A Visual Analytics Approach for Radar Signal Clustering

Radar signal clustering seeks to identify radar signals that belong to each radar radiation source from the radar signal data. With the rapid development of radar technology, the unevenly distributed and highly overlapping attributes of radar signals often result in unsatisfactory results by means of traditional radio signal clustering methods. This paper proposes an interactive clustering approach for radar signal data. Firstly, we propose an improved clustering method that integrates a kernel density estimation, density clustering and permutation entropy technology to process the radar signal data. Then, we design and implement a visual analysis system that presents the radar signal data and automatic clustering results from multiple perspectives. The system also provides a set of interactions to involve users in the entire process of radar signal clustering.