摘要
针对当前网络切片资源分配方案不灵活的问题,提出一种基于强化学习的5G无线资源管理技术,该方法旨在实现自适应的网络切片动态优化和端到端服务的可靠性。首先,计算不同用户优先级实现基于用户优先级差异化的分配策略;然后,将用户业务的QoE视为服务质量主观看法与特定网络指标之间的映射;最后,以最大化系统的用户QoS需求和吞吐量为目标,实现网络切片资源分配方案。仿真表明,所提方法能够有效提升系统的吞吐量和公平性,可以为5G无线资源管理提供参考。
A 5G wireless resource management technology based on reinforcement learning is proposed to address the issue of inflexibility in current network slicing resource allocation schemes.This method aims to achieve adaptive network slicing dynamic optimization and end-to-end service reliability.Firstly,calculate different user priorities to realize the allocation strategy based on user priority differentiation.Then,consider the QoE of user service as a mapping between subjective views on service quality and specific network indicators.Finally,with the goal of maximizing the user QoS requirements and system throughput,a network slicing resource allocation scheme is implemented.Simulation shows that the proposed method effectively improves the throughput and fairness of the system and provides a reference for 5G wireless resource management.
作者
张伟
ZHANG Wei(Guangdong Branch of China United Network Communications Co.,Ltd.,Guangzhou 510630,China)
出处
《移动通信》
2023年第12期66-70,共5页
Mobile Communications
关键词
切片资源分配
用户分组
QOE
强化学习
slice resource allocation
user grouping
QoE
reinforcement learning