Specific Emitter Identification Based on Homomorphic Filtering and Support Vector Machine-2K

Specific emitter identification (SEI) is the process of identifying or discriminating different emitters by extracting the radio frequency fingerprints from the received signals. A novel SEI scheme with two steps is proposed in this paper. In the first step, the new fingerprint features are extracted as the emitter-irrelated information is suppressed by homomorphic filtering. Then, Two View SVM-2K (Support Vector Machine on two Kernels) classifier is exploited to classify emitters effectively based on the above features. Simulation results show that the proposed method achieved better classification performance than the benchmark method.