Passive Radar Super-Resolution Imaging Technology

Passive radar not only has the characteristics of available and abundant illuminators of opportunity and stealth, but also can resolve the problems of ground clutter jamming and low-flying target detection, which brings broad application prospects. On the basis of summarizing passive radar technology, exploratory and innovative research on passive radar imaging algorithms is developed in the thesis. Considering the actual factors, three passive radar imaging algorithms for small rotation angle and short-time accumulation are investigated. The main research works are as follows:1. Two typical imaging modes of passive radar including multi-illuminators and single-receiver(MISR) and single-illuminator and multi-receivers(SIMR) are introduced, as well as their geometric model and echo signal model. Regarding the echo signal model of MISR, the echo data are converted to a parametric model containing spatial frequencies and amplitude information of scatterers through time sampling. The RELAX algorithm for estimating the spatial frequencies and amplitude information is proposed to reconstruct the target image. The effectiveness and correctness of the proposed algorithm are verified by simulation experiments.2. A passive radar super-resolution imaging algorithm based on compressive sensing(CS) is proposed. Based on the sparse representation of discretized echo signal, the CS-based algorithm is used to estimate the spatial frequencies and amplitude information, and a search algorithm is applied to reconstruct the target image. Numerical simulation indicates that in the case of low signal-to-noise ratio(SNR) and limited measured data, compared with the ESPRIT and RELAX algorithms, the proposed CS algorithm can achieve more robust performance.3. A novel algorithm for passive radar imaging via sparse representation of autocorrelation matrix(SRAM) is put forward. Based on the fact that the estimation error of autocorrelation matrix of the echo signal obeys asymptotic Gauss distribution, the SRAM algorithm is applied to reconstruct the sparse matrix. The scatterers’ locations are obtained by searching the non-zero elements of the matrix, and meanwhile, the scatterers’ amplitudes can be calculated through the least-square estimation method. Under low SNR and limited observation data, the imaging performance of the proposed SRAM algorithm is better than that of the CS-based algorithm.