摘要
在边缘计算资源协同与调度QoE优化前提下,对两阶段边缘服务组合及调度提出一种改进的天牛须粒子群算法.该算法将天牛须搜索算法中的天牛抽象成粒子,将单个个体的天牛须搜索算法拓展至群体,并引入二阶振荡机制和动态因子.不仅改进了位置更新公式和动态参数机制,改进群体觅食时的位置更新时的动态参数机制,丰富了群体移动时的位置多样性,并且提高了算法的全局搜索能力.通过QoE限制条件下的服务组合与调度仿真实验结果分析得出,该算法能够在满足用户请求QoE的条件下使得请求的整体执行时间开销达到最小.
Under the premise of QoE optimization,a new Beetle Antennae Particle Swarm Optimization is proposed for two-stage edge service composition and scheduling.The algorithm abstracts the concept of"beetle"in Beetle Antennae Search Algorithm into"particle",extends the individual in the Beetle Antennae Search Algorithm to the swarm,and introduces the second-order oscillation mechanism and dynamic factor.Thus,the position update formula and dynamic parameter mechanism of the swarm when foraging are improved,the position diversity of the swarm is enriched,and the global search ability of the algorithm is also improved.Simulation experiments of service composition and scheduling are carried out under the condition of QoE restriction.The analysis results show that the algorithm can meet the QoE constraints in user request and minimize the overall execution time of the request.
作者
简琤峰
裘科意
张美玉
JIAN Cheng-feng;QIU Ke-yi;ZHANG Mei-yu(Computer Science and Technology College,Zhejiang University of Technology,Hangzhou 310023,China)
出处
《小型微型计算机系统》
CSCD
北大核心
2019年第7期1397-1403,共7页
Journal of Chinese Computer Systems
基金
国家自然基金面上项目(61672461,61672463)资助
关键词
边缘计算
QOE
服务组合与调度
天牛须粒子群算法
edge computing
QoE
service composition and scheduling
beetle antennae particle swarm optimization
