基于Actor-Critical架构的5G Massive MIMO波束能效的研究与应用

Research and Application on Multi-beam Energy Efficiency of 5G Massive MIMO Based on Actor-Critical Architecture

大规模阵列天线技术(Massive Multiple Input Multiple Output,Massive MIMO)作为第五代移动通信(5G)的无线核心技术,实现了多波束空间覆盖增强,然而5G Massive MIMO的多波束射频高能耗、多波束碰撞和增加的干扰造会成5G网络能效下降,运营成本增高。基于3D数字地图、基站工程参数、终端上报的测量报告/最小化路测(Measurement Report/Minimization of Drive Test,MR/MDT)数据、用户/业务分布构建的三维数字孪生栅格,通过卷积长短期记忆(Convolutional Long Short Term Memory,Conv-LSTM)算法对栅格内的用户分布、业务分布进行分析和预测,通过Actor-Critic架构对5G波束配置和优化策略进行评估,实现不同场景、时段的5G波束最佳能效,智能适应5G网络潮汐效应,实现“网随业动”。

As the key wireless technology of the 5th generation mobile communication(5G), Massive MIMO realizes the enhancement of space coverage by multiple narrow beams. However, the high energy consumption, multi-beam collisions and increased jamming of 5G Massive MIMO can cause the decrease of the energy efficiency and the increase of operating expense. Based on the 3D digital map, base-station engineering parameters, MR/MDT data reported by terminals, and user/service distribution, this paper constructs a 3D digital-twin grid. The Conv-LSTM(Convolutional Long Short Term Memory) algorithm is used to analyze and predict the user distribution and service distribution within the grids. By evaluating 5G beam configuration and optimization strategies through the Actor-Critical architecture, the optimal energy efficiency of 5G beams for different scenarios and periods is achieved, thus intelligently adapting to the tidal effect of 5G networks and realizing “network following service”.