Study Of High Resolution ISAR Imaging For Mono-and Bistatic Radar

As an important tool for microwave detection, ISAR (Inverse synthetic apertureradar) plays an increasingly significant role in many fields, e.g. territory defense andspace detection. Due to the capability of obtaining images of non-cooperated targetsunder the condition of all weather, all day and far distance, it has great value for militaryand civil application. The increased development of requirements also leads to furtherstudy on ISAR imaging techniques. This dissertation makes a detailed research on thenew techniques in high-Resolution Imaging processing for Mono-and Bistatic Radar.1. This dissertation introduces the research background and significance of thisdissertation, and describes the history and present situation of ISAR imaging.2. This dissertation demonstrates the principle of high range and azimuthresolution in ISAR imaging, provides the basic method for translational motioncompensation and ISAR imaging, and reviews the research results about ISAR imagingmethods at home and abroad.3. This dissertation makes a study on the frequency synthesis processing oflinearly modulated stepped frequency signals. Firstly, the form of linearly modulatedstepped frequency signals is presented, as well as the problem of frequency bandwidthsynthesis caused by the phase error of target translational motion. Then it illustrates abandwidth coherence processing method of echoed signals with low SNR based on highorder cumulants technique, and also lists the imaging flow. Simulation and the real dataexperiment both demonstrate the effectiveness of the method proposed in this chapter.4. This dissertation considers the problem that it may result in artifacts andscatterers omission to judge scatterers number in a range cell with energy in traditionalsuper-resolution ISAR imaging. It proposes a statistical RELAX for super-resolutionISAR imaging. To extract the scatterers, this scheme well exploits the statisticsinformation of background noise and clutter of the environment where the target locates.The echoed signals in all the range cells are detected with Gaussian test and thethreshold is set according to the expected false alarm rate. When the range cellscontaining scattererss is determined, scatterers extraction can be implemented viastatistical RELAX super-resolution techniques. The effectiveness of this method isconfirmed by the simulation and real data experiments results.5. This dissertation puts forward a modified CS imaging method, with respect tothe problem of performance reduction or even inefficiency for ISAR imaging based onCompressive sensing (CS) with low SNR. The range cells which contain targets and noise are separated by energy threshold, and the SNR of the measured data is increasedby coherent projection in azimuth. With the iterative weighted norm-1optimization, thetrue scatterers energy can be enhanced while the noise is reduced. Since the modifiedCS algorithm utilizes a few pulses to obtain high-resolution image, it is suitable forISAR imaging in the condition of low SNR. Real data experiments indicate that thisapproach is effective to reduce noise and clutter. High-resolution ISAR images can beacquired with limited few echoed pulses, which also verifies the robustness of thismethod in the situation of quite limited pulses.6. This dissertation for the time-variant property of bi-static angles in two poles ofbistatic radar, conventional ISAR imaging methods are incapable to achieve highresolution target images. In this chapter, a super-resolution imaging method based onRadon transform combined with TCDS (Time Chirp-rate distribution-Search) isproposed. First of all, the signal model for bistatic ISAR is presented, and it analyzes thetime-variant property of the bistatic angles in two poles areas, as well as its influence onrange envelope and azimuth. A Radon-TCDS-Relax super-resolution imaging algorithmis brought up for the effect from high-order azimuth terms of target motion model inpole areas. This method attempts to search for the frequency modulation rate and its rateof the high-order azimuth terms in each range cell and its error is compensated.Scatterers extracting and imaging are achieved by Radon-TCDS-RELAX and TCD-RIDrespectively. With the comparison between simulated data and traditional RID imagingmethod, the scheme proposed in this chapter shows great advantage and effectiveness.7. This dissertation makes a study on bandwidth synthesis of ultra widebandfrequency modulated signals with LMSF. It firstly analyzes the problem of frequencybandwidth synthesis brought by phase error from the target motion. Then it gives anaccurate estimation method for motion parameters within one burst in ultra widebandstepped frequency synthesis based on image entropy minimization. In addition, MTRCphenomenon in ultra wideband ISAR imaging is analyzed, followed by thecorresponding solution. Simulation experiments confirm that the algorithm presented inthis chapter is effective for high-resolution ultra wideband ISAR imaging.