Development Of Engineering Software On Detecting Signal Rpocessing Of Inverse Syntehtic Aperture Radar

Inverse Synthetic Aperture Radar (ISAR) is available to search and detect thetarget water-area under all-day, all-weather condition. The information of marinevessels can be estimated and ISAR images of interesting targets can be obtained andutilized for feature extraction and target identification, which would be ofadvantage to enhance the nation’s coastal defense. ISAR is a new generation coastaldefense system, whose signal processing technologies include two parts: detectingand imaging. In this paper, the detecting signal processing technology of ISAR isstudied and a complete signal processing program is proposed. The engineeringsoftware based on practical radar system is developed and the effectiveness isdemonstrated through simulation and ocean experiments.Detecting and estimation are the main purposes of the detecting mode in ISAR.The range-velocity power spectrum is calculated firstly through pulse compressionand accumulation, and then target is detected with the use of const false alarm ratio(CFAR) detection algorithm. The range and velocity of target can be estimatedaccurately with linear interpolation method. By combining the location of target andthe range-velocity complex spectrums in sum and difference channels, the azimuthof target can be estimated via sum-difference amplitude-comparison monopulse.Mean-level (ML) CFAR method is widely used in practical radar system forreal-time processing, which has an excellent performance for single target detectingin homogeneous environment.Tracking is another important task of ISAR. The state of moving target could bepredicted and smoothed by tracking, and the steady track will be utilized formaritime surveillance. Target tracking includes several aspects such as trackinitiation, correlation and quality management. The filtering method for tracking ischosen with theoretical analysis firstly, then the radar plots centroid technologybased on centroid algorithm is studied due to the relationship between antennarotation speed and coherent integration time. The experimental result demonstratesthe effectiveness of centroid technology utilized in the practical radar. Theprocessing method for multi-targets tracking is investigated and a complete trackingprogram is designed based on practical radar system, which has been verified via simulation.An integral signal processing software based on practical hardware platform,which includes range-velocity processing, target parameter estimation andmulti-targets tracking, is developed for detecting mode of ISAR. The correctness oftheoretical method, real time of algorithm and performance of program are takeninto consideration when developing. The engineering software is verified throughpractical onshore and shipborne ISAR experiments.The angle measurement accuracy of sum-difference amplitude-comparisonmonopulse is analyzed based on practical ISAR system. The sum-differenceamplitude-comparison monopulse is used for azimuth estimation, whose accuracy isrelated to the signal-to-noise (SNR). In practical, due to the deficiency of SNR insingle pulse echo, coherent integration is necessary to improve the performance oftarget detection. With the rotation of antenna, the deviation of the target from thebeam center is changed during the coherent integration time. As a result, mainmonopulse ratio detected after coherent integration is approximately the amplitudeand phase weighted sum of all the instantaneous monopulse ratios, which mayaffect the azimuth estimation. In order to reduce this impact, a non-coherentintegration monopulse model is established. In addition, the impact brought bynoise on azimuth estimation is discussed. For the extreme case with low SNR, theazimuth output is set by the use of the two basic antenna beams, where the deviationin practical radar is depicted to be less than the theoretical angle measurementaccuracy.In the lase part of this paper, a novel range and azimuth tracking method viacentroid algorithm with wideband echo is proposed, which makes use of multipleechoes with forgotten factor. The experimental results demonstrate that the rangeand azimuth estimation of the target would become smooth, which will be ofadvantage of tracking performance. Furthermore, even for the extreme case thatwhere target echo is lost temporarily, the range and azimuth could be estimatedaccurately.