Specific Emitter Identification Via Multiple Distorted Receivers

Specific emitter identification (SEI) is a technique that identifies different emitters based on specific characteristics of each individual emitter, which finds broad applications in wireless communication authentication, radio monitoring, electromagnetic environment sensing, and information warfare. In this paper, we consider the SEI problem with unknown receiver distortion, which has not been well studied in the existing work. Two cooperative SEI algorithms are proposed which are composed of two steps, i.e., the feature extraction and the cooperative identification. In the first step, the received signal at each receiver is processed by the intrinsic time-scale decomposition (ITD), and the non-Gaussian features, i.e., the kurtosis and the skewness, are extracted from the decomposed signal. In the second step, the support vector machine (SVM) and the back propagation (BP) neural network are applied to fuse the features extracted from multiple distorted receivers respectively, and then determine the unknown emitters. Simulation results show that the proposed two cooperative identifiers using multiple distorted receivers outperform that using any single receiver in terms of the performance of correct identification. The significance of this paper is that the receive diversity can be achieved by the proposed cooperative identifier by using multiple distorted receivers without receiver distortion compensation.