Classification of Low RCS Targets Based on PCL Radar Network Using Statistical Hypothesis Test

In this paper, we propose a new framework of target classification for a passive coherent location(PCL) radar network. The framework uses radar cross sections(RCSs) obtained from multiple bistatic radars, and is computationally more efficient compared with the conventional method that uses time-varying RCSs obtained from a monostatic radar. Firstly, we construct the training set of the bistatic RCS distribution of each target using the scenario-based method and a PCL radar network with multiple transmitters and a receiver. Next, assuming that a test sequence consists of bistatic RCSs, we classify each target using statistical hypothesis test algorithms, such as Z-test, Wilcoxon test, and sign test. The proposed framework demonstrated better performance than the conventional method, in terms of computational efficiency.