物联网边缘计算资源分配模糊建模与优化研究

Research on Fuzzy Modeling and Optimization of Resource Allocation of Edge-computingbased Internet of Things

针对物联网边缘计算资源分配问题,提出模糊多目标优化资源分配算法.先对物联网边缘计算架构进行设计,提出一种含有虚拟化控制器的边缘网关型物联网边缘计算系统架构.然后,利用模糊优化理论,建立了物联网边缘计算系统资源分配模糊多目标函数模型,利用非线性六角模糊数对目标函数进行求解,提出了多目标向单目标转换算法和一种基于改进遗传算法的模因算法对模糊目标函数进行优化.最后,对所提出的物联网边缘计算系统资源分配的模糊优化方法进行实验和仿真,并与禁忌搜索元启发式算法、捕食搜索算法和多目标遗传算法进行对比仿真,本文算法在系统时延、边缘设备平均能耗、满意度等方面,都得到比较优异的资源分配结果.

According to the issue of the edge computing in Internet of Things(IoT)resource allocation,it is proposed a fuzzy multi-objective optimization resource allocation algorithm in this paper.At first,the IoT edge computing architecture is designed and an edge gateway type IoT edge computing system architecture with a virtualization controller is proposed.Then,a fuzzy multi-objective function model for resource allocation of edge computing systems in IoT using fuzzy optimization is built.By solving the objective function using NHFN(Nonlinear Hexagonal Fuzzy Number),a multi-objective to single-objective conversion algorithm is proposed.And a memetic algorithm based on improved genetic algorithm to optimize the fuzzy objective function is proposed.Finally,experiments and simulation on the proposed fuzzy optimization method for resource allocation of IoT edge computing system are made,meanwhile,the simulation on tabu search meta-heuristic algorithm,predatory search algorithm and Multi-Objective Genetic Algorithm are compared.The algorithms proposed in this paper have obtained excellent results of resource allocation in terms of system delay,average energy consumption of edge devices,and satisfaction.