双连接场景中基于强化学习的基站选择算法
    点此下载全文
引用本文:陈美娟,管铭锋.双连接场景中基于强化学习的基站选择算法[J].南京邮电大学学报:自然科学版,2019,39(6):9~14
摘要点击次数: 281
全文下载次数: 196
作者单位
陈美娟 南京邮电大学 通信与信息工程学院江苏南京210003 
管铭锋 南京邮电大学 通信与信息工程学院江苏南京210003 
基金项目:国家自然科学基金(61771085)和重庆市基础与前沿研究计划(CSTC2015icyjA40040)资助项目
中文摘要:为了满足日益增长的数据服务需求,网络运营商密集部署了大量5G小型基站。在3GPP R14中,定义了5G双连接的场景,即用户可以同时接入5G基站和4G基站。但是,目前主流的基站选择算法并不适用于这种场景。因此,为了解决5G双连接网络中的基站选择问题,文中提出一种基于强化学习的基站选择算法。该算法以用户设备为中心,以最大化用户吞吐量为目标。算法将基站选择问题映射为一个强化学习问题:将用户设备作为学习者,将无线接入技术选择策略作为动作空间,将当前时刻连入基站所获得的吞吐量作为回报值,从而计算出下一时刻选择各个基站的概率。仿真结果表明,在双连接场景中,相比于传统的RSS算法,文中算法可以减少用户设备切换次数,并提高统计时间段内用户的总吞吐量。
中文关键词:双连接  基站选择  5G  强化学习  后悔度
 
BS selection algorithm based on reinforcement learning in dual connectivity scenarios
Abstract:The conventional coprime array direction of arrival(DOA) estimation algorithm has many problems,such as low degree of freedom,small array aperture,phase ambiguity in a specific direction and poor estimation performance on the low signal to noise ratio,etc.To solve these problems, a DOA estimation method for monostatic unfolded coprime array multiple input multiple output (MIMO) radar based on the multiple signal classification(MUSIC) algorithm is proposed. The method combines MIMO radar and unfolded coprime array. By using the unfolded coprime array as a transceiver array, the larger array aperture and the higher degree of freedom are obtained than the uniform linear array and the traditional coprime array. The algorithm still has sharp spectral peaks even SNR is as low as -20 dB.However,the algorithm involves eigenvalue decomposition of high dimensional matrices,leading to the high computational complexity.Consequently,a DOA estimation method for monostatic unfolded coprime array MIMO radar is proposed based on propagation method,thus avoiding the eigenvalue decomposition and reducing the computational complexity.When SNR is greater than 2 dB,the performance is similar.Subsequently,the theory of phase free ambiguity is deduced.Both of the proposed methods are strictly free from traditional phase ambiguity and phase ambiguity in new scenes.Finally,the effectiveness of the methods is verified by simulation.
keywords:direction of arrival(DOA)  multiple input multiple output (MIMO)  unfolded coprime array  phase ambiguity
查看全文  查看/发表评论  下载PDF阅读器

你是第3044585访问者
版权所有《南京邮电大学学报(自然科学版)》编辑部
Tel:86-25-85866913 E-mail:xb@njupt.edu.cn
技术支持:本系统由北京勤云科技发展有限公司设计