Research On The Key Techniques Of Signal Processing In Space-based AMTI Radar

Air Moving Target Indication (AMTI) using space-based early warning radar hasimportant significance in military applications. The related signal processing methodshave become a hot research area in the world. For space-based radar (SBR), seriousclutter and interference is induced because of large illumination region. High speedmoving of the radar platform along the orbit brings serious spectrum expansion. Anddue to the Earth's rotation, the crab angle is time variant. High radial velocity betweenradar and target gives rise to large range migration. All these make traditional AMTImethods which have been widely used for airborne early warning radar can not bedirectly applied to space-based radar (SBR). Many challenges have been encountered insignal processing for SBR's AMTI. The research on the signal processing for SBR'sAMTI is therefore has great significance.Aiming at the frontier of this research topic on AMTI for SBR, this dissertationfocuses on four key techniques: the space-based radar signal modeling and cluttersimulation, clutter suppression using space-time adaptive processing (STAP),orthogonal waveform design for mitigation range folded clutter and integrationdetection of high-speed weak target. The main contents of the dissertation are asfollows.1. The models for the space-based radar clutter and the moving target signal areestablished. The characteristics of the clutter for SBR and the influence of earth'srotation are studied. Space-time clutter modeling of SBR is studied. The space-timeclutter simulation data are provided to evaluate the performance of the STAP algorithm.Target echo data are obtained to evaluate the performance of the integration detectionalgorithm. These establish the fundamental base for the following research.2. The STAP clutter suppression and reduced-dimension STAP algorithms for SBRare researched. The performance of the STAP algorithms in the influence of Earth'srotation and ambiguity is investigated. The reduced-dimension STAP based on JointDomain Localized (JDL) algorithm is researched for clutter suppression. A modifiedJDL algorithm based on transform matrix which is composed of measured data is proposed. It is comparably researched with DFT-JDL in SBR environment. Aconclusion that the JDL algorithm based on the transform matrix is better than theDFT-JDL algorithm in space-based radar environment is drawn.3. The requirements of the space-based AMTI radar waveform are researched. Ithas been confirmed that the waveform should has the abilities to mitigate the rangeambiguities and the Doppler ambiguities. In order to meet the requirements forlow-speed target detection, the waveform needs to mitigate the clutter range foldovercaused by the clutter range ambiguities and the greater radar observation's region.Whereas for high-speed target detection, the waveform should be designed to have highDoppler tolerance for large Doppler. The mitigation mechanism of range folded clutterusing orthogonal waveform is studied. The form of the orthogonal waveform isanalyzed. The orthogonal waveform groups by combining chirp waveform withorthogonal binary code by simulated annealing are approached. Simulation results showthat the method has good clutter range folded mitigation performance and high Dopplertolerance.4. The conditions for power accumulation on high-speed weak target detection byspace-based AMTI radar are studied. The high location and the large beam irradiationarea of SBR can meet the condition of illumination time. But target's echo will gothrough several range cells in the accumulation time, which makes SBR more difficultto obtain target's power accumulated for high-speed weak target detection. A hybridintegration algorithm based on Keystone transform and dynamic programmingalgorithm is proposed. The detection time duration is divided into multiple coherentintervals (CPI). Coherent integration is processed based on Keystone transform withineach CPI. Some initial detection results are obtained. Dynamic programming algorithmis applied to determine the target trace. Non-coherent integration is processed along thesame target trace. Final detection results show the low false alarm ratio of the algorithm.Through the hybrid integration algorithm, the detection on low SNR targets can berealized.