摘要
在超密集网络(UDN)中,密集部署小基站会产生大量能耗及温室气体排放等问题。为此,我们提出了一种联合小基站睡眠和功率控制的分步自适应动态功率控制(ADPC)算法用以改善网络能效。此算法分为两个阶段:第一个阶段,提出基于网络负载状况的小基站睡眠策略;第二阶段,对处于激活状态的小基站进行功率控制。并进一步提出基于强化学习(RL)及深度神经网络(DNN)的框架以优化小基站功率控制。通过与典型Q-学习框架的功率控制算法进行仿真比较表明,所提方案具有很好的自适应能力,使网络能效提升13%,对于更加密集的场景仍具有很好的性能。
In ultra-dense networks(UDN),dense deployment of small base stations will cause the issues of huge energy consumption and greenhouse gases emission.For this,based on combining the sleep and power control of small base stations,we propose a two-step adaptive dynamic power control(ADPC)algorithm to improve network energy efficiency.In the first phase,we propose a small base station sleep strategy based on network load fluctuation.In the second stage,we further propose a framework based on reinforcement learning(RL)and deep neural network(DNN)to optimize the power of active small base stations.Comparing with the typical power control algorithm based on Q-learning framework,simulation results show that our proposed scheme has good self-adaptation ability and can improve network energy efficiency by 13%,and still has good performance for more dense scenes.
作者
郑冰原
孙彦赞
吴雅婷
王涛
Zheng Bingyuan;Sun Yanzan;Wu Yating;Wang Tao(Shanghai Institute for Advanced Communication and Data Science,Key laboratory of Specialty Fiber Optics and Optical Access Networks,Joint International Research Laboratory of Specialty Fiber Optics and Advanced Communication,Shanghai University,Shanghai 200444,China)
出处
《电子测量技术》
2020年第9期133-138,共6页
Electronic Measurement Technology
基金
国家重点研发计划(2017YFE0121400)
国家自然科学基金(61501289,61671011,61420106011)项目资助
作者简介
通信作者:郑冰原,硕士研究生,主要研究方向为超密集网络、能效优化。E-mail:bingyuanzheng@163.com;孙彦赞,副教授,主要研究方向为无线通信资源管理、干扰协调、绿色通信。E-mail:yanzansun@shu.edu.cn;吴雅婷,副教授,主要研究方向无线通信OFDM、MIMO系统资源管理。E-mail:yt-wu@shu.edu.cn;王涛,教授,主要研究方向为无线通信资源管理,Relay通信,绿色通信。E-mail:twang@shu.edu.cn
