Research on traffic prediction model for power communication optical network based on deep learning joint optimization algorithm

With the rapid development of energy transformation and digital technology, the surge of massive distributed energy access, interconnection between data centers under the electric computing collaboration scenario and other new power services have put forward unprecedented needs and challenges to the communication network, and the new era of optical transmission network as the base of arithmetic power needs to solve the problem of large capacity and low latency. Therefore, this paper analyzes the characteristics of traffic data in optical transmission networks and constructs a time series to predict network traffic through neural networks, and the simulation results show that the time series model established in this paper can achieve higher prediction accuracy compared with other comparative schemes.