A Transverse Spectrum Deconvolution Technique for MIMO Short-Range Fourier Imaging

Abstract

The growing need for high-performance imaging tools for terrorist threat detection and medical diagnosis has led to the development of new active architectures in the microwave and millimeter range. Notably, multiple-input multiple-output systems can meet the resolution constraints imposed by these applications by creating large, synthetic radiating apertures with a limited number of antennas used independently in transmitting and receiving signals. However, the implementation of such systems is coupled with strong constraints in the software layer, requiring the development of reconstruction techniques capable of interrogating the observed scene by optimizing both the resolution of images reconstructed in two or three dimensions and the associated computation times. In this paper, we first review the formalisms and constraints associated with each application by taking stock of efficient processing techniques based on spectral decompositions, and then, we present a new technique called the transverse spectrum deconvolution range migration algorithm allowing us to carry out reconstructions that are both faster and more accurate than with conventional Fourier domain processing techniques. This paper is particularly relevant to the development of new computational imaging tools that require, even more pronouncedly than in the case of conventional architectures, fast image computing techniques despite a very large number of radiating elements interrogating the scene to be imaged.Index Terms-MIMO radar, microwave imaging, millimeter wave radar, multistatic radar.

Accelerating short range MIMO imaging with optimized Fourier processing

“…A new short range MIMO imaging technique has been described in this paper following a first publication. 7 The results allow us to anticipate applications in many contexts requiring large volumes of data to be processed rapidly. Partitioning strategies may allow to further accelerate reconstructions 10 and future efforts will enable these algorithms to be ported to more efficient and highly parallelizable architectures in order to ensure their real time application.…”

“…It is then possible to use Stolt interpolation techniques, 4 as well as non-uniform Fourier transforms. 6 A technique is presented in this paper in order to facilitate these treatments, 7 allowing substantial gains in memory consumption, computational speed and quality of reconstructed images.…”

3-D Short-Range Imaging With Irregular MIMO Arrays Using NUFFT-Based Range Migration Algorithm

“…(10) Equation 10gives an analytical approach for generic planar MIMO array imaging. Note a similar equation was derived based on the principle of stationary phase for rectilinear MIMO array imaging in [17] and [18]. To efficiently implement the operation in (10), the spectrum…”

“…To quantitatively compare the focusing performance of the three imaging algorithms, we introduce a power spreading ratio that is defined as the power spreading outside the object normalized to the power within the object volume. It is given by α = P o /P i (18) where P o denotes the power spreading outside the object in the focused image and P i is the power within the object volume. Table II shows the power spreading ratios of the images (i.e., Figs.…”

“…Taking advantage of the fast Fourier transform (FFT), this MIMO-RMA algorithm significantly improves the efficiency of image formation. To further improve the efficiency, a transverse spectrum deconvolution technique [18] was proposed to implement the 5-D to 3-D mapping of the MIMO-RMA in the f -k domain. However, both algorithms are only applicable to the signals acquired with 2-D uniform rectilinear MIMO arrays.…”