Airborne Radar Target Recognition Technology Research

This thesis is focused on the study of classification algorithm for radar airborne objectives. Two approaches of recognition are studied. The main creative work includes:a) A novel radar target recognition algorithm based on multiple-polarized range profile is presented. It is a recognition method for wideband polarization radar. First, the Radar Cross Section of targets in four kinds of polarization is given. Then, the approach to construct the multiple-polarized range profile matrix is introduced. At last, the classification is performed by the Support Vector Machine (SVM). Classification program is compiled and the algorithm is verified by the real targets data and simulation results.b) Improving Greco software to compute the Radio Cross Section of mobile. Micro-Doppler effect signal modeling is performed. Empirical Mode Decomposition is adopted to extract features of micro-Doppler signal and classification using nearest neighbor algorithm is performed. The recognition results of three kinds of simple targets are shown. Program is compiled and the algorithm is validated by the simulation results and the real targets data.