异构网络中基于深度强化学习的用户关联与资源分配策略
Strategy of user association and resource allocation based on deep reinforcement learning in heterogeneous networks
符平博 1陶旭 2张见 2李晖2
作者信息
- 1. 南京信息工程大学电子与信息工程学院 南京 210044
- 2. 南京信息工程大学电子与信息工程学院 南京 210044;无锡学院江苏省集成电路可靠性技术及检测系统工程研究中心 无锡 214105
- 折叠
摘要
由于异构网络非凸性和组合性的特点,联合用户关联和资源分配来实现能量效率(energy efficiency,EE)和频谱效率(spectral efficiency,SE)同时最大化的最优全局策略仍然是非常具有挑战性的.基于深度强化学习(deep reinforcement learn-ing,DRL)的方法成为在保证异构网络下行链路用户设备(user equipments,UEs)服务质量(quality of service,QoS)的同时实现联合EE-SE性能最大化的必要解决方案.此外,为解决状态-动作空间下计算量大的问题,引入了多智能体架构的深度强化学习算法(MAD3QN)来获得近乎最优控制策略.仿真结果表明,MAD3QN算法在系统容量方面比DDQN算法和DQN算法分别提高了9.2%和18.2%,在联合EE-SE性能方面分别提高了8.5%和16.6%,提升了系统的有效性.
Abstract
Due to the non convexity and combinatorial characteristics of heterogeneous networks,it is still very challenging to combine user association and resource allocation to achieve the optimal global strategy that maximizes both energy efficiency(EE)and spectral efficiency(SE)simultaneously.The method based on deep reinforcement learning(DRL)has become a necessary solution for maximizing joint EE-SE performance while ensuring the quality of service(QoS)of user equipments(UEs)in heterogeneous networks.In addition,to solve the problem of high computational complexity in the state action space,the double DQN algorithm with multi-agent dueling architecture(MAD3QN)was introduced to obtain almost optimal control strategies.The simulation results show that the MAD3QN algorithm has increased system capacity by 9.2%and 18.2%respectively compared to the DDQN algorithm and DQN algorithm,and improved joint EE-SE performance by 8.5%and 16.6%respectively,enhancing the effectiveness of the system.
关键词
深度强化学习/用户关联/资源分配/能量效率/频谱效率Key words
deep reinforcement learning/user association/resource allocation/energy efficiency/spectral efficiency引用本文复制引用
基金项目
国家自然科学基金(61661018)
江苏省基础研究计划青年基金(BK20210064)
无锡市科技创新创业资金(WX03-02B0137-022200-34)
出版年
2024
NETL
NSTL
万方数据