基于MDP的协作认知边缘计算网络资源分配方案

MDP based resource allocation scheme for cooperative cognitive mobile edge computing networks

为了缓解移动边缘计算网络中用户设备续航有限、频谱稀缺的问题,提出一种可无线充能的认知边缘计算网络中的资源优化方案。主用户可以对次用户进行无线充能与协作中继,尽快完成自身数据传输后使信道空闲,次用户接入信道后进行边缘计算。利用马尔科夫决策过程(Markov Decision Proces,MDP)对次用户的能量收集时间长度、卸载能耗和操作模式等进行联合优化设计,最大化次用户能获得的长期期望计算量。仿真结果表明,所提方案能够提升系统频谱效率,并且所提方案获得的长期期望计算量就短期优化方案而言具有显著提升。

To mitigate the problem of limited battery life and scarce spectrum for user devices in mobile edge computing networks,a resource optimization scheme in the wireless powered cognitive edge computing network is proposed.The primary user can charge the secondary user wirelessly,and execute cooperative relay with the secondary user to complete the data transmission as soon as possible,and to leave the primary channel idle.Then the secondary user can access the primary channel to execute edge computing.Markov decision process(MDP)is adopted to optimize the length of the energy harvesting time,offloading energy consumption and operation mode of the secondary user,so as to maximize the long-term expected calculated amount of the secondary user.Numerical results show the proposed scheme can improve the spectrum efficiency.Compared with the short-term optimization scheme,it can significantly improve the long-term expectation calculation amount.