Spatial-Temporal Traffic Prediction Based Load Balancing in Cellular Networks

The rapid growth of cellular network traffic reveals the need for real-time load balancing strategies to improve network performance and resource utilization. To address this issue, a load balancing algorithm based on traffic prediction is proposed. First, a traffic prediction model based on an adaptive hybrid spatial-temporal graph neural network is designed for cellular network traffic prediction. Second, using the predicted traffic data, an ant colony optimization-based load balancing algorithm is proposed, which can dynamically adjust network resource allocation to cope with future traffic peaks and load imbalances. Simulation results show that the load balancing algorithm based on traffic prediction can more intelligently schedule network resources compared to traditional methods, improving the utilization rate of cellular network resources and providing a solution for the sustainable development of mobile communications.