Discrimination of Deception Targets in Multistatic Radar Based on Clustering Analysis

Abstract

Multistatic radar has attracted considerable interest. One of the reasons is that it can provide improved performance against deception jamming. However, the existing anti-jamming methods always need multiple pulse repetition intervals (PRIs), and the slow fluctuating targets will be misjudged as deception targets. In this paper, hierarchical clustering analysis method is first utilized in multistatic radar to discriminate deception targets within one PRI in a defined amplitude ratio feature space, in which deception targets concentrate on one position, while physical targets possess dispersion characteristic. Two approaches are provided to terminate the clustering process by optimizing the number of clusters or designing the minimum cost among clusters. The first one works in most cases without any prior information, but is not in Neyman-Pearson sense. The other one can achieve a constant probability of false alarm, but obtain its optimal performance only when the jamming-to-noise ratios of all generated false targets are approximately equal. According to the final clustering results, deception targets can be discriminated by deciding that the targets in the cluster with more than one target are all false targets. The simulated results are covered to illustrate the feasibility and validness of the proposed approach.