基于改进粒子群算法的边云协同计算卸载策略

Edge-cloud Collaborative Computation Offloading Strategy Based on Particle Swarm Optimization Algorithm

移动边缘计算是一种可以为用户提供比云计算更低延迟和更低能耗的智能服务,但边缘服务器的资源是有限的,并不能完全替代云计算。为了充分利用云计算中海量网络资源的优势,研究了一个多用户、多边缘服务器和一个云数据中心的边缘-云协同场景。首先以最小化时延和能耗的加权和为目标,建立边云协同系统模型。然后在计算资源约束下,设计了一个非线性混合整数卸载决策问题,并采用基于适应度值和迭代次数的惯性权重,结合模拟退火算法改进粒子群算法求解计算卸载决策。仿真结果表明,改进的粒子群算法相比标准粒子群算法具有更好的收敛性,降低了8.7%的系统成本。并且在性能上优于其他3种基准对比算法,最多降低了72.3%的系统成本,能够有效降低系统网络中执行任务的时延和能耗。

Mobile edge computing is an intelligent service that can provide users with lower latency and lower energy consumption than cloud computing,but the resources of edge servers are limited and cannot completely replace cloud computing.In order to make full use of the advantages of massive network resources in cloud computing,an edge-cloud collaborative scenario with multiple users,multiple edge servers and a cloud data center is studied.Firstly,the edge-cloud collaborative system model is established with the goal of minimizing the weighted sum of delay and energy consumption.Then,under the constraint of computing resources,a nonlinear mixed integer offloading decision problem is designed,and the inertia weight based on fitness value and iteration number is used,and the simulated annealing algorithm is combined to improve the particle swarm optimization algorithm to solve the computation offloading decision.Finally,the simulation results show that the improved particle swarm optimization algorithm has better convergence than the standard particle swarm optimiza-tion algorithm,and reduces the system cost by 8.7%.Moreover,the performance is better than the other three benchmark comparison algorithms,and the system cost is reduced by 72.3%at most,which can effectively reduce the delay and energy consumption of executing tasks in the system network.