High Range Resolution Profiling of Moving Targets via Compressive Sensing in Stepped Frequency Radar

Abstract

With respect to the problems of position mismatch, oversampling redundancies, distortion and dispersion in range profile caused by radical motion between radar and the interested targets in stepped-frequency radar system, a combined range-velocity estimation method is presented based on the compressive sensing (CS) theory. Considering the sparsity of the interested targets in the illuminated scene by the radar beam, a sparse presence model of the targets is established, and then the range and velocity of the targets are co-estimated directly in the range-velocity plane employing multi-snapshots data, thus the redundancies and mismatch problems existing in traditional method are avoided, besides the output signal-noise-ratio improves dramatically in this case. To reduce the complexity of the computing process, a dynamic sensing matrix is constructed using information acquired from pre-estimated range and velocity, following with which a singular value decomposition (SVD) of the reduced echo data is employed to further reduce the computation load. The effectiveness of the proposed method and robust performance to noise are validated by simulation and experimental results.