Radar emitter identification based on EPSD-DFN

Aimed at the problem that the radar emitter characteristic parameters are not fully used in the identification process and the stability of the radar parameter of sample map is not stable enough, the estimation parameter sample diagram (EPSD) is used to describe the characteristic parameter description model of the radar radiant source. A radar emitter identification method combining EPSD with Deep Feedforward Network (DFN) is proposed. By calculating the statistic corresponding to the numerical sequence in the radar emitter parameter of sample map, the numerical sequence is replaced by the obtained statistic in the parameter sample diagram, thus the parameter sample map mixed with the numerical and numerical sequence is simplified as the estimation parameter sample diagram. DFN is used to train the training samples of EPSD, and network precision adjustment is achieved through test results. Simulation results verify the effectiveness and stability of the PESM-DFN method.