Research On ISAR Motion Compensation Technology

Inverse Synthetic Aperture Radar (ISAR) is extensively used in both military and civilianapplications. When ISAR is used in no-cooperative targets, motion compensation is a key problem inISAR imaging, it affects the quality of ISAR imaging directly. The motion of a no-cooperative targetrelative to the radar is complex, it can be decomposed into two parts: the translational motion of areference point in the target and the rotational motion of others about the reference point. Where, therotational part is the fundamental element of the ISAR imaging, and the translational part is harmful.The translational motion influences ISAR imaging in two aspects: First, it causes the rangemisalignment of the range profiles which makes the coherent accumulation of signal in cross-rangeimpossible. Second, it introduces an additive phase component which makes defocusing incross-range. Therefore the translational part should be compensated. Motion compensation can bedivided into two steps: range alignment and phase compensation. In this article, we mainly discuss theabove two techniques.Compressed Sensing is a theory proposed in recent years, the CS theory states that if the signal issufficiently sparse in a certain transformed domain, we can project this signal from high dimension tolow dimension using a observation matrix which is irrelevance with the base matrix, and the originalsignal can be recovered by solving a optimization problem.This idea can reduce sample rate and sharethe pressure in data transmission and store. This theory can be used in ISAR imaging, thus highquality imaging results can be obtained. This article can be divided into six parts:Chapter1explains the basic idea of the ISAR such as the history, the latest development and theapplications of ISAR. Then the history of compressed sensing is introduced.Chapter2exposits the problem of the ISAR motion compensation in detail after the principle ofthe ISAR imaging is described.Chapter3introduces the principles of range alignment. We mainly discuss global minimumentropy algorithm and improved envelope correlation algorithm. The results show that these methodsare effective. Then some classical range alignments are introduced briefly.Chapter4discusses three kinds of phase compensation algorithms: phase gradient autofocus,rank one phase estimation and maximum-likelihood estimation. Then we compare their advantagesand disadvantages by imaging results of real data.Chapter5mainly introduces the compressed sensing theory. Then we analyze the pulse compression and imaging methods which based on the compressed sensing theory. Lastly we brieflyintroduce some range alignment algorithms and phase compensation algorithms, then analyze theperformances of these algorithms in random PRI signal conditions.Chapter6summarizes the whole work of this thesis and points out the problems to be solved inthe future.