Frequency-coded pulse radar imaging via compressive sensing

A novel compressive sensing (CS) based frequency-coded pulse radar (CS-FCPR) imaging approach is proposed in this paper. FCPR enables high resolution range and Doppler estimation, while transmitting narrowband pulses. Considering spatial sparsity of the radar target scene, FCPR targets echo is first analyzed and the problem of joint range-Doppler estimation is formulated to fit the CS classical framework. Direct solution of this problem using traditional sparse signal reconstruction algorithms needs the statistics of the noise. A cross validation (CV) based robust-SL0 (CV-RSL0) algorithm without requiring the prior information of the statistics of the noise is presented to extract the parameters of targets. The targets parameter extraction performance of the proposed algorithm can rapidly approach the lower bound of the best estimator with the signal to noise ratio improving. Numerical results are compared with the traditional matched filter to illuminate the validity and superiority of this method.