Squeeze-Excitation Based Communication Behavior Recognition Approach for Spectrum Sensing Data
Kaixin Cheng, Lei Zhu, Changhua Yao, Lei An, Guizhen Han · 2021 7th International Conference on Computer and Communications (ICCC) · 2021
Communication behavior recognition under electronic countermeasures conditions has become an increasingly important topic for communication security and military application. This paper proposes a squeeze-excitation-based communication behavior recognition approach to adapt to the demand for rapid response under electronic countermeasures conditions. Unlike previous studies, a more practical communication scene is constructed to acquire convincing data. To speed up the recognition process and realize better performance on recognition, the squeeze-excitation network is utilized to apply the attention mechanism in the recognition approach to speed up the learning process. The comparative experiments show that the proposed approach can realize ideal recognition performance much faster compared with the traditional method, which is adaptive to the electronic countermeasure environment.