The Multidimensional Quadratic Phase Fourier Transform: Theoretical Analysis and Applications

Sarga Varghese, Gita Rani Mahato, Manab Kundu · 2025

Read The Multidimensional Quadratic Phase Fourier Transform: Theoretical Analysis and Applications by Sarga Varghese, Gita Rani Mahato, Manab Kundu (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-2505-02526.txt.

Introduction

A transform that twists the classic Fourier lens with a quadratic phase can see signals the standard lens blurs. Sarga Varghese, Gita Rani Mahato, and Manab Kundu extend that lens to many dimensions in this 2025 paper. Their transform generalizes several famous cousins at once, including fractional and linear canonical forms. Core properties are proven, including energy conservation and a way back from transformed data.

New convolution and correlation operations for several variables are built inside the framework. A Boas-type theorem ties the transform to a classical inequality style. Two applications close the paper: designing multiplicative filters and solving integral equations with the new convolution. Multi-dimensional signals, from images to fields, are the natural audience.

Catalog record

CREATORSarga Varghese, Gita Rani Mahato, Manab Kundu
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE93 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2505-02526

Browse the full library · About the Tower of Babel

Full text

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

The Multidimensional Quadratic Phase Fourier Transform: Theoretical Analysis and Applications Sarga Varghese, Gita Rani Mahato, Manab Kundu Abstract . The quadratic phase Fourier transform (QPFT) is a generalization of several well known integral transforms, including the linear canonical transform (LCT), fractional Fourier transform (FrFT), and Fourier transform (FT). This paper introduces the multidimensional QPFT and investigates its theoretical properties, including Parseval’s identity and inversion theorems. Generalized convolutions and correlation for multiple variables, extending the conventional convolution for single-variable functions, is proposed within the QPFT setting. Additionally, Boas type theorem for the multidimensional QPFT is established. As applications, multiplicative filter design and the solution of integral equations using the proposed convolution operation are explored. 1. Introduction Integral transforms [ 1 , 2 ] have been used for nearly two centuries to solve various problems in applied mathematics, mathematical physics, and engineering science. By converting functions from one domain to another, integral transforms simplify complex problems, enabling the analysis and solution of differential equations, and the study of signal properties. The Fourier transform (FT), one of the widely used integral transform, is a significant tool used in signal processing and harmonic analysis [ 3 ] . Despite its long history, the field continues to evolve, with new transforms like the fractional Fourier transform(FrFT), linear canonical transform (LCT), and wavelet transform emerging in the past three decades. In 1980, Victor Namias [ 4 ] introduced the fractional Fourier transform (FrFT), which he used to solve problems in quantum mechanics. By the 1990s, it gained attention for its applications in signal processing and optics. The FrFT generalizes the FT by introducing a fractional order parameter, allowing the transform to rotate signals in the time-frequency plane [ 7 , 5 , 6 , 8 ] . The FrFT was introduced to address the limitations of the FT in analyzing signals with time-varying frequency content. It is a special case of the LCT, a broader class of integral transforms that extends FrFT by incorporating four transformation parameters. The LCT, in turn, allows for more flexible time-frequency analysis [ 9 , 10 , 11 , 12 ] . Further extending these concepts, the quadratic phase Fourier transform (QPFT) includes quadratic phase terms to model complex signal behaviors, making it useful in applications such as radar, sonar, and optical systems. Also, it is used in linear time-invariant (LTI) system, signal analysis, solving integral equations, etc. QPFT was introduced by Castro et al. [ 13 , 14 ] motivated by the work of Saitoh [ 15 ] , who used the reproducing kernel theory of QPFT for solving heat equation. Later, Prasad et al. [ 16 , 17 , 18 , 19 ] discussed the QPFT framework by developing the wavelet transform, and studied its underlying features. QPFT can be used to identify single and multi-component linear frequency modulated signals by introducing a modified ambiguity function and the Wigner distribution associated with QPFT [ 21 ] . For more recent works in QPFT theory, one can refer to [ 20 , 22 , 23 ] . In this paper, we are mainly focusing on (a) new convolution and correlation for LTI system and solving integral equations, (b) Boas theorem for analyzing the compactly supported signals, for the multidimensional cases. The theoretical foundation of linear time-invariant (LTI) systems heavily relies on convolution and correlation processes. These methods work very well in a variety of fields, including pattern recognition, optics, and signal processing. Any continuous LTI structure’s efficiency can be calculated by integrating the impulse response of the system with the incoming signals [ 24 ] . In recent times, Wei et al. and Li et al. have made remarkable contributions in the field of convolution and correlation. They proposed new structures for several integral transforms and investigated their sampling theorems for band-limited signals [ 25 , 26 , 27 , 28 , 29 , 30 , 31 , 32 ] . Boas’ theorem [ 33 ] characterizes high-frequency signals by relating the vanishing of signal frequency content to a specific integral transform. This result is crucial for understanding high-frequency signals in Fourier analysis. It states that the signal frequency content f vanishes almost everywhere on the band (-1,1) if and only if the Boas transform (Bf)(x) satisfies the equation: \displaystyle(B(Bf))(x)=-f(x),\quad x\in\mathbb{R}. Here, B is defined as: \displaystyle(Bf)(x)=\frac{1}{\pi}\int_{0}^{\infty}\frac{f(x+t)-f(x-t)}{t^{2}}\sin(t)\,dt. In [ 34 ] ,Tuan proposed another description of high frequency signals, which enables to describe band-pass and band-stop frequency signals fully. The integral equations, one of the most useful mathematical tools in both pure and applied mathematics, has enormous applications in many physical problems. In 1825 Abel, an Italian mathematician, first produced an integral equation in connection with the famous tautochrone problem. Singular integral equations encountered by Abel can easily be solved by using integral transforms. Also, many initial and boundary value problems associated with ordinary differential equation (ODE) and partial differential equation (PDE) can be transformed into problems of solving some approximate integral equations. The development of science has led to the formulation of physical laws that are often expressed as differential equations, which play a crucial role in solving engineering problems. These differential equations can be transformed into equivalent integral equations. Integral equations are equally important, as many physical problems are directly governed by them, and they can also be converted back into differential equations. Solving integral equations is essential for addressing complex physical and engineering challenges, providing a versatile framework for modeling and analyzing real-world phenomena. Their ability to bridge differential equations and practical applications highlights their significance in scientific and engineering research [ 35 , 36 , 37 ] . Multidimensional signals, depends on multiple independent variables such as space, time, or frequency, play crucial role in modern signal processing and systems theory. These signals are essential in modeling and analyzing complex systems where interactions across multiple dimensions are involved, such as in image processing, video compression, geophysical signal analysis, and multidimensional control systems. The mathematical framework of multidimensional systems involves functions and polynomials in several complex or real variables, which has been instrumental in advancing areas like iterative learning control, distributed network synthesis, and nonlinear system analysis through multidimensional transforms. Multidimensional signal processing has been traditionally based on the concepts and theory of linear systems and Fourier analysis (or other related transforms) [ 38 , 40 , 41 , 39 ] . Recently, there have been significant advancements in the developement of multidimensional integral transforms. Zayed [ 42 ] (2018) introduced a new two-dimensional FrFT and established its properties. Building on this work, Roopkumar et al. [ 43 ] (2019) provided rigorous proofs of the inversion theorems for the multidimensional FrFT on L^{p}(\mathbb{R}^{N}) for p=1,2 , and derived the Parseval’s theorem for L^{2}(\mathbb{R}^{N}) . Additionally, they developed a novel approach to define convolution and correlation for the multidimensional FrFT. Later, Kundu et al. [ 44 ] (2022) extended the theory of the linear canonical transform to the multidimensional domain, establishing its convolution properties and further enriching the framework of multidimensional signal processing and Ahmad et al. [ 45 ] studied uncertainty principles associated with multidimensional LCT. In [ 46 ] , Castro et al. introduced the multidimensional quadratic-phase Fourier transform (MQPFT), extending the one-dimensional QPFT to higher

Read the complete text