摘要
在电力系统和能源互联网的深度融合大背景下,电力通信网融合第五代移动通信技术(fifth generation mobile communication technology,5G)成为一种必然趋势,5G电力通信网业务编排成为支撑电力业务服务质量需求的关键手段。针对电网中现有业务编排优化存在的不足,提出了一种基于软件定义网络(software defined network,SDN)的5G电力通信网业务编排自主决策方法,利用Sarsa(state action reward state action)算法,通过不断与环境交互,在信息不确定的情况下实现接入网信道选择与核心网路由选择的联合优化,从而最小化电力通信终端数据包传输总时延,为电力业务超低时延需求提供保障。仿真结果表明,所提算法能够在传输环境突变的情况下,保持较低传输总时延,并降低波动,有效地提高了电力业务的传输质量。
Under the background of the deep integration of power system and energy Internet,the power communication network integrating the fifth generation mobile communication technology(5G)has become an inevitable trend.The service orchestration in 5G power communication network has become a significant measure to support the quality of service(QoS)requirements of power services.In view of the shortcomings of the existing business orchestration optimization in the power grid,this paper proposes a software defined network(SDN)based service orchestration autonomous decisionmaking method for 5G power communication network.The proposed method utilizes state-action-reward-state-action(Sarsa)algorithm to achieve the joint optimization of channel selection in access network and routing selection in core network under uncertain information through continuously interacting with the environment.The total transmission delay of data packets of power communication terminals is minimized to support the ultra-low delay requirements of power services.Simulation results show that the proposed algorithm can keep low total transmission delay and reduce fluctuation under the condition of transmission environment mutation,effectively improving the transmission quality of power services.
作者
朱校汲
翟明岳
卢文冰
ZHU Xiaoji;ZHAI Mingyue;LU Wenbing(North China Electric Power University,Changping District,Beijing 102206,China)
出处
《全球能源互联网》
CSCD
2023年第3期289-296,共8页
Journal of Global Energy Interconnection
基金
国家电网有限公司科技项目(基于5G电力调控业务应用创新研究,5700-202141442A-0-0-00)。
关键词
5G电力通信网
业务编排
信道选择
路由选择
自主决策
5G power communication network
service orchestration
channel selection
routing selection
autonomous decision-making