Sparse reconstruction of multi-component Doppler signature exploiting target dynamics

In this paper, we propose a novel technique for a non-parametric sparse reconstruction of the Doppler signature of multiple ground moving targets exploiting prior knowledge about the target dynamics. The Doppler signature of multiple non-stationary targets is commonly modeled as a multi-component polynomial phase signal, whose order depends on the complexity of target dynamics and the coherent processing interval. In recent years, sparse signal reconstruction and compressive sensing algorithms have been extensively deployed in both parametric and non-parametric estimation of Doppler signatures. The performance of sparsity based Doppler signature reconstruction algorithms is significantly compromised in the presence of strong noise and a large number of missing samples. Herein, we present an effective technique for a non-parametric sparse reconstruction of a multi-component Doppler signature by exploiting the a priori information on target dynamics. Simulation results show that the proposed technique significantly outperforms the conventional sparse reconstruction based methods in such challenging conditions.