A Rapid Radar Signal Sorting Method Based on Density and Gaussian Mixture Model Data Stream Clustering

Radar signal sorting is one of the crucial techniques in radar reconnaissance. However, as the electromagnetic environment increasingly complex and the density of radar pulses surges, the efficiency of clustering-based sorting algorithms is severely degraded. To better align with the streaming data characteristics of radar pulses and avoid the storage and computation of large amounts of pulse data, this paper proposes a rapid radar signal sorting method based on density and Gaussian mixture model data stream clustering (DGStream). Firstly, by introducing the concepts of core clusters and outlier clusters based on a Gaussian mixture model (GMM), the proposed method achieves effective storage of online summary information for radar emitter signals with reduced space complexity. Meanwhile, by optimizing strategies for cluster evolution and incremental updates during the online phase, and by directly outputting clustering results using GMM in the offline phase, the approach achieves a significant reduction in computational load and further enhances overall efficiency. Experimental simulation results demonstrate that the proposed method excels in terms of efficiency, providing a practically valuable method for sorting radar pulse streams.