Algorithm Of Radar Pattern Clustering In Passive Radar Systems

Abstract

This article discusses the application of signal pattern clustering algorithms to increase the efficiency of passive radar complexes at the stage of detection of marks from targets. The solution of such a problem will allow in the future to select more effective algorithms for secondary (trajectory) processing of targets. Implementation of effective clustering of marks will allow in the future to adequately form the trajectories of target groups, which will lead to energy gain and reduce the number of false runs by an order of magnitude. To solve the problem of radar pattern clustering, we developed the cDBSCAN clustering algorithm, a density-based spatial clustering algorithm with the presence of noise, using a cascade of heuristic rules. The results on the resulting simulation model showed that the algorithm correctly detects patterns in the sample with high accuracy, more than 98%. At the same time, the value of false detection of patterns is about 2%.