摘要
针对现有云环境市场调度策略多为追求收益最大化而对服务方信誉度考虑不足,提出基于信用价格及成本计算的市场调度模型,使服务方收益最大化且不失在客户中的信誉程度。模型引入信用评价系数、信用价格以及任务赔偿金和收益判别函数等经济学概念,将服务方在用户中的信用评价融入基于成本计算的任务抢占调度机制,以服务方收益判别函数值进行任务抢占决策判断。模拟实验表明,模型实现服务方收益最大化且优先调度低信誉度任务。
For the existing market schedule strategy in a cloud environment seeking to maximize their rate of return and insufficient attention to credibility of service, a market schedule model in a cloud environment based on credit price and cost computing is pro-posed, with maximizing their rate of return and good credit for service. Credit price, task penalty and income discriminant function are introduced to model. Credibility of service is introduced in preempted schedule strategy based on cost calculating, and the income discriminant function of service is thought as decision to preempt the current task. The simulation shows that, income maximization is implemented and a low-trust task is preempted to schedule.
出处
《微型电脑应用》
2013年第9期9-11,共3页
Microcomputer Applications
基金
国家自然科学青年科学基金项目(61005008)
江苏现代教育技术研究基金技术应用重点项目(2012-R-23010)
关键词
云计算
调度算法
信用价格
任务赔偿金
收益判别函数
Cloud Computing
Scheduling Algorithm
Credit Price
Task Penalty
Income Discriminant Function
