Study On Adaptive Constant False Alarm Rate Detection Under Complex Environment

As a major means to achieve adaptive target detection in radar system,constant false alarm rate(CFAR)processing technique plays an important role in radar signal processing.With the development of radio technology,now the radar system requires a higher detection performance in the complex electromagnetic environment.However,the conventional CFAR processing methods are only applicable to some specific environments,and it is difficult to maintain a good detection performance in complex environment.Therefore,how to adaptively choose the appropriate CFAR processing method according to background environment has became the main problem of target detection.This thesis mainly studies the adaptive CFAR algorithm under complex environment.Firstly,two conventional CFAR processing methods are introduced,including the mean level CFAR algorithm and the ordered statistical CFAR algorithm.For the mean level CFAR algorithm,CA-CFAR,GO-CFAR and SO-CFAR are discussed in details.The fundamental principles of these methods are introduced in this thesis.Then their detection performance in different clutter environments is analyzed by simulation experiments,and their advantages and disadvantages as well as the applicable environment are also discussed.The adaptive CFAR processing algorithm is dedicated to improving the performance of radar target detection in complex environment.Secondly,two conventional adaptive CFAR methods,VI-CFAR and HCE-CFAR,are studied in details,and their performance in different clutter environments is analyzed.Considering the performance degradation of VI-CFAR in multi-target environment,an improved method,named VIHCEOS-CFAR,is proposed based on VI-CFAR,HCE-CFAR and OS-CFAR.The simulation results show that the VIHCEOS-CFAR can achieve a better detection performance in different clutter environments.Finally,two adaptive CFAR processing methods based on neural network and K-nearest neighbor algorithm are discussed.The basic idea of these two methods is to use classification algorithm to distinguish the background type of the cell to be detected,and then chooses CFAR processing method adaptively according to the result of type recognition.The simulation results show that the two methods discussed in this thesis have a good detection performance in homogeneous environment and multi-target environment,and also have the great ability to control false alarm rate at clutter edge,which further proves that these two methods have the strong adaptability and robustness in complex environment,thus improving the target detection performance of radar system.