Radio Frequency Signal Detection Method based on Dual Clustering Algorithm

Signal detection is an important part of electromagnetic spectrum detection. Currently, the main method of electromagnetic spectrum detection is the energy detection algorithm, which allows the detection of RF signals without a priori information. In this paper, we propose an RF signal detection method based on the dual clustering algorithm, which determines the frequency band where the signal is located by combining eight energy points around each energy point for judgmental clustering, and then secondary longitudinal clustering. By adding a suspicious signal discrimination module to determine whether there is a suspicious signal in the monitoring frequency band. Given the starting and ending frequencies and related csv files, the system can automatically generate waterfall diagrams to mark signals and detect suspicious signals. This paper compares with the DBSCAN-based clustering algorithm, and the detection effect of this method is better than that of the DBSCAN clustering algorithm.