Abstract: In radar signal sorting based on pulse description words (PDW), the traditional clustering algorithm needs to set the clustering center and the number of clustering in advance. To solve this problem, this paper proposes a new radar signal sorting method, which combines pulse repetition interval (PRI) transform and clustering based on the data field theory. Firstly, noise points are removed based on the potential value according to the data field theory, and then the PRI transformation algorithm is used to obtain the PRI estimate. Based on the PRI value, the preclassification of normalized PDW data are carried out to calculate all kinds of center points, and the kinds with the euclidean distance of its center points less than the radiation factor are combined to determine the initial clustering center and the number of clustering automatically. Finally, the improved K-means algorithm is used for clustering and sorting. Simulation results show that the proposed method can deal with the complex signal environment with frequency agility, repeated frequency fluctuation, jitter, overlapping parameters and partial pulse loss, and the correct sorting probability is obviously improved.

Key words: radar signal sorting, pulse description words (PDW), data field, pulse repetition interval (PRI) transform, K-means clustering