Multifunctional Radar Work Mode Recognition Based on Transformer and Residual Module

The rise of multifunctional radar and the development of low probability of intercept(LPI) technology have brought great challenges to electronic reconnaissance. With the development of radar technology, radar has enough resources and does not use pure search mode alone, and the detection intent of radar is mainly reflected in the tracking data rate. To overcome these problems, firstly, the radar work modes and LPI technology are analyzed and pulse data with different detection intent based on the LPI strategy are generated. Subsequently, a ResNet-Transformer with Batchnorm based work mode recognition network(RT-Net) is proposed which uses residual module to perceive features, and uses the attention mechanism to obtain the global information, and the use of Batchnorm further improves the generalization and stability of the network. Finally, simulations are carried out to prove the superiority of the proposed network with respect to other networks.