Satellite Network Traffic Prediction Based on LSTM and GAN
Jingyi Cai, Shutian Song, Haipeng Zhang, Ruiliang Song, Bo Zhang, Xiang Zheng
Abstract
Satellite networks are characterized by rapid topology changes, quick updates in the coverage of subsatellite points, and large variations in service traffic access in different regions, but they are also likely to cause congestion and blockage in the network. In order to solve this problem, a network traffic prediction method based on long short-term memory (LSTM) and generative adversarial networks (GAN) was put forward. Firstly, the network traffic simulation dataset is constructed based on the population distribution density. Secondly, the dataset is augmented with GAN to prevent the occurrence of training overfitting problems. Finally, LSTM-based model is trained and tested on the dataset obtain a network traffic prediction model with an accuracy of 95.43%, which can provide effective data support for the coordination of satellite network resource scheduling.