Adaptive time-frequency Kernel Local Fisher Discriminant Analysis to distinguish range deception jamming

Abstract

A deception jamming recognition method is proposed based on Adaptive Kernel Local Fisher Discriminant Analysis. The digital radio frequency memory (DRFM) in jammer creates multiple repeat false targets, are commonly utilized in practical applications for limitation of defense radar tracking and discrimination unit. So as to face with decision scheme groups of discriminating among targets and RGPO signals, an analytic form of the embedding transformation and the solution is resorted which can be simply calculated by solving a generalized eigenvalue problem. The practical utility and scalability of the LFDA algorithm can diminish non-linear dimensionality states by applying the kernel trick. The experimental consequences demonstrate that the probability of recognition accuracy performance of the proposed KLFDA in RGPO deception jamming algorithm is greater than 90% when SNR is higher than 4dB. © 2015 IEEE.