Clutter Environment Perception via Residual Network Equipped with Pyramid Split Attention

With the increasing complexity of the radar working environment, the traditional radar working mode has been unable to allocate its own resources reasonably, which severely limits the performance of the radar. Perception and understanding of the target environment can provide prior knowledge for radar data processing, which is the key to improving radar performance. In order to make full use of the amplitude and phase characteristics of clutter complex data, a dual-channel network architecture is used in this article. In order to obtain better recognition performance, we combine the residual network with Pyramid Split Attention (PSA). Experiments based on measured data show that the method proposed in this paper has excellent recognition performance for typical clutter environments, and is of great significance to the intelligent development of radar.