Sparse reconstruction-based time-delay estimation algorithm in the exponential correlation domain

For the estimation problem of multiple adjacent targets, traditional correlation function method based on matched filter (MF) is often limited by resolution, resulting in insufficient accuracy. Subsequently, an exponential (EF) with high resolution was constructed introducing a controllable exponent p into frequency response [17]. In this paper, we propose sparse reconstruction and filter-based algorithm to estimate time delays targets. This uses autocorrelation (the output filter) as template constructs overcomplete model received signal domain. An interior-point xmlns:xlink="http://www.w3.org/1999/xlink">l 1 -norm regularization then employed solve proposed model. Notably, does not require prior knowledge target number. Theoretical analysis Simulation results demonstrate that, can effectively reduce coherence between targets thus achieve high-accuracy time-delay estimation. When input SNR higher than -24 dB, it has accuracy for multi-target (especially targets) previous MF or EF algorithms. Particularly, at signal-to-noise ratios, error approach bound derived from Cramér–Rao lower bound; when -4 achieves completely accurate