A Hypergraph-Based Multifunction Radar Signal Sorting Method for Mitigating Batch-Increasing in Complex Electromagnetic Environments

Sorting signals from multifunction radars (MFRs) in complex electromagnetic environments is challenging due to the “batch-increasing” problem, where a single MFR’s operating modes are misclassified as signals from multiple emitters. To effectively mitigate this issue, we propose a hypergraph-based method that takes into account interference pulses, missing pulses, and parameter estimation errors inevitable in intercepted radar interleaved pulse sequences (RIPS). Our method begins by removing interference pulses from the RIPS. Then, we apply a selection rule combined with fuzzy c-means clustering to obtain partial clustering labels for the pulse sequences. Next, we construct a hypergraph based on the pulse description word (PDW) parameters and the data potential value of remaining pulses. Finally, using the obtained partial clustering labels, we perform hypergraph learning to effectively sort the MFR signals. Simulation results demonstrate that the proposed method effectively removes interference pulses from the RIPS. The hypergraph construction, which considers the interrelationships between PDW parameters, enhances sorting performance. Compared to existing approaches, our method achieves higher accuracy and robustness under nonideal conditions, effectively alleviating the “batch-increasing” problem in MFR signal sorting.