基于半马尔科夫决策过程的虚拟传感网络资源分配策略

Semi-Markov Decision Process-based Resource Allocation Strategy for Virtual Sensor Network

针对传统无线传感网络(WSN)中资源部署与特定任务的耦合关系密切,造成较低的资源利用率,进而给资源提供者带来较低的收益问题,根据虚拟传感网络请求(VSNR)的动态变化情况,该文提出虚拟传感网络(VSN)中基于半马尔科夫决策过程(SMDP)的资源分配策略。定义VSN的状态集、行为集、状态转移概率,考虑传感网能量受限以及完成VSNR的时间,给出奖赏函数的表达式,并使用免模型强化学习算法求解特定状态下的行为,从而最大化网络资源提供者的长期收益。数值结果表明,该文的资源分配策略能有效提高传感网资源提供者的收益。

The close relationship between resource deployment and specific tasks in traditional Wireless Sensor Network(WSN)leads to low resource utilization and revenue.According to the dynamic changes of Virtual Sensor Network Request(VSNR),the resource allocation strategy based on Semi-Markov Decision Process(SMDP)is proposed in Virtual Sensor Network(VSN).Then,difining the state,action,and transition probability of the VSN,the expected reward is given by considering the energy and time to complete the VSNR,and the model-free reinforcement learning approach is used to maximize the long-term reward of the network resource provider.The numerical results show that the resource allocation strategy of this paper can effectively improve the revenue of the sensor network resource providers.