Interference Signal Suppression Algorithm Based on CNN-LSTM Model

Abstract

Sensors and anti-interference technology have a complementary relationship. The anti-interference capability directly affects the measurement accuracy, reliability, and stability of sensors. In complex electromagnetic or natural environments, sensors are inevitably influenced by various interference sources. Effective anti-interference technology is the key to ensuring the normal operation of sensors, and suppressing interference signals is one of the key links to improving communication quality. This paper proposes a CNN-LSTM-based interference signal suppression algorithm, aiming to enhance the anti-interference capability of wireless communication systems through deep learning technology. The algorithm utilizes CNN to extract the spatial features of the signal and LSTM to capture the temporal dynamic characteristics of the signal, outputting a predicted signal to effectively suppress interference signals. The performance of the experimental simulation algorithm under different interference scenarios was evaluated and compared with three models: LSTM, BO-LSTM, and CNN-GRU. The results demonstrated that this algorithm had a small error and a high degree of regression fitting. Finally, the effectiveness of the algorithm was verified by using the signal propagation model based on ITU-R P.1546 and the publicly available noise datasets collected from the actual environment. The research shows that this algorithm can significantly suppress the influence of interference signals and environmental noise on useful signals, providing a basis for promoting the evolution of sensors towards higher reliability and robustness.