Swarm Sparse Range Profile Recovery from Fewer Measurements of FMCW Radar
Longhui Wang,Zhanyang Ai,Qiexiang Wang,Jian Wang,Xudong Zhang
DOI: https://doi.org/10.1109/taes.2024.3411566
2024-01-01
Abstract:In this article, we propose an algorithm for recovering swarm sparse range profiles from fewer measurements of frequency-modulated continuous-wave (FMCW) radar. First, we give the FMCW radar signal model and show the drawbacks of existing algorithms for the swarm sparse range profile recovery. Next, we propose an algorithm, called swarm orthogonal matching pursuit (sOMP), which recovers swarm sparse signals by exploiting the cluster feature of swarm targets in radar scenes. The highlights of sOMP lie in the two element index selection modes and the likelihood-ratio-based mode selection strategy. The probability distributions involved in the likelihood ratio are derived from the Gaussian approximation of the sensing matrix taken from the complex uniform spherical ensemble or the complex Gaussian ensemble. For the application of sOMP in FMCW radar, we propose two schemes to address the issue of Gaussian approximation of the sensing matrix taken from the Fourier ensemble. Then, the performance of sOMP is analyzed theoretically. In particular, under equivalent parameter settings, the sufficient conditions derived for sOMP can also be regarded as those for generalized orthogonal matching pursuit (gOMP). Compared with latest sufficient conditions for gOMP, our results are looser in terms of the constraint of the minimum magnitude of nonzero elements. Numerical simulations confirm that sOMP in both schemes outperforms the state-of-the-art recovery algorithms. Moreover, real FMCW radar experiments with drone swarms also demonstrate the superior performance of sOMP.