| Clutter suppression has always been the key problem in the moving targetdetection process of airborne radar. Based on the space-time domain couplingcharacteristic of the clutter, space-time adaptive processing (STAP) is proposed, whichcan be used to improve the performance of the clutter suppression and the moving targetdetection. At present, the space-time adaptive processing technique has been graduallydeveloped from the theory research to the practical application over the last more thanthirty years at home and abroad.With the purpose of improving the clutter suppression performance and the targetdetection ability, this dissertation focuses on the engineering problems for STAPtechnique applying to airborne radar. The main contents are described as following.Firstly, this dissertation introduces the clutter mathematic model of airborne radar,then analyses the two dimensional space-time clutter spectrum with the influence ofchannel mismatch and intrinsic clutter motion. In addition, the basic principle of STAPis described.Secondly, several robust STAP algorithms are introduced in the presence of smalltraining sample size and interference subspace leakage, both of which can make thecovariance matrix estimation inaccurate. Then their applicabilities are also analyzed.Simulations show that the methods by combining the covariance matrix taper (CMT)approach with either principal components (PC) or diagonal loading (DL) have theadvantage.Finally, focusing on the problem that the non-homogeneous training data can makethe adaptive processing performance suffer in complicated environments, a hybridapproach based on the joint domain localized (JDL) and direct data domain (DDD)algorithms is discussed, and a knowledge-based STAP (KB-STAP) processor whichutilizing echo data is explored in depth. Simulations show that utilizing KB-STAPprocessor can suppress clutter more effectively, and improve the target detectionperformance. |
PDF Full Text Request