A Split Pulse Fusion Method for Radar Reconnaissance

In modern battlefields, the electromagnetic environment is complex and changeable, and radar detection and acquisition equipment perform unstably. This often causes pulse splitting in the received pulse stream. As a result, the corresponding data stream shows arbitrary statistical distribution and contains many outliers, leading to errors in signal sorting or identification. To address this problem, a split pulse fusion method for radar reconnaissance is proposed. Firstly, the split pulse fusion algorithm model is constructed, and the pulse is simplified to time point. Then, combining with the idea of robust statistics, a pulse time point clustering algorithm (PTPC) is proposed, and the optimal number of range gates is calculated by fast estimation of the complete pulse flow timing position. Subsequently, the fusion pulse descriptor word (PDW) stream is generated by updating the candidate neighborhood radius, and the fusion result is optimized according to the pulse fusion index. The simulation results show that the proposed method has better performance than the baseline methods in most scenarios with pulse flow characteristic changes, which verifies the stability and superiority of the proposed method.