Enhanced Self-Detection Estimation Compensation for CS-MIMO Radar: Noise and Sparsity

Yanjun Yan,Kai Chen,Yijiu Zhao,Chuanrong Liu,Houjun Wang,Lei Qian,Xiaotian Zhang

DOI: https://doi.org/10.1109/autotestcon47465.2024.10697494

2024-01-01

Abstract:In electronic warfare, the development of multiple-input multiple-output (MIMO) radars and electronic information countermeasures (EIC) equipment affects the power of weaponry, communications between forces, and the coordination of command and control warfare. In an electronic countermeasures environment, the problems of background noise interference and lack of sparse priori information associated with radar signals can degrade the processing speed and accuracy of the channel receiver. Meanwhile, the compressed sensing (CS) method to solve the massive data processing problem of MIMO radar systems is difficult to obtain ideal reconstruction results under low signal-to-noise ratio conditions, especially in the fields of wave-absorbing material frontend characterisation and radar electronic countermeasures. To address the non-adaptive problem of MIMO radar systems to noise and sparse priors, this letter proposes a CS-MIMO radar device self-test-compensation framework with sparsity and noise adaptation, which is divided into two parts: noise compensation and sparsity adaptive selection of measurement optimization methods. First, we establish a channel noise model and theoretically explain the noise folding principle by analyzing the effect of noise on the reconstruction performance of the CS-MIMO radar receiver. Then, we analyze the statistical properties of the compressive measurement noise and its effect on the reconstruction performance and design a compressive domain projection adaptive filter (CPAF) to obtain the position information of the noise subspace. This information is further used to construct a noise compensation matrix and optimize the adaptive filter parameters to filter out the channel noise introduced into the CS-MIMO radar and compress the subband noise background of the receiver. In the second step, we take advantage of the unique spatiotemporal co-sparse characteristics of CS-MIMO radar to estimate the sparse parameters and generate the selection compression measurement matrix through the double improvement of dynamic sparse adaptive and structural sparse, and the selection matrix intelligently selects the measurement of the activation component signals while "shielding" the inactivation components, and we propose the Optimized Selection Matrix Sparse Adaptive Compressed Sensing Algorithm (OSMSAMP). The simulation results support that our improved self-test-compensation framework significantly improves the target detection, parameter estimation and clutter suppression performance of CS-MIMO radar systems.