The Application of Unsupervised Clustering in Radar Signal Preselection Based on DOA Parameters

With the deterioration of electronic environment, using signals DOA (direction of arrival) parameters has great significance to preselect multiple radar pulses. Cluster analysis as an important means of data classification, is gradually applied to radar signal sorting. In this paper, a novel method of signal sorting flowsheet is proposed based on Fuzzy Clustering to sort emitters DOA as data objects, with dynamic clustering for reference and Gaussian distance function instead of Euclidean distance. This method avoids establishing enormous similar matrix and adapt to the change of the number of emitters. The result of simulation demonstrates that this method is effective.