科学研究
首页 >> 科学研究 >> 学术报告 >> 正文
学术报告(8月11日):能量收集系统中基于增强学习的多接入控制和电池预测

报告题目:能量收集系统中基于增强学习的多接入控制和电池预测

  

报告人:楚曼 博士

时间:20208月11日上午10:00-11:00

地址:  Zoom 会议, ID:665 6126 2304,密码:123456

会议链接https://cuhksz.zoom.com.cn/j/66561262304?pwd=NG4xN2lDRnh6czVNUENRcXZCSk41UT09

报告人单位:香港中文大学

主办单位:新葡的京集团35222vip

报告摘要:

Energy harvesting (EH) is a promising technique to fulfill the long-termand self-sustainable operations for Internet of things (IoT) systems.In this paper, we study the joint access control and battery predictionproblems in a small-cell IoT system including multiple EHuser equipments (UEs) and one base station (BS) with limited uplink accesschannels.Each UE has a rechargeable battery with finite capacity.The system control is modeled as a Markov decision process without complete priorknowledge assumed at the BS, which also deals with large sizes in both state andaction spaces.First, to handle the access control problem assuming causal battery and channelstate information,we propose a scheduling algorithm that maximizes the uplink transmissionsum rate based on reinforcement learning (RL) withdeep Q-network (DQN) enhancement.Second, for the battery prediction problem,with a fixed round-robin access control policy adopted,we develop a RL based algorithm to minimize the prediction loss(error) without any model knowledge about the energy source and energy arrivalprocess.Finally, the joint access control and battery prediction problem isinvestigated, where we propose a two-layer RLnetwork to simultaneously deal with maximizing the sum rate and minimizingthe prediction loss: the first layer is for battery prediction,the second layer generates the access policy based on the output from the firstlayer.Experiment results show that the three proposed RL algorithmscan achieve better performances compared with existing benchmarks.

报告人简介:楚曼博士分别于2011年和2014年获得西安交通大学学士和硕士学位。2015年-2016年在香港中文大学Professor Vincent Lau团队进行访问;2016-2018年在加州大学戴维斯分校Professor Shuguang Cui团队进行访问; 2019年获得西安交通大学博士学位。目前是香港中文大学(深圳)理工学院的博士后/助理研究员。

研究领域: Deep Learning, Federated Learning, 能量收集,无线资源优化等。

 

上一条:学术报告(9月12日):基于神经影像的智能计算模型构建
下一条:学术报告(8月4日):移动网络中的无人机基站部署
扫一扫分享此页
扫码关注