基于改进SAMP的频率捷变和PRI抖动雷达群目标检测

针对群目标场景下目标个数未知和网格失配影响稀疏重构效果的问题，提出一种改进的稀疏度自适应匹配追踪（sparsity adaptive matching pursuit,SAMP）算法用于频率捷变和脉冲重复间隔（pulse repetition interval,PRI）抖动雷达的动目标检测.改进算法回溯阶段的原子选择策略，通过正则化挑选出能量相近的原子进行网格失配矫正，可以有效减小重构误差、抑制重构伪峰.根据重构信号的功率变化率自适应扩展支撑集长度（群目标个数估值），减少算法在同一步长的搜索时间.仿真结果表明，相比于SAMP算法和传统的群目标检测算法，本文改进的算法在提升重构效果的同时也能减小距离和速度测量误差.

Frequency Agile and PRI Jitter Radar Group Target Detection Based on Improved SAMP

To improve sparse reconstruction effect caused by unknown number of targets and mesh mismatch in group target scene, an improved sparsity adaptive matching pursuit (SAMP) algorithm was designed for target detection of frequency agile and pulse repetition interval (PRI) jittered radar. First, improving the atom selection strategy in the backtracking stage, selecting atoms with similar energies through regularization for grid mismatch correction, the algorithm was arranged to effectively reduce the reconstruction error and suppress the reconstruction spurious peak. Then, adaptively expanding the search step (group target number estimation) according to the power change rate of the reconstructed signal, the algorithm was arranged to speed up algorithm search. Simulation results show that, comparing with SAMP and traditional group target detection algorithm, the improved SAMP algorithm can improve reconstruction effect and reduce the target range and velocity estimation error.