Time-Delay Estimation of Off-grid Targets for One-bit Compressed Sampling Radar via Polar Interpolation

Sparse reconstruction-based time-delay estimation for one-bit compressed sampling radar usually requires targets to be located on discrete time-delay grids. For the off-grid targets, the estimation performance significantly degrades due to the model error resulting from the discretized sparse representation dictionary. In this paper, we propose a binary iterative hard thresholding with polar interpolation (Polar-BIHT) method to estimate the time-delays of off-grid targets. The method expresses radar echoes by polar interpolation approximation, and thus transforms the time-delay estimation into a one-bit sparse reconstruction sub-problem and a one-bit nonlinear optimization sub-problem. By solving the two sub-problems efficiently, it can accurately estimate the time-delays. The simulation results show Polar-BIHT is superior to the binary iterative hard thresholding method. Moreover, Polar-BIHT still has low computational cost, which is significant in practical application.