Beamfocusing and Power Allocation for AN-Based PLS in Multiuser XL-MIMO with Multiple Eavesdroppers

Xiangjun Ma, Ali Arshad Nasir, Daniel Benevides da Costa · 2025

Read Beamfocusing and Power Allocation for AN-Based PLS in Multiuser XL-MIMO with Multiple Eavesdroppers by Xiangjun Ma, Ali Arshad Nasir, Daniel Benevides da Costa (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-14100.txt.

Introduction

Xiangjun Ma, Ali Arshad Nasir, and Daniel Benevides da Costa study a 2025 problem in wireless security for very large antenna arrays. In near-field links, a sender can focus beams tightly and add shaped noise that harms eavesdroppers but not true users. The aim is to lift the worst secrecy rate across many users.

The team splits the task between beam shaping and power sharing, solved by two search methods in turn. Tests show the noise does help even with many listeners, and even when their channel facts are unknown. The joint design beat past methods that skip the noise or use plain beam rules.

Catalog record

CREATORXiangjun Ma, Ali Arshad Nasir, Daniel Benevides da Costa
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE28 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-14100

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Beamfocusing and Power Allocation for AN-Based PLS in Multiuser XL-MIMO with Multiple Eavesdroppers Xiangjun Ma, Ali Arshad Nasir, Daniel Benevides da Costa Abstract This paper investigates the downlink (DL) physical layer security (PLS) in a near-field (NF) extra-large multiple-input multiple-output MIMO (XL-MIMO) system. To enhance the secrecy rate (SR), null-space artificial noise (AN) is transmitted alongside the confidential message, ensuring orthogonality with legitimate user equipment (LUE) channels. The objective is to maximize the minimum SR by optimizing the NF beamfocusing matrix and power allocation between the signal and AN, considering various channel state information (CSI) conditions and transmit power constraints. The proposed approach uses successive convex approximation (SCA) for beamfocusing optimization and golden section search (GSS) for power allocation. The following open questions are addressed: (i) Can AN transmission further enhance SR for multiple LUEs in the presence of multiple eavesdropping user equipment (EUEs)? (ii) Can null-space AN transmission achieve attractive SR performance even without CSI availability for EUEs? Both questions are affirmatively answered and explored in detail, with an algorithm presented for joint beamfocusing design and AN-aided power allocation. The proposed method outperforms state-of-the-art approaches that either omit AN transmission or rely on maximal-ratio transmission (MRT) for beamfocusing. I Introduction As a promising technique in the upcoming sixth-generation (6G) networks, extremely large-scale multiple-input multiple-output (XL-MIMO) systems are equipped with a large number of antennas, resulting in an oversized antenna aperture. This significantly improves spectral efficiency (SE) compared to conventional MIMO systems [ 1 ] . On the other hand, the electromagnetic (EM) radiation from the transmitter to the receiver can be divided into far-field (FF) and near-field (NF) regions, differentiated by the Rayleigh distance. Since the Rayleigh distance is positively correlated with the antenna aperture, NF communications become dominant in XL-MIMO systems. This necessitates the use of spherical-wave model to accurately characterize the channel. Due to the broadcast nature of the wireless propagation channel, transmitted signals from base station (BS) to legitimate UEs (LUEs) are vulnerable to wiretapping by malicious eavesdropping UEs (EUEs). For FF communication, the beamforming design based on the planar-wave model often leads to significant leakage of confidential messages, as EUEs are typically located in the same direction as LUEs [ 2 ] . In contrast, the spherical-wave model accounts for both the angle and distance information of the receiver, allowing precoding to focus on a specific region, known as beamfocusing. This approach enables positive secrecy rate (SR) even when EUEs are positioned in the same direction. Several groundbreaking studies have recently emerged on physical layer security (PLS) for NF communications. For the single-UE scenario, the authors in [ 3 ] investigated secrecy performance of XL-MIMO, where simulation results demonstrated that the spherical-wave model can achieve higher SR. An NF secure transmission framework using hybrid beamforming was proposed in [ 2 ] , where a two-stage algorithm was developed to maximize SR. The authors in [ 4 ] investigated the SR maximization problem in NF wideband communications, where a novel analog beamfocusing scheme was introduced to address the interplay between near-field propagation and wideband beamsplit. For the multi-UE scenario, [ 5 ] investigated the sum SR in XL-MIMO and proposed a novel leakage subspace precoding strategy, demonstrating superior performance compared to conventional methods. In [ 6 ] , a hybrid beamfocusing scheme was introduced to maximize the minimum SR in NF communications, ensuring a more uniform SR distribution among all LUEs. Since the passive EUE is undetectable, its channel state information CSI is hard to estimate. A practical approach to enhance secrecy is injecting artificial noise (AN). In [ 7 ] , the PLS of a reconfigurable intelligent surface (RIS)-aided MIMO system was investigated, where the beamforming matrix and AN covariance matrix were jointly optimized to maximize the sum-rate while limiting confidential message leakage to EUEs. The work in [ 8 ] examined the secrecy performance of a RIS-aided cell-free network, formulating a max-min security energy efficiency optimization problem to jointly design active beamforming and AN matrices at BSs and passive beamforming matrices at RISs. However, both studies [ 7 ] and [ 8 ] focused on conventional FF communication. A novel AN-aided beamfocusing scheme was developed in [ 9 ] for a NF communication system. However, the study did not address the challenging max-min SR optimization in the presence of multiple LUE-EUE case. Additionally, the adopted maximal-ratio transmission (MRT) beamfocusing usually does not yield optimal SR and the assumption of perfect CSI of EUEs is not practical as well. Building on the preceding discussion, this paper investigates secure downlink (DL) transmission in an NF XL-MIMO system, where both LUEs and EUEs are aligned in the same direction relative to the BS. To enhance the SR, null-space AN is transmitted alongside the signal, ensuring orthogonality to LUE channels. The objective is to maximize the minimum SR by optimizing the NF beamfocusing matrix and the power allocation between the signal and AN, considering various CSI conditions and subject to the BS’s power budget. The proposed approach leverages successive convex approximation (SCA) for beamfocusing optimization and golden section search (GSS) for power allocation. We have demonstrated the superiority of the proposed algorithm over existing state-of-the-art approaches that either omit AN transmission or rely on MRT for beamfocusing. Our simulation results show that the proposed null-space AN transmission, under no CSI availability of EUEs, can achieve a performance close to the upper benchmark, which assumes CSI availability of EUEs. Notations : The superscripts T and H denote the and Hermitian transpose, respectively. \|\bullet\| and \|\bullet\|_{\text{F}} denote the 2-norm and Frobenius norm, respectively. \mathbf{I}_{N} is the N\times{N} identity matrix. The notation [\mathbf{a}]_{n} refers to the n -th element of vector \mathbf{a} . The indicator function \mathds{1}_{\mathcal{P}}(x) is defined as \mathds{1}_{\mathcal{P}}(x)=1 if x\in\mathcal{P} and \mathds{1}_{\mathcal{P}}(x)=0 if x\notin\mathcal{P} . Finally, [x]^{+}=\max\{0,x\} . II System Model Fig. 1: DL secrecy transmission of XL-MIMO. We consider a DL secrecy transmission scenario in an XL-MIMO system, as illustrated in Fig. 1 . In this setup, a BS equipped with N_{\mathrm{b}} antennas communicates with K single-antenna LUEs, while E single-antenna EUEs attempt to intercept the confidential messages intended for all K LUEs. We define the LUE and EUE sets as \mathcal{K}=\{1,\cdots,K\} and \mathcal{E}=\{1,\cdots,E\} , respectively, and further define \mathcal{U}=\mathcal{K}\cup\mathcal{E} . The total number of BS antennas is given by N_{\mathrm{b}}=N_{\mathrm{b}}^{x}N_{\mathrm{b}}^{z} , where N_{\mathrm{b}}^{x} and N_{\mathrm{b}}^{z} represent the number of elements in the uniform planar array (UPA) along the x -axis and z -axis, respectively. Moreover, we assume that the number of UPA elements in each direction is odd, i.e., N_{\mathrm{b}}^{\nu}=2\bar{N}_{\mathrm{b}}^{\nu}+1 , where \bar{N}_{\mathrm{b}}^{\nu} is a non-negative integer and \nu\in\{x,z\} . II-A Channel Model The inter-antenna distance is set to \lambda/2 , where \lambda is the wavelength. The channel from the BS to UE- u ( u\in\mathcal{U} ) is denoted as \mathbf{h}_{u\mathrm{b}}^{\mathrm{H}}\in\mathbb{C}^{1\times{N}_{\mathrm{b}}} , and can be modeled as: \mathbf{h}_{u\mathrm{b}}^{\mathrm{H}}=\beta_{u\mathrm{b}}^{\frac{1}{2}}e^{-\mathrm{j}\frac{2\pi}{\lambda}d_{u\mathrm{b}}}[\mathbf{a}_{\mathrm{b}}(\theta_{u\mathrm{b}},\phi_{u\mathrm{b}},d_{u\mathrm{b}})]^{\mathrm{T}}+\sum_{l=1}^{L}(\mathbf{h}_{u\mathrm{b}}^{(l)})^{\mathrm{H}}. (1) where L is the number of non-line-of-sight (NLoS) paths. Additionally, \beta_{u\mathrm{b}} , d_{u\mathrm{b}} , \theta_{u\mathrm{b}} , and \phi_{u\mathrm{b}} represent the LSFC, distance, azimuth angle, and elevation angle of the line-of-sight (LoS) path from the BS to UE- u , respectively. Furthermore, \mathbf{a}_{\mathrm{b}}(\theta_{u\mathrm{b}},\phi_{u\mathrm{b}},d_{u\mathrm{b}})\in\mathbb{C}^{N_{\mathrm{b}}} denotes the transmitting array response of the BS, and (\mathbf{h}_{u\mathrm{b}}^{(l)})^{\mathrm{H}} represents the l -th NLoS path, as discussed in [ 10 ] . Define \displaystyle{f}^{x}(\theta,\phi,d,\bar{n})=\exp\bigg(-\mathrm{j}\frac{2\pi}{\lambda}\bigg(\bar{n}\frac{\lambda}{2}\cos(\theta)\sin(\phi)

Read the complete text