基于改进的正则化子空间追踪算法的ISAR成像

摘要

针对ISAR成像面对的大带宽和高速率采样等大数据问题,提出了基于改进的正则化子空间追踪算法(Modified Regularize Subspace Pursuit,MRSP)的ISAR成像方法.该方法采用压缩感知理论,将子空间追踪(Sub-space Pursuit,SP)算法与正则化思想结合,采用回溯反馈和贪婪精选的方法进行支撑集选择,并利用黄金比例分割法确定正则化过程实现支撑集的二次筛选,最终能实现信号的精确重构.仿真实验表明,该算法对较少的回波数据进行重构,既能获得更加清晰的成像结果,又能进一步提高运算速度,是一种有效可行的重构算法.%In order to solve the large data problem from radar imaging due to large bandwidth and high speed sam-pling,and so on,an improved method of ISAR imaging algorithm based on modified regularize subspace pursuit method (MRSP)was proposed.The algorithm combined a regularization technique and the subspace pursuit(SP)algorithm to select the atoms in the support set,with feedback method and greedy algorithm.The golden section method was used to determine the regularization process for second selecting.Simulation results showed that the algorithm could achieve a more clear ISAR image with lesser echo data and faster speed in computation.