Research on an inverse synthetic aperture radar imaging algorithm based on non-convex regularization model

一种基于非凸正则化模型的逆合成孔径雷达成像算法研究

¹ Nanjing Research Institute of Electronic Engineering, Nanjing 210007, China
² School of Mathematical Sciences, Nankai University, Tianjin 300071, China

Received: 7 August 2023

Abstract

Inverse synthetic aperture radar(ISAR) is widely used in military and civilian fields because of its ability to image non-cooperative maneuvering targets. Researches show that the compressed sensing technology can be used to improve the resolution and reduce the amount of data required on the ISAR imaging. In this paper, we focus on a classical non-convex regularization model in the field of compressed sensing. For this model, we propose a new algorithm which is based on the MM iteration algorithm framework and adopts the idea of support shrinkage technique, called as iteration support shrinkage algorithm. The new algorithm is simple and efficient, and numerical experiments show that it performs well in ISAR imaging.

摘要

逆合成孔径雷达因其可对非合作机动目标成像而广泛地应用于军事和民用中, 研究表明压缩感知技术可以有效提高逆合成孔径雷达成像的分辨率并且有效降低雷达硬件负担。针对典型的非凸ι_p(0 < p < 1)正则化模型, 基于Majorization-Minimization(MM)迭代算法框架并采用支撑集收缩策略提出了一种新的雷达成像算法——迭代支撑集收缩算法。迭代支撑集收缩算法是一个简单高效的算法, 数值实验表明迭代支撑集收缩算法在逆合成孔径雷达成像中表现优异。