摘要
针对边缘计算的大数据量和强实时性交互处理需求,研究面向边缘服务器的计算能力、服务时间、带宽、内存等多目标优化的任务资源协同调度算法.首先提出面向多目标优化的边缘计算资源协同调度模型,在此基础上提出了改进混沌蝙蝠群协同调度算法,通过二维扰动因子和混沌因子减少群组算法陷入局部最优的概率,同时提升找到最优调度解的概率.实验证明该算法能够在满足多目标优化的前提下边缘计算能力更突出,能够满足强实时性大数据处理需求.
Aiming at the large data volume and strong real-time interactive processing requirements of edge computing,this paper focuses on the multi-objective optimization task resource collaborative scheduling algorithm for edge server-oriented computing power,service time,bandwidth and memory. Firstly,a collaborative computing model for edge computing resources oriented to multi-objective optimization is proposed. Based on this,an improved chaotic bat group cooperative scheduling algorithm is proposed. The two-dimensional perturbation factor and chaotic factor are used to reduce the probability that the group algorithm falls into local optimum and improvethe probability of the optimal scheduling solution. Experiments show that the algorithm can make the edge time calculation ability of the edge cloud more prominent under the premise of multi-objective optimization,and load balancing shows that the algorithm can reduce energy consumption.
作者
简琤峰
陈家炜
张美玉
JIAN Cheng-feng;CHEN Jia-wei;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11期2424-2430,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
资源调度
二维扰动
混沌蝙蝠群
edge computing
resource scheduling
two dimensional
perturbation
chaotic bat algorithm
