High‐angular‐ and range‐resolution imaging using MIMO sonar with two‐dimensional deconvolution

Abstract

A real‐aperture high‐resolution imaging method utilising a multiple‐input multiple‐output (MIMO) sonar with two‐dimensional (2‐D) deconvolution processing is proposed. Via derivation, the authors show that the conventional beamformer (CBF) output of the MIMO sonar can be approximately expressed as the 2‐D convolution between the original angular and range distributions of scatterers and a 2‐D point spread function (PSF). Hence, to improve the imaging performance degraded by the convolution effect, the 2‐D Richardson‐Lucy (R‐L) algorithm is used to deconvolve the CBF output of the MIMO sonar. Specifically, the authors give the iteration formula of the 2‐D R‐L algorithm attached to the MIMO sonar, and take the absolute values of the CBF outputs of the virtual array as the deconvolution inputs. The analytical expression of the 2‐D PSF is also given, which is designed as the angular‐ and range‐domain (amplitude) responses of a far‐field ideal scatterer located in the normal direction of the virtual array. Meanwhile, the authors point out that the mismatch made by the approximation may degrade the imaging performance, and suggest that a small number of iterations in the 2‐D R‐L algorithm can be used to alleviate the mismatch problem. Via numerical simulations and a tank experiment, the authors show that the proposed method can simultaneously increase the angular and range resolutions and suppress the sidelobes, when compared to the existing MIMO sonar imaging method.

Low-Sidelobe Imaging Method Utilizing Improved Spatially Variant Apodization for Forward-Looking Sonar

“…Deconvolution algorithms have been proposed to improve the angular resolution in spatial spectrum estimation by processing beamformer output data. These methods were subsequently extended to high-frequency sonar imaging to achieve higher resolution in both the angular and range dimensions simultaneously [7,8,12,[22][23][24][25][26][27]. However, the deconvolution results require several iterations, and the performance is related to the number of iterations.…”