Waveform optimization for compressive sensing radar imaging

The compressive sensing radar (CSR) has been demonstrated to provide high-quality radar images using fewer data samples than conventional methods. Similar to traditional radar systems, the imaging performance of CSR is also affected by the transmitted waveform. Since a direct optimization of the sensing matrix with respect to the waveform is prohibitive, a CSR waveform design method minimizing a weighted coherence of the sensing matrix is proposed in this paper. A universal CSR model is first established and the cost function for waveform optimization is derived. Then, the simulated annealing (SA) algorithm is introduced to reduce the cost function with binary phase-coded waveform as an example. The proposed algorithm generates the waveform with smaller coherence and inner products compared to the other waveforms widely used in literatures, thus can achieve better performance with superresolution and high accuracy in the CS imaging formation.