Cooperative spectrum sensing algorithm based on LSTM network and covariance matrix
Youyao Liu, Juan Li · 2022 14th International Conference on Measuring Technology and Mechatronics Automation (ICMTMA) · 2022
The traditional spectrum sensing algorithm has the problems of difficult detection threshold and low detection performance at low SNR, long network training time and high complexity of the sensing algorithm combined with deep learning. In this paper, a cooperative spectrum sensing algorithm based on Long short-term memory neural network (LSTM) and covariance matrix is proposed. The normalized covariance matrix of received signals is input to LSTM neural network as feature vector for training and classifier is obtained. Finally, the trained model is used for spectrum sensing. The results show that compared with the traditional algorithm and the algorithm using raw signals directly to train the neural network, the algorithm proposed in this paper has a great improvement in detection performance and a short training time.