基于半休眠模式的云计算任务卸载策略及性能研究  

Research on semi-sleep mode based task offloading strategy and system performance in cloud computing

作　　者：宋家 余靖[1] 金顺福[1] Song Jia;Yu Jing;Jin Shunfu(School of Information Science and Engineering,Yanshan University,Qinhuangdao 066004)

机构地区：[1]燕山大学信息科学与工程学院,秦皇岛066004

出　　处：《高技术通讯》2021年第12期1269-1279,共11页Chinese High Technology Letters

基　　金：国家自然科学基金(61872311,61973261)资助项目。

摘　　要：云系统中的任务卸载是一种将密集的移动应用程序迁移到云端计算的机制。考虑云用户服务质量和云系统能耗水平,在云端引入周期性同步半休眠模式,提出一种云计算任务卸载策略。针对不同负载情况下的虚拟机服务速率,构建多服务台同步多重工作休假排队模型。基于拟生灭过程和矩阵几何解方法,得到云用户的平均响应时间与云系统的能源消耗。进行系统实验,评估卸载策略的系统性能,并验证其有效性。基于权重分配构建系统成本函数,改进海鸥优化算法,给出云计算任务卸载策略的智能优化方案,实现系统成本的最小化。Task offloading in cloud computing is a mechanism for migrating intensive mobile applications to cloud environment.By considering the service quality of cloud users and the energy consumption level of cloud system,a periodic synchronous semi-sleep mode is introduced in the cloud and a task offloading strategy is proposed.According to the service rates of virtual machine under different loads,a multi-server queueing model with synchronous multiple working vacation is constructed.Using the quasi-birth-and-death process and matrix-geometric solution method,the average response time of cloud users and the energy consumption of cloud system are derived.System experiments are carried out to evaluate the system performance and verify the validity of the offloading strategy.With the weight factors of different performance measures,a system cost function is constructed.By improving a seagull optimization algorithm,the proposed task offloading strategy in cloud computing is optimized with the minimum system cost.

关 键 词：云计算 任务卸载 半休眠 工作休假排队 智能优化 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]