Radar Imaging with Quantized Measurements Based on Compressed Sensing

Abstract

In this paper, we consider the problem of radar imaging with quantized data. The quantized CS (QCS) method is used to reconstruct the radar image of sparse targets from quantized data. The reconstruction problem is derived in the maximum a posteriori (MAP) estimation framework and formulated as a convex optimization problem. We compare the proposed method with the traditional l1-regularization method using 1-D simulated data with different quantization bits. For coarse quantization with 1 or 2 bits, the simulation results show that the QCS method outperforms the l1- regularization method in high SNR situations. For high- resolution quantization with more bits, we derive the conditions under which the l1-regularization method and the QCS method are equivalent. This statement is explained theoretically and confirmed by simulation results.