Radar Signal Separation and Recognition based on Semantic Segmentation
Hou Changbo, Lijie Hua, Guowei Liu, Lin Yun · 2020
Signal recognition is a key technology in the current information and communication field, and it plays an important role in the civilian and military fields. With the diversification of signal modulation forms, the electronic reconnaissance environment becomes more and more complex, and the signal density continues to increase. Time and frequency domain aliasing will become a serious problem in the field of signal recognition. Consequently, the separation and recognition of multi-component signals is an aspect to be studied in the radar system. This paper extracted the time-frequency images (TFIs) of received signals by Choi-Williams distribution (CWD). Second, the semantic segmentation network DeepLab V3+ is used to segment the TFIs of the signal. Finally, we obtained the recognition result and visualized the result. The simulation results show that the method proposed in this paper can effectively solve the multi-signal separation and recognition of time-frequency domain aliasing.