基于网络模体的空闲计算资源捕获算法

Idle computing resource capture algorithm based on network motifs

针对新兴应用的低时延需求以及移动边缘计算网络中移动设备趋于密集化,同时闲置计算资源丰富,使得系统计算资源利用率低下,考虑基于设备到设备通信的计算卸载有望满足新兴应用低延迟要求。另外,移动设备将计算密集型应用通过设备到设备通信卸载到空闲设备,能够提高计算资源利用率,缓解移动边缘计算网络计算压力。考虑移动边缘计算网络,利用网络模体对系统进行分析,捕获系统空闲计算资源,实现设备到设备卸载。具体来说,考虑基于网络基本组成结构、卸载数据流向的空闲计算资源发掘问题,结合链路传输质量及中断概率,利用KM(Kuhn-Munkras)算法,提出了一种基于网络模体的设备匹配资源搜索算法,以捕获终端邻近空闲计算资源,实现设备到设备卸载。仿真结果表明,该算法能有效捕获空闲计算资源,提高系统计算资源利用率。

Aiming at the low latency requirements of emerging applications and Mobile Edge Computing(MEC) network;In MEC, mobile devices tend to be intensive and idle computing resources are abundant, which leads to low utilization of system computing resources;At the same time, mobile devices offloading computing intensive applications to idle devices through D2D communication, which can improve the utilization of computing resources and relieve the computing pressure of MEC network. Considering MEC network, the network model is used to analyze the system, capture the idle computing resources of the system, and realize D2D unloading. Specifically, considering the problem of idle computing resource discovery based on the basic network structure and the direction of unloading data, combining with the link transmission quality and outage probability, using KM(Kuhn-Munkras)algorithm, a device matching resource search algorithm based on network model is designed and proposed to capture the idle computing resources near the terminal and realize D2D offloading. Finally, the simulation results show that the proposed algorithm can capture idle computing resources in MEC network and improve the utilization of system computing resources.