基于T-GCN的4G/5G基站节能减排智能决策系统

T-GCN Based Intelligent Decision-making System of 4G/5G Base Stations for Energy Saving and Emission Reduction

随着4G/5G移动互联网的快速发展,为了满足不断增长的流量需求并提升蜂窝网络的覆盖率,基站的流量负荷呈爆炸式增长。在全球能源短缺的背景下,要实现碳达峰、碳中和的目标,在不降低用户通信质量的前提下,如何对基站进行精准开关控制,使其能耗减小到最低是一个重要问题。为此,在栅格模型和基站能耗计算模型的基础上,提出了一种基于时间图卷积网络(Temporal-Graph Convolutional Network, T-GCN)预测和自设计启发式算法关断决策的基站智能决策系统,实现了基站的智能开启和关闭。同时保证符合实际约束,从而得以提高网络资源管理的效率并优化网络能耗性能。通过仿真实验,流量预测效果良好,在一定范围内得到了理想的基站开关决策结果。

With the rapid development of 4G/5G mobile internet,in order to meet the growing traffic demand and improve the coverage of cellular networks,the traffic load of base stations is increasing explosively.In the context of global energy shortage,in order to achieve the goal of carbon peak and carbon neutrality,without reducing communication quality,it is an important issue to know how to accurately switch the base stations to reduce their energy consumption to the minimum.Therefore,based on the grid model and the base station energy-consuming calculation model,an intelligent base station decision system based on Temporal-Graph Convolutional Network(T-GCN)prediction and self-designed heuristic algorithm is proposed to realize intelligent opening and closing of base stations.At the same time,it is guaranteed to meet the practical constraints,so as to improve network resource management efficiency and optimize network energy consumption performance.Simulation experiment show that the flow-predicting consequence is good and an ideal decision result of base station switch is obtained in a certain range.