A Method of Working Mode Recognition for Multi-Function Radar Based on RGAT

Multi-function radar can flexibly adjust their working modes and carry out multiple tasks in parallel, posing great challenges to radar intelligence reconnaissance. Identifying the working mode of multi-functional radar is the foundation of subsequent threat assessment, adaptive confrontation and guided attack, which directly determines the effectiveness of radar countermeasures. In this paper, a new method of multi-function radar working mode recognition based on Relational Graph Attention Network (RGAT) is proposed. The simulation results show that the accuracy of this paper's method for multi-functional radar working mode identification reaches 98.40% under 5% estimation error, which is higher than that of GAT model and GCN model, and it can effectively identify different working modes; in the case of 10% of spurious pulses and 40% of missing pulses, the accuracy of the identification is more than 94%, which is highly robust.