针对移动边缘计算中用户移动性导致服务器间负载分布不均,用户服务质量(Quality of Service,QoS)下降的问题,提出了一种移动性感知下的分布式任务迁移方案。首先,以优化网络中性能最差的用户QoS为目标,建立了一个长期极大极小化公平性问...针对移动边缘计算中用户移动性导致服务器间负载分布不均,用户服务质量(Quality of Service,QoS)下降的问题,提出了一种移动性感知下的分布式任务迁移方案。首先,以优化网络中性能最差的用户QoS为目标,建立了一个长期极大极小化公平性问题(Max Min Fairness,MMF),利用李雅普诺夫(Lyapunov)优化将原问题转化解耦。然后,将其建模为去中心化部分可观测马尔可夫决策过程(Decentralized Partially Observable Markov Decision Process,Dec-POMDP),提出一种基于多智能体柔性演员-评论家(Soft Actor-Critic,SAC)的分布式任务迁移算法,将奖励函数解耦为节点奖励和用户个体奖励,分别基于节点负载均衡度和用户QoS施加奖励。仿真结果表明,相比于现有任务迁移方案,所提算法能够在保证用户QoS的前提下降低任务迁移率,保证系统负载均衡。展开更多
Service composition is considered to be an effective way for resource sharing and node collaboration in Mobile Ad hoc NETworks (MANETs) environment.In order to improve the reliability and shorten the response time of ...Service composition is considered to be an effective way for resource sharing and node collaboration in Mobile Ad hoc NETworks (MANETs) environment.In order to improve the reliability and shorten the response time of composite services,this paper first analyzes the node environments and network environments that affect the component services' availability,and then proposes an Environment-aware Quantitative Evaluation Model for Service Availability (EQEM-SA).In addition,based on EQEM-SA,a service field concept is proposed as well as the availability attenuation equation,where the availability value is considered as the field strength.Then the service selection approach based on service field model is presented.The effectiveness of the proposed approach is illustrated and compared with related references,and the results of experimental evaluations indicate that our approach significantly reduces the failure rate and shortens the service delay.展开更多
基金supported in part by the National Key Basic Research Program of China(Grant Nos.2009CB320406and2009CB320504)Foundation for Innovative Research Groups of the National Natural Science Foundation of China(Grant No.60821001)National Natural Science Foundation of China(Grant No.61070206)
文摘Service composition is considered to be an effective way for resource sharing and node collaboration in Mobile Ad hoc NETworks (MANETs) environment.In order to improve the reliability and shorten the response time of composite services,this paper first analyzes the node environments and network environments that affect the component services' availability,and then proposes an Environment-aware Quantitative Evaluation Model for Service Availability (EQEM-SA).In addition,based on EQEM-SA,a service field concept is proposed as well as the availability attenuation equation,where the availability value is considered as the field strength.Then the service selection approach based on service field model is presented.The effectiveness of the proposed approach is illustrated and compared with related references,and the results of experimental evaluations indicate that our approach significantly reduces the failure rate and shortens the service delay.
