Study On The Algorithms Of Radar Target Recognition Using High Resolution Range Profiles

In this paper, the problem of recognition of radar target using high resolution range profiles is mainly studied by the means of techniques in modern signal process and in modern pattern recognition. The aim is to make the arithmetic of high resolution radar target recognition have less complexity and identify targets more accuracy. In the paper, there are:At first, it makes a brief introduction to the techniques in recognition of radar target using high resolution range profiles. Some representative misleading areas and key points in the process of the study on recognition of radar target are pointed out.The principle and simulation method of high resolution range profiles is studied,and the pose sensibility of range profiles is investigated. After introducing the characteristics and classic simulation method of high resolution radar clutter, a neural network type empirical iteration method and a neural network Trained by natural gradient algorithm is proposed to simulate correlated non-Gaussian radar clutter by zero-memory nonlinearities (ZMNL).In the paper, a method is presented which gets radar target high dimensional structure information from one dimensional range profiles based on Independent Component Analysis (ICA). Then after discussing the limitations of ICA, two modified algorithms of radar target 3D imaging using clustering analysis and quantization statistical analysis are proposed. The emulation results are presented, analyzed and compared.In conclusion, the main works of the dissertation are summarized and the future research areas are pointed out.