基于多变量联合优化的MEC网络能耗最小化算法

Energy Minimization of Mobile Edge Computation Networks Based on Multi-variable

基于无人机(unmanned aerial vehicle,UAV)支持的移动边缘计算网络能够为移动终端提供按需计算服务,但是其能耗仍是亟待解决的问题。为此,提出基于UAV的移动边缘计算的最小化能耗算法(energy consumption minimization algorithm,ECMA),进而减少网络的总体能耗。先构建关于时隙分配、任务分配和UAV移动轨迹3个变量的目标问题;再将目标问题拆分成两个子问题:时隙分配和任务分配的联合子问题和UAV移动轨迹的优化问题,并分别利用迭代算法和连续凸近似法求解这两个子问题。仿真结果表明,相比于正交多址接入算法(orthogonal multiple access algorithm,OMAA)和等功率分配算法(equal power allocation algorithm,EPAA)算法,提出的ECMA算法降低了网络能耗。

An unmanned aerial vehicle(UAV)-assisted mobile edge computing(MEC) networks fulfil the on-demand computation services for the mobile terminals(MTs).But the energy efficiency is still a major issue in this network.Therefore,for UAV-based mobile edge computation networks,energy consumption minimization algorithm(ECMA) is proposed in this paper.First,a target problem with three variables of time slot allocation,task allocation and UAV movement trajectory is constructed.Then,the target problem is divided into two subproblems:joint time slot allocation and task computation capacity,and UAV trajectory optimization.To solve the joint time allocation and computation task capacity,we proposed an iterative algorithm.Finally,to optimize the trajectory of the UAV,we used the successive convex approximation technique.Simulation results show that the proposed ECMA algorithm can effectively reduce energy consumption of networks,compared with orthogonal multiple access algorithm(OMAA) and equal power allocation algorithm(EPAA).