ISAR Imaging Based On Compressive Sensing

Inverse Synthetic Aperture Radar (ISAR) imaging technology has been widely used inmilitary and civilian fields because it has advantages of long working distance and little impactby the environment. Generally, ISAR system needs a relatively long period of observationtime for coherent processing in order to obtain high resolution in azimuth direction. However,the targets of ISAR usually have a non-cooperative character, and strong flexibility. In addition,the work pattern of ISAR is usually to track multiple targets, so it is almost impossible toobserve a single target for an enough long term. Therefore the situation of losses of azimuthdata often happens in practice. If the echo data is incomplete, the traditional Range-Dopplermethod will make the image blur because of high sidelobe, and it will greatly affect the qualityof ISAR imaging. Compressive sensing theory has received great attention since it is proposed.It has brought significant changes to many areas with its new signal processing concept. Forradar imaging, compressive sensing theory can make the radar get high quality image by usingvery little data, which not only solves the problem of large amounts of the radar system data,but also ensures ISAR imaging to be able to obtain high resolution image under the conditionof data loss.This paper firstly introduces the basic principles of ISAR imaging, and analyzes theRange-Doppler algorithm based on the ideal turntable model. After the commonly used motioncompensation technology are compared and studied, they are carried out with the measureddata of Yak-42. Then this paper systematically discusses the basic theory of compressivesensing, focusing on studying and discussing the signal reconstruction methods of basispursuit algorithm and orthogonal matching pursuit algorithm, and verifies the validity of twomethods through simulation results of sine signals. In the end, this paper studies the method ofISAR imaging based on compressive sensing theory, to solve the problem of losses of echoazimuth data during the process of ISAR imaging. Basing on whether the losses of data haveaccurate motion compensation, this paper proposes two schemes of ISAR imaging based oncompressive sensing theory, namely the scheme with data after motion compensation usingcompressive sensing theory and the scheme with data before motion compensation usingcompressive sensing theory. The results of the above two schemes used for the measured data of ISAR imaging show that the algorithms of using compressive sensing theory to solve theproblem of data loss of ISAR imaging are obviously superior to the traditional algorithm. Butwhen the scattering points are influenced by interferences or noises, neither BP algorithm norOMP algorithm can prevent reduction of the intensity, and the quality of ISAR imagingdeclines. In order to solve the problem, this paper puts forward the regularization algorithmused for ISAR imaging with echo data containing noises, and proves its advantage by theresult of the data of MIG-25fighter.