571615.840.99002.60932.540.97154.8440.82641817.220.95672.71672.670.98424.0130.74812018.930.94652.80302.780.99015.5360.9152221.540.97912.87922.840.98784.5830.6552524.310.97242.96292.920.98472.5140.40643028.910.9637Average accuracy for range estimation0.9757Average accuracy for angle estimation0.7509Average accuracy for velocity estimation0.9675 Table 6Average accuracy of tracking parameters of two objects by experimentation for the conventional beamformer.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (cm/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (cm/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-12.33702.340.999818.98160.8431110.700.97272.50502.440.974619.59180.91881211.800.98332.58192.520.977820.4200.98041412.900.92142.62152.640.99418.23190.95781514.850.99002.66302.660.997617.48180.97031615.840.99002.82412.720.961919.01140.73651817.220.95672.91902.810.963117.75150.84512018.930.94653.02002.990.988518.73190.98562221.540.97913.11222.980.957517.39170.97762524.310.97243.20923.060.954914.99180.79923028.910.9637Object-22.17082.110.9721.58−1−0.63291110.700.97272.32422.240.96493.4530.86961211.800.98332.35852.280.96524.8640.8231412.900.92142.46662.360.95874.1840.95691514.850.99002.49802.410.96564.5920.43571615.840.99002.60932.540.97154.8440.82641817.220.95672.71672.670.98424.0130.74812018.930.94652.80302.780.99015.5360.9152221.540.97912.87922.840.98784.5830.6552524.310.97242.96292.920.98472.5140.40643028.910.9637Average accuracy for range estimation0.9757Average accuracy for angle estimation0.7509Average accuracy for velocity estimation0.9675 Average accuracy of tracking parameters of two objects by experimentation for the conventional beamformer. Performance analysis of the ULA-based radar with the conventional beamformer The performance of the ULA-based phased array radar with the conventional beamformer can be carried out by measuring the overall accuracy for object detection and deviation between the results of the simulation and experimentation process. The overall accuracy of both simulation and experimentation process is calculated with the help of Eq. (94) and recorded in Table 7. Table 7Overall accuracy of the simulation and experimentation process for object detection utilizing the conventional beamformer.ProcessAverage accuracy for range estimationAverage accuracy for angle estimationAverage accuracy for velocity estimationOverall accuracySimulation0.98600.81920.95640.9205 or 92.05%Experimentation0.97570.75090.96750.8980 or 89.80% Table 7Overall accuracy of the simulation and experimentation process for object detection utilizing the conventional beamformer.ProcessAverage accuracy for range estimationAverage accuracy for angle estimationAverage accuracy for velocity estimationOverall accuracySimulation0.98600.81920.95640.9205 or 92.05%Experimentation0.97570.75090.96750.8980 or 89.80% Overall accuracy of the simulation and experimentation process for object detection utilizing the conventional beamformer. From Table 7, one can find that the overall accuracy of the simulation and experimentation process is 92.05% and 89.80% respectively. It is obvious that there is no such big deviation between the simulation accuracy and experimentation accuracy, hence, the system performs well in this scenario. However, the maximum unambiguous range for the experimental prototype is very low in terms of real-life application. Thus, when the range of objects is increased, the detection accuracy of the prototype will decline. So, the higher accuracy can be retained in that case if the maximum detectable range and scanning area of the PAA are increased by adding more antenna elements to the array. The experimental prototype is not capable of increasing the object velocity in meter per second range which can be a significant issue for the practical object detection problem. Moreover, the process of interference rejection cannot be performed with this experimental setup and so the overall accuracy is considerably affected with noisier environment. For this reason, a robust algorithm is required for the beamformer used in the radar system so that it can be outperformed the conventional beamformer as well as implemented in practical scenario. Performance analysis of the proposed hybrid robust beamformer In order to analyze the performance of the proposed hybrid robust beamformer, a challenging scenario is considered that can reflect the real-life challenges. In this challenging scenario, there are three sources and two interferences and the environment is kept so noisy. So, the SNR in this scenario is set at 0.01 dB, source signal power is 1 dB, and interference signal power is 100 dB. Here, other five beamformers (Conventional, MVDR, LCMV, MVDR with Diagonal Loading, and LCMV with Diagonal Loading) are chosen to make comparison with the proposed hybrid robust beamformer. Apart from that, the whole process is done with three different number of antenna elements and they are 10 × 10 URA (100 elements), 20 × 20 URA (400 elements), and 30 × 30 URA (900 elements) in order to observe the difference in performance of the proposed hybrid robust beamformer. The simulation is performed in the MATLAB53 software environment. For the proposed beamformer, the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} value is chosen \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:0.2$$\end{document} for the better performance in the given scenario. The process for the selection of the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} value in this research is trial-and-error. However, an adaptive technique can be an effective solution for selecting the \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:\beta\:$$\end{document} value rapidly, which may be considered the further improvement of the proposed beamformer. From Figs. 22, 23, and 24, one can observe the comparison of beam patterns from various beamformers, including Conventional, MVDR, LCMV, MVDR-DL, LCMV-DL, and the proposed hybrid robust beamformer, using a 10 × 10 URA, 20 × 20 URA, and 30 × 30 URA, respectively. Here, the term DL is used as the short form of Diagonal Loading. The beam patterns are normalized with respect to the angle, showing normalized power (dB) as a function of angle in degrees. From Figs. 22, 23, and 24, one can also observe that the conventional beamformer (denoted by the red line) can create main beams to the source directions though there are some look direction errors. However, it cannot create nulls properly at the interference directions. The performance of the simplified MVDR beamformer (denoted by the deep green line) and the simplified LCMV beamformer (denoted by the blue line) are worse than that of the conventional beamformers. These beamformers (MVDR and LCMV) solve an optimization problem that involves inverting the array correlation matrix of the received signals. Due to the lower SNR hence the challenging scenario, the correlation matrix becomes ill-conditioned, meaning that the inversion process introduces large numerical errors, leading to a beam pattern that fails to suppress interference and even steers the beam in the wrong direction. For this reason, these two beamformers cannot provide any appropriate results in the given scenario. The better performance of the conventional beamformer as compared with the MVDR and LCMV beamformers in the challenging scenario is primarily due to its simplicity, which makes its less sensitive to errors in correlation matrix estimation, steering vector mismatches, and other numerical issues. Fig. 22Power response vs. signal direction graph with number of URA elements = 100, source signal power = 1 dB, DoAs = 0⁰, 25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and -50⁰. Power response vs. signal direction graph with number of URA elements = 100, source signal power = 1 dB, DoAs = 0⁰, 25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and -50⁰. Fig. 23Power response vs. signal direction graph with number of URA elements = 400, source signal power = 1 dB, DoAs = 0⁰, -25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and -50⁰. Power response vs. signal direction graph with number of URA elements = 400, source signal power = 1 dB, DoAs = 0⁰, -25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and -50⁰. Fig. 24Power response vs. signal direction graph with number of URA elements = 900, source signal power = 1 dB, DoAs = 0⁰, -25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and − 50⁰. Power response vs. signal direction graph with number of URA elements = 900, source signal power = 1 dB, DoAs = 0⁰, -25⁰ and 45⁰, number of interference = 2, power of interference = 100 dB, and direction of interference signals = 60⁰ and − 50⁰. The ill-condition issue of the correlation matrix is then resolved by adding a regularization factor to the matrix according to the given scenario. As a result, this time, the MVDR-DL beamformer (denoted by the dotted green line) can create main beams to the source signals, however, it cannot create nulls at the interference directions yet due to the challenging scenario. Then, the issue of the MVDR-DL beamformer is mitigated by the LCMV-DL beamformer (denoted by the dotted magenta line), which creates deeper nulls at the interference directions while maintaining its main beams at the source directions. The proposed beamformer (denoted by the black line) can efficiently create its main beam at the signal directions and can also create nulls at the interference directions. From Figs. 22, 23, and 24, it is clear that the LCMV-DL beamformer creates deeper nulls in comparison to the proposed beamformer. However, this advantage of the LCMV-DL beamformer is offset by its higher sidelobe levels, which are more pronounced than those of the proposed beamformer. In contrast, the proposed beamformer exhibits superior sidelobe control, providing better performance in terms of interference suppression while maintaining relatively low sidelobe levels, making it more robust in challenging environments. Overall, the proposed hybrid robust beamformer stands out as a reliable solution, offering a balanced trade-off between null depth and sidelobe control, making it ideal for environments with multiple sources and interference, as highlighted by the performance analysis conducted with three signal sources and two interference sources. Therefore, the proposed beamformer can be implemented to the URA-based phased array scanning radar system to obtain the best possible outcome under noisier environment. The signal power comparison of sources and interferences among various beamformers (Conventional, MVDR, LCMV, Diagonal-loaded MVDR, Diagonal-loaded LCMV, and Proposed) with different number of array elements (10 × 10, 20 × 20, and 30 × 30 URAs) is depicted in Table 8. From Table 8, for the 10 × 10 and 20 × 20 URAs, a very small look direction errors are observed for the proposed beamformer. Yet for the 30 × 30 URA, no noticeable look direction error is seen. Apart from that, 30 × 30 URA is able to create deeper nulls to the interferences as compared with that of the 10 × 10 and 20 × 20 URAs. Here, the incident angles of Source-1, Source-2, and Source-3 are 0°, − 25°, and 45°, respectively. The incident angles of Interference-1 and Interference-2 are 60 and − 50, respectively. Table 8Signal power comparison of sources and interferences among various beamformers with different number of array elements. Significant values are in bold.ConventionalMVDRLCMVMVDR-DLLCMV-DLProposed10 × 10 URAPeak array gain (dB)202020202020Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−4.23−12.33−10.48−1.47−0.031 −0.001 Normalized source-2 signal power (dB)−2.10−8.04−16.94−1.49−0.046 −0.001 Normalized source-3 signal power (dB)−2.92−5.16−7.84−2.75−0.044 0 Normalized interference-1 signal power (dB)−14.66−9.31−23.19−0.007−65.71 −66.97 Normalized interference-2 signal power (dB)−17.64−5.26−19.98−0.603−65.16 −56.36 20 × 20 URAPeak array gain (dB)262626262626Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−12.48−5.89−6.82−1.78−0.002 −0.001 Normalized source-2 signal power (dB)−11.90−6.11−0.54−1.72−0.002 −0.001 Normalized source-3 signal power (dB)−10.61−12.08−2.80−1.690 0 Normalized interference-1 signal power (dB)−40.29−17.22−5.72−0.25−77.34 −65.33 Normalized interference-2 signal power (dB)−25.66−17.54−7.59−0.15−77.47 −50.93 30 × 30 URAPeak array gain (dB)29.529.529.529.529.529.5Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−15.76−23.02−28.74−1.680 0 Normalized source-2 signal power (dB)−21.80−16.49−15.76−1.69−0.001 0 Normalized source-3 signal power (dB)−31.35−7.67−7.27−1.690 0 Normalized interference-1 signal power (dB)−34.92−15.70−17.83−0.02−84.35 −69.45 Normalized interference-2 signal power (dB)−35.94−8.85−8.910−84.36 −74.25 Table 8Signal power comparison of sources and interferences among various beamformers with different number of array elements. Significant values are in bold.ConventionalMVDRLCMVMVDR-DLLCMV-DLProposed10 × 10 URAPeak array gain (dB)202020202020Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−4.23−12.33−10.48−1.47−0.031 −0.001 Normalized source-2 signal power (dB)−2.10−8.04−16.94−1.49−0.046 −0.001 Normalized source-3 signal power (dB)−2.92−5.16−7.84−2.75−0.044 0 Normalized interference-1 signal power (dB)−14.66−9.31−23.19−0.007−65.71 −66.97 Normalized interference-2 signal power (dB)−17.64−5.26−19.98−0.603−65.16 −56.36 20 × 20 URAPeak array gain (dB)262626262626Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−12.48−5.89−6.82−1.78−0.002 −0.001 Normalized source-2 signal power (dB)−11.90−6.11−0.54−1.72−0.002 −0.001 Normalized source-3 signal power (dB)−10.61−12.08−2.80−1.690 0 Normalized interference-1 signal power (dB)−40.29−17.22−5.72−0.25−77.34 −65.33 Normalized interference-2 signal power (dB)−25.66−17.54−7.59−0.15−77.47 −50.93 30 × 30 URAPeak array gain (dB)29.529.529.529.529.529.5Normalized peak array gain (dB)000000Normalized source-1 signal power (dB)−15.76−23.02−28.74−1.680 0 Normalized source-2 signal power (dB)−21.80−16.49−15.76−1.69−0.001 0 Normalized source-3 signal power (dB)−31.35−7.67−7.27−1.690 0 Normalized interference-1 signal power (dB)−34.92−15.70−17.83−0.02−84.35 −69.45 Normalized interference-2 signal power (dB)−35.94−8.85−8.910−84.36 −74.25 Signal power comparison of sources and interferences among various beamformers with different number of array elements. Significant values are in bold. Robustness margin analysis Fig. 25 presents the beam pattern comparison for a 30 × 30 URA under varying SNRs, showing the normalized power as a function of angle. The plot illustrates the beamforming performance across three signal sources and two interference sources at SNR levels ranging from − 10 dB to 10 dB. The beamformer performs well up to 0 dB, with the main beam focusing on the signal sources and creating nulls for interference. However, beyond − 1 dB, performance degrades, with increased sidelobes approaching 0 dB, which is unexpected and inappropriate for beamforming. Although extreme SNR values are presented, Fig. 25 depicts the parametric sweep with intermediate SNRs that offer greater clarity regarding the robustness margin, especially in the − 1 dB to 0 dB range, where the beamformer’s performance begins to deteriorate. Fig. 25Beam pattern comparison for 30 × 30 URA across different SNRs under three signal sources and two interference sources. Beam pattern comparison for 30 × 30 URA across different SNRs under three signal sources and two interference sources. Implementation of the proposed hybrid robust beamformer in the URA-based radar system From Fig. 26, one can observe the scan map of the URA-based scanning radar system with the scanning range from − 90° to + 90°. The position of the static interference is set at (2500 m, 100 m, 0 m). In Fig. 26a, one can observe that the conventional beamformer detects three predefined moving objects, but the positions of Object-2 and Object-3 are uncertain, which is visible from the scan map. Moreover, it cannot reject the interference that leads to false alarms. Fig. 26Scan map of the 30 × 30 URA-based phased array scanning radar system utilizing (a) Conventional beamformer, (b) Simple MVDR beamformer, (c) Simple LCMV beamformer, and (d) Proposed hybrid robust beamformer. Scan map of the 30 × 30 URA-based phased array scanning radar system utilizing (a) Conventional beamformer, (b) Simple MVDR beamformer, (c) Simple LCMV beamformer, and (d) Proposed hybrid robust beamformer. In Fig. 26b, one can observe that the simplified MVDR beamformer detects three objects, but their positions are totally uncertain, for which exact positions cannot be retrieved from the signal echoes. Again, it cannot reject the interference, which is strongly visible on the scan map. In Fig. 26c, one can observe that the LCMV beamformer detects only two objects and cancels out one object with the interference, which is not desired. Apart from that, the detected two objects’ positions are not certain so that the exact position cannot be retrieved from the echoes. In Fig. 26d, one can observe that the proposed hybrid robust beamformer detects all three objects with higher precision and rejects the interference properly. The object blips are so sharp, for which the exact position of the objects can be obtained from the scan map. After this comparison, it is undoubtedly stated that the proposed hybrid robust beamformer outperforms the other beamformers while implementing in the URA-based phased array scanning radar system. Multiple object detection & visualization Three objects are expected to be encountered and one interference to be rejected by the proposed system’s surveillance region. Table 9 displays the detection parameters acquired after rejecting one interference and identifying three objects at different positions in the MATLAB53 software environment. The fact that the fluctuation between true and measured values is negligible, as seen in Table 9, assures that the proposed scanning radar outperforms other systems in terms of detecting multiple objects. Table 9Detection parameters of three objects by simulation utilizing the proposed hybrid robust beamformer.Detection parametersAxisObject-1Object-2Object-3Object position (m)x160021003030y40−12001000z000Actual range (m)1600.502418.683190.76Estimated range (m)16002424.993200Actual angle (deg.)13018Estimated angle (deg.)03018Components of object velocity (m/s)x40−100−80y705050z000Actual velocity (m/s)80.62111.80−94.33Estimated velocity (m/s)80.51111.70−92.03 Table 9Detection parameters of three objects by simulation utilizing the proposed hybrid robust beamformer.Detection parametersAxisObject-1Object-2Object-3Object position (m)x160021003030y40−12001000z000Actual range (m)1600.502418.683190.76Estimated range (m)16002424.993200Actual angle (deg.)13018Estimated angle (deg.)03018Components of object velocity (m/s)x40−100−80y705050z000Actual velocity (m/s)80.62111.80−94.33Estimated velocity (m/s)80.51111.70−92.03 Detection parameters of three objects by simulation utilizing the proposed hybrid robust beamformer. In Fig. 27, one can observe each object’s 3D coordinates and their visual representation. The large peaks over the detection threshold are object echoes, and the rest of the peaks under the threshold are noise echoes, which are not detected by the proposed radar system to maintain its performance. Again, from Fig. 27, it is noted that the Object-1, Object-2, and Object-3 peak height coordinates in the X, Y, and Z axes are (1600 m, 0 m, − 112.54 dB), (2100.11 m, − 1212.5 m, − 116.46 dB), and (3043.38 m, 988.85 m, − 115.23 dB), respectively. Fig. 27Visualization of the identified three objects in 3D coordinate system of the proposed 30 × 30 URA-based phased array scanning radar system with the hybrid robust beamformer using MATLAB R2021a (Version: 9.10.0.1602886, URL: https://www.mathworks.com) software. Visualization of the identified three objects in 3D coordinate system of the proposed 30 × 30 URA-based phased array scanning radar system with the hybrid robust beamformer using MATLAB R2021a (Version: 9.10.0.1602886, URL: https://www.mathworks.com) software. Range and angle estimation The three objects that are simulated by the pulse all have an elevation of zero. A variation in the object positions results in distinct changes to both the actual and estimated ranges of the objects illustrated in Fig. 28. The actual and estimated values of the range are documented in Table 10. Similarly, by changing the positions of objects, variations are observed in the actual angle and angle estimated of the objects visualized in Fig. 29. The data on the actual and estimated angles of the objects are recorded in Table 10. Fig. 28Changes in range by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Changes in range by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Table 10Tracking parameters (range and angle) of three objects by simulation for the proposed radar system.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Object-111430143160.9955.611431.31432.082221480148266.6578.171481.51484.063331525152781.9479.731527.21529.083341575157781.3680.831577.11579.07335160016004001600.516001061621162054.0259.161621.91621.082271653165557.5159.816541656.082281689169146.1460.171689.631692.072291706170458.4260.417071705.0722101720172164.2661.261721.21722.0922Object-2120002001.11−1022.72−1024.702246.322248.212727220322035.21−1082.64−1081.872302.422304.892828320802078.36−1107.11−1114.072356.292358.122828420962096.50−1171.06−1166.372400.962399.112929521002100.11−1200.01−1212.482418.682424.993030621502148.54−1282.70−1294.472503.562508.363131722602256.25−1138.30−1153.732530.482534.122727823402342.24−1114.00−1117.602591.642595.212526923602358.11−1121.01−1136.842612.712617.8425261024902491.12−909.14−924.972650.782657.302020Object-3128502890.20901.96785.82989.322995.121815229002910.30928.85912.773045.123050.081817329402960.20988.12972.143101.613115.741918430003008.21930.16955.533140.893156.321718530303043.381000.02988.863190.763200.001818631003108.89909.91947.783230.783250.151617731403156.21966.6930.33285.413290.461716832003215.32870.12904.613316.193340.151516932463261.45869.04858.123360.323372.4515151033103321.46738.81722.963391.453399.231312 Table 10Tracking parameters (range and angle) of three objects by simulation for the proposed radar system.Obs. no.Act. X-pos. (m)Est. X-pos. (m)Act. Y-pos. (m)Est. Y-pos. (m)Act. range (m)Est. range (m)Act. angle (Deg.)Est. angle (Deg.)Object-111430143160.9955.611431.31432.082221480148266.6578.171481.51484.063331525152781.9479.731527.21529.083341575157781.3680.831577.11579.07335160016004001600.516001061621162054.0259.161621.91621.082271653165557.5159.816541656.082281689169146.1460.171689.631692.072291706170458.4260.417071705.0722101720172164.2661.261721.21722.0922Object-2120002001.11−1022.72−1024.702246.322248.212727220322035.21−1082.64−1081.872302.422304.892828320802078.36−1107.11−1114.072356.292358.122828420962096.50−1171.06−1166.372400.962399.112929521002100.11−1200.01−1212.482418.682424.993030621502148.54−1282.70−1294.472503.562508.363131722602256.25−1138.30−1153.732530.482534.122727823402342.24−1114.00−1117.602591.642595.212526923602358.11−1121.01−1136.842612.712617.8425261024902491.12−909.14−924.972650.782657.302020Object-3128502890.20901.96785.82989.322995.121815229002910.30928.85912.773045.123050.081817329402960.20988.12972.143101.613115.741918430003008.21930.16955.533140.893156.321718530303043.381000.02988.863190.763200.001818631003108.89909.91947.783230.783250.151617731403156.21966.6930.33285.413290.461716832003215.32870.12904.613316.193340.151516932463261.45869.04858.123360.323372.4515151033103321.46738.81722.963391.453399.231312 Tracking parameters (range and angle) of three objects by simulation for the proposed radar system. Fig. 29Changes in angle by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Changes in angle by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Velocity estimation The estimated and actual velocities of the objects depicted in Fig. 30 change substantially when the target positions are varied. From the data on velocities documented in Table 11, one can observe that the velocities of Object-1 and Object-2 are positive, whereas Object-3 shows negative velocities. The negative velocity indicates that the moving object is coming towards the radar. Therefore, the first and second objects are moving away from the radar, while the third object is coming near it, which is accurately decided by the proposed radar system. Fig. 30Changes in velocity by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Changes in velocity by varying positions of Object-1, Object-2, and Object-3 at 0° of elevation for the proposed system. Table 11Tracking parameters (velocity) of three objects by simulation for the proposed radar system.Obs. no.X-component of actual velocity (m/s)Y-component of actual velocity (m/s)Net actual velocity (m/s)Estimated velocity (m/s)Object-11407080.6278.052407080.6278.663407080.6278.944407080.6279.065407080.6279.646407080.6279.877407080.6279.958407080.6280.129407080.6280.3410407080.6280.51Object-21−10050111.80109.062−10050111.80109.453−10050111.80110.124−10050111.80110.465−10050111.80110.896−10050111.80111.017−10050111.80111.238−10050111.80111.459−10050111.80111.5010−10050111.80111.70Object-31−8050−94.33−94.252−8050−94.33−94.143−8050−94.33−93.764−8050−94.33−93.045−8050−94.33−92.466−8050−94.33−92.377−8050−94.33−92.318−8050−94.33−92.239−8050−94.33−92.1010−8050−94.33−92.03 Table 11Tracking parameters (velocity) of three objects by simulation for the proposed radar system.Obs. no.X-component of actual velocity (m/s)Y-component of actual velocity (m/s)Net actual velocity (m/s)Estimated velocity (m/s)Object-11407080.6278.052407080.6278.663407080.6278.944407080.6279.065407080.6279.646407080.6279.877407080.6279.958407080.6280.129407080.6280.3410407080.6280.51Object-21−10050111.80109.062−10050111.80109.453−10050111.80110.124−10050111.80110.465−10050111.80110.896−10050111.80111.017−10050111.80111.238−10050111.80111.459−10050111.80111.5010−10050111.80111.70Object-31−8050−94.33−94.252−8050−94.33−94.143−8050−94.33−93.764−8050−94.33−93.045−8050−94.33−92.466−8050−94.33−92.377−8050−94.33−92.318−8050−94.33−92.239−8050−94.33−92.1010−8050−94.33−92.03 Tracking parameters (velocity) of three objects by simulation for the proposed radar system. Performance analysis of the proposed scanning radar system From Table 12, it is depicted that the average accuracy for range estimation of the proposed system is 99.80%, the average accuracy for angle estimation is 99.94%, and the average accuracy for velocity estimation is 98.67%. The overall accuracy is obtained by Eq. (94) and documented in Table 13. Table 12Average accuracy of tracking parameters (range, angle and velocity) of three objects by simulation for the proposed radar system.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (m/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (m/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-11431.31432.080.999522180.6278.050.96711481.51484.060.998333180.6278.660.97511527.21529.080.998833180.6278.940.97871577.11579.070.998833180.6279.060.98031600.516000.999710080.6279.640.98771621.91621.080.999522180.6279.870.990616541656.080.998722180.6279.950.99161689.631692.070.998622180.6280.120.993817071705.070.998922180.6280.340.99651721.21722.090.999522180.6280.510.9986Object-22246.322248.210.999227271111.80109.060.97492302.422304.890.998928281111.80109.450.97852356.292358.120.999228281111.80110.120.98472400.962399.110.999229291111.80110.460.98792418.682424.990.997430301111.80110.890.99182503.562508.360.998131311111.80111.010.99292530.482534.120.998627271111.80111.230.99492591.642595.210.998625260.96111.80111.450.99692612.712617.840.99825260.96111.80111.500.99732650.782657.300.997520201111.80111.700.9991Object-32989.322995.120.998118150.8333−94.33−94.250.99923045.123050.080.998418170.9444−94.33−94.140.9983101.613115.740.995419180.9474−94.33−93.760.99393140.893156.320.995117180.9412−94.33−93.040.98613190.763200.000.997118181−94.33−92.460.97983230.783250.150.99416170.9375−94.33−92.370.97883285.413290.460.998517160.9412−94.33−92.310.97813316.193340.150.992815160.9333−94.33−92.230.97723360.323372.450.996415151−94.33−92.100.97583391.453399.230.997713120.9231−94.33−92.030.975Average accuracy for range estimation0.9980Average accuracy for angle estimation0.9440Average accuracy for velocity estimation0.9867 Table 12Average accuracy of tracking parameters (range, angle and velocity) of three objects by simulation for the proposed radar system.Act. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{t}\:$$\end{document}(m)Est. range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{R}_{m}\:$$\end{document}(m)Accuracy for range, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{R}$$\end{document}Act. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{t}\:$$\end{document}(Deg.)Est. angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\theta\:}_{m}\:$$\end{document}(Deg.)Accuracy for angle, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{\theta\:}$$\end{document}Act. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{t}$$\end{document} (m/s)Est. velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{{v}_{D}}_{m}$$\end{document} (m/s)Accuracy for velocity, \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$\:{\mu\:}_{{v}_{D}}$$\end{document}Object-11431.31432.080.999522180.6278.050.96711481.51484.060.998333180.6278.660.97511527.21529.080.998833180.6278.940.97871577.11579.070.998833180.6279.060.98031600.516000.999710080.6279.640.98771621.91621.080.999522180.6279.870.990616541656.080.998722180.6279.950.99161689.631692.070.998622180.6280.120.993817071705.070.998922180.6280.340.99651721.21722.090.999522180.6280.510.9986Object-22246.322248.210.999227271111.80109.060.97492302.422304.890.998928281111.80109.450.97852356.292358.120.999228281111.80110.120.98472400.962399.110.999229291111.80110.460.98792418.682424.990.997430301111.80110.890.99182503.562508.360.998131311111.80111.010.99292530.482534.120.998627271111.80111.230.99492591.642595.210.998625260.96111.80111.450.99692612.712617.840.99825260.96111.80111.500.99732650.782657.300.997520201111.80111.700.9991Object-32989.322995.120.998118150.8333−94.33−94.250.99923045.123050.080.998418170.9444−94.33−94.140.9983101.613115.740.995419180.9474−94.33−93.760.99393140.893156.320.995117180.9412−94.33−93.040.98613190.763200.000.997118181−94.33−92.460.97983230.783250.150.99416170.9375−94.33−92.370.97883285.413290.460.998517160.9412−94.33−92.310.97813316.193340.150.992815160.9333−94.33−92.230.97723360.323372.450.996415151−94.33−92.100.97583391.453399.230.997713120.9231−94.33−92.030.975Average accuracy for range estimation0.9980Average accuracy for angle estimation0.9440Average accuracy for velocity estimation0.9867 Average accuracy of tracking parameters (range, angle and velocity) of three objects by simulation for the proposed radar system. Table 13Overall accuracy of the proposed radar system for object detection utilizing the hybrid robust beamformer.Average accuracy for range estimationAverage accuracy for angle estimationAverage accuracy for velocity estimationOverall accuracy0.99800.94400.98670.9929 or 99.29% Table 13Overall accuracy of the proposed radar system for object detection utilizing the hybrid robust beamformer.Average accuracy for range estimationAverage accuracy for angle estimationAverage accuracy for velocity estimationOverall accuracy0.99800.94400.98670.9929 or 99.29% Overall accuracy of the proposed radar system for object detection utilizing the hybrid robust beamformer. If we look at the overall accuracy of the ULA-based phased array radar system utilizing conventional beamformer for both simulation and experimentation process depicted in Table 7, it is less than that of the proposed URA-based phased array radar system utilizing hybrid robust beamformer. The proposed radar system with robust algorithm improves the average angle estimation accuracy that leads to higher overall accuracy of the system. Apart from that, the average range estimation accuracy is not degraded though the maximum unambiguous range is considerably increased. Therefore, the proposed 30 × 30 URA-based scanning radar system shows outstanding performance regarding all the defined extreme scenarios compared to the 1 × 16 ULA-based radar prototype. Comparative study Tables 14 and 15 provide a brief overview of the various techniques used in radar systems. Table 14 clearly shows that the radar system that is being proposed can identify several moving objects at the same time and locate them wherever in the area being monitored. Other methods are not able to achieve this at the same time. Table 14Qualitative comparative study.Radar systemsObject range detectableObject angle detectableObject velocity detectableNumber of objectsRef.8,9NoNoNoMultiple stationaryRef.10,11NoYes (Same range azimuth)NoMultiple stationaryRef.12–14YesYesNoMultiple movingRef.15YesNoYesMultiple movingRef.16YesNoNoMultiple (Discrimination is not specified)Ref.17YesYes (Poor azimuth resolution)NoMultiple movingRef.21YesYes (Poor azimuth resolution)NoMultiple movingRef.19YesYes (Poor azimuth resolution)YesMultiple movingRef.23YesYes (Same range azimuth)YesMultiple moving (At the same distance)Ref.3,7Yes (Poor range resolution)YesYesMultiple movingProposedYesYesYesMultiple moving Table 14Qualitative comparative study.Radar systemsObject range detectableObject angle detectableObject velocity detectableNumber of objectsRef.8,9NoNoNoMultiple stationaryRef.10,11NoYes (Same range azimuth)NoMultiple stationaryRef.12–14YesYesNoMultiple movingRef.15YesNoYesMultiple movingRef.16YesNoNoMultiple (Discrimination is not specified)Ref.17YesYes (Poor azimuth resolution)NoMultiple movingRef.21YesYes (Poor azimuth resolution)NoMultiple movingRef.19YesYes (Poor azimuth resolution)YesMultiple movingRef.23YesYes (Same range azimuth)YesMultiple moving (At the same distance)Ref.3,7Yes (Poor range resolution)YesYesMultiple movingProposedYesYesYesMultiple moving Qualitative comparative study. From Table 15, one can observe the quantitative comparative study among the existing and proposed radar systems. The average accuracy of the models in2,3,6,7, and the proposed model is very similar. However, their systems were tested in ideal conditions and did not account for any interference. In this context, the proposed system is able to enhance the detection capability even under challenging scenarios. The proposed system shows better average angle accuracy as compared with3,6 due to the robust algorithm. In6,7, the average velocity accuracy was not mentioned, and the proposed method outperforms2,3 in this context. The overall accuracy of the proposed scanning radar system is larger than that of all the mentioned systems. Table 15Quantitative comparative study.Ref.7Ref.3Ref.6Ref.2ProposedRadar typeScanningScanningScanningScanningScanningRadar antenna type1 × 16 ULA1 × 16 ULA40 × 40 URA35 × 35 URA30 × 30 URAMaximum range3.6 m3.6 m5000 m5000 m5000 mScanning range−45° to + 45°−34° to + 34°−120° to + 120°−90° to + 90°−90° to + 90°No. of objects for detection32343InterferenceNot consideredNot consideredNot consideredNot considered Considered Beamformer typeConventionalConventionalConventionalConventional Hybrid robust ScenarioStableStableStableStable Challenging Analysis processSimulationExperimentalSimulationSimulationSimulationAverage range estimation accuracy99.62%98.66%99.93%99.92%99.80%Average angle estimation accuracy95.03%92.13%93.16%95.22%94.40%Overall accuracyNot mentioned96.47%Not mentioned97.72% 99.29% Table 15Quantitative comparative study.Ref.7Ref.3Ref.6Ref.2ProposedRadar typeScanningScanningScanningScanningScanningRadar antenna type1 × 16 ULA1 × 16 ULA40 × 40 URA35 × 35 URA30 × 30 URAMaximum range3.6 m3.6 m5000 m5000 m5000 mScanning range−45° to + 45°−34° to + 34°−120° to + 120°−90° to + 90°−90° to + 90°No. of objects for detection32343InterferenceNot consideredNot consideredNot consideredNot considered Considered Beamformer typeConventionalConventionalConventionalConventional Hybrid robust ScenarioStableStableStableStable Challenging Analysis processSimulationExperimentalSimulationSimulationSimulationAverage range estimation accuracy99.62%98.66%99.93%99.92%99.80%Average angle estimation accuracy95.03%92.13%93.16%95.22%94.40%Overall accuracyNot mentioned96.47%Not mentioned97.72% 99.29% Quantitative comparative study. Computational complexity The condition number-based method is used in this study to obtain the optimal weights for the proposed beamformer. For the grid search method, a grid spacing of 0.5λ is used to ensure adequate resolution for adaptive computations. On the other hand, a maximum iteration limit of 100 is set for the stopping criteria. Besides, a threshold for the change in the objective function (optimal diagonal loading factor) is set. The grid search process is stopped when the difference between successive iterations falls below 10−6. From Fig. 31, one can observe the convergence comparison between Newton-Raphson and grid search (conventional) method. The condition number difference is not converged to zero by the grid search method even after about 100 iterations whereas it is converged to zero by the Newton-Raphson numerical method after only 10 iterations. For this reason, Newton-Raphson numerical method is applied to reduce the computational complexity regarding the optimal weight calculation of the proposed beamformer. Fig. 31Convergence comparison graph between Newton-Raphson and Grid search method for obtaining the optimal weights for the proposed beamformer. Convergence comparison graph between Newton-Raphson and Grid search method for obtaining the optimal weights for the proposed beamformer. Table 16 presents a performance comparison between two computational methods, namely the Grid Search and Newton-Raphson, based on their Floating-Point Operations (FLOP) count and runtime. The hardware specifications are provided, detailing the processor model (Intel® Core™ i3-3220 from the 3rd Generation), core type (Dual Core), clock speed (3.30 GHz), RAM (12 GB), HDD (512 GB), and SSD (128 GB). The methods are evaluated in the MATLAB53 software, with their corresponding FLOP counts and runtime for different iterations. From Table 16, it is observed that the grid search method requires approximately 218 billion FLOPs for 100 iterations, resulting in a runtime of 107.379 s. On the other hand, the Newton-Raphson method needs about 21.87 billion FLOPs for 10 iterations, with a runtime of 34.8696 s, confirming a fast convergence compared with the grid search method. Table 16Flop count and runtime analysis of different computational methods.Hardware specificationSoftwareMethodsFLOP countRuntime (s)Processor: Intel® Core™ i3-3220 (3rd Generation)Core Type: DualClock Speed: 3.30 GHzRAM: 12 GBHDD: 512 GBSSD: 128 GBMATLAB R2021a (Version: 9.10.0.1602886)Grid search~ 218 Billion (for 100 iterations)107.379Newton-Raphson~ 21.87 Billion (for 10 iterations)34.8696 Table 16Flop count and runtime analysis of different computational methods.Hardware specificationSoftwareMethodsFLOP countRuntime (s)Processor: Intel® Core™ i3-3220 (3rd Generation)Core Type: DualClock Speed: 3.30 GHzRAM: 12 GBHDD: 512 GBSSD: 128 GBMATLAB R2021a (Version: 9.10.0.1602886)Grid search~ 218 Billion (for 100 iterations)107.379Newton-Raphson~ 21.87 Billion (for 10 iterations)34.8696 Flop count and runtime analysis of different computational methods. Processor: Intel® Core™ i3-3220 (3rd Generation) Core Type: Dual Clock Speed: 3.30 GHz RAM: 12 GB HDD: 512 GB SSD: 128 GB Hardware limitations Processor: The Intel® Core™ i3-3220, being a 3rd generation processor, has a relatively lower clock speed and fewer cores compared to more modern processors. This results in slower computational speeds for both methods, especially when handling large FLOP counts or multiple iterations. RAM: The 12 GB RAM is adequate for moderate tasks, but might be limiting for large-scale computations or more complex simulations. Storage: The HDD of 512 GB and SSD of 128 GB are somewhat outdated for computational-heavy tasks. The slower speed of the HDD could create bottlenecks during disk access, especially if the methods are storing or retrieving large datasets during processing. The smaller SSD capacity may also limit the system’s ability to handle large datasets efficiently. Core Count: With only 2 cores, the processor’s ability to parallelize computations is limited. Methods like Grid Search, which involve multiple iterations and evaluations, may significantly benefit from a more multi-core processor for faster processing. Conclusion In this study, an analysis of the phased array scanning radar system’s performance utilizing the hybrid robust beamformer has been presented relative to that of previous radars. In addition, comprehensive qualitative and quantitative comparative studies have been conducted to examine the capabilities of both existing systems and the proposed model for identifying numerous moving objects using the given requirements. First, a simulation work along with experimental validation of a 1 × 16 ULA-based radar system with a conventional beamformer has been performed under a well-conditioned scenario, and it shows an overall accuracy of 92.05% and 89.80% in terms of the simulation and experimentation process, respectively, for object detection. However, the maximum unambiguous range for the experimental prototype has been very low in terms of real-life application. Thus, when the range of objects has increased, the detection accuracy of the prototype has declined. So, the higher accuracy can be retained in that case if the maximum detectable range and scanning area of the PAA are increased by adding more antenna elements to the array. Moreover, the process of interference rejection cannot be performed with this experimental setup and so the overall accuracy is considerably affected with noisier environment. To enhance the radar performance in challenging situations, the hybrid robust beamforming model on a 30 × 30 URA-based phased array scanning radar system has been employed. In the hybrid robust model, the MVDR and LCMV beamformer with diagonal loading have been combined with a hybrid weighting factor. The condition number-based method has been used to obtain the optimal weights for the proposed beamformer and Newton-Raphson numerical method has been used to reduce computational complexity. Comparing the proposed radar system to conventional radar systems, the proposed phased array radar shows an outstanding overall accuracy of 99.29% for multiple object detection and interference rejection, according to the simulation results. The scanning criteria of the proposed radar is two-dimensional, which can be extended to the three-dimensional scanning by employing additional calibration for elevation channels and more complex steering vector modeling. The performance of the proposed system can be investigated by introducing the non-planar arrays such as conformal or spherical arrays. Moreover, all the targets in this study are non-fluctuating. A future advancement can be done by including Swerling I–IV and micro-Doppler signatures for more realistic diversity. Supplementary Information Below is the link to the electronic supplementary material. Supplementary Material 1 Supplementary Material 1