LSTM-Based Jamming Detection for Satellite Communication with Alpha-Stable Noise
Mingqian Liu, Lei Jin, Bodong Shang · 2021
In this paper, we propose an intelligent detection method for satellite communication jamming based on the long and short-term memory (LSTM) networks. The LSTM network can effectively predict the value of the signal at a particular moment. Specifically, a myriad filter is used to suppress the alpha-stable noise in the receiving signal to guarantee that the suppressed noise follows the Gaussian distribution approximately. For the LSTM network training, a priori noise-free jamming signal is used to ensure that the network has a specific memory of the jamming signal. Then, we utilize the trained LSTM network for satellite communication jamming detection. Furthermore, the mean square error between the predicted value and the actual value is examined as the test statistic. It is observed that the mean square error with the only input of noise is larger than that with the inputs of both signal and noise. In addition, a threshold is obtained by recording the noise-only signal into the network multiple times. Consequently, the detection of jamming signals is achieved based on the proposed model. Simulation results show that the detection rate of the satellite communication jamming can reach 98% when the jamming-to-noise ratio is 7dB.