基于Monte Carlo熵权决策的云计算数据预取算法  

Cloud Computing Data Prefetching Algorithm Based on Monte Carlo Entropy Weight Decision-making

作　　者：王辉[1] 高广耀[1] 

机构地区：[1]河南广播电视大学现代教育技术中心,郑州450008

出　　处：《科技通报》2015年第4期130-132,共3页Bulletin of Science and Technology

摘　　要：云计算数据预取算法设计是实现云平台环境下通信链路优化和任务调度均衡分配的基础技术。在传统的云计算据查询模式下,当由于缓存空间不足而导致新的缓存数据无法进入缓存时,导致数据预取拥堵,性能不好。提出一种基于Monte Carlo熵权决策的云计算数据预取算法,构建云计算数据查询模板模型,进行Hybrid缓存置换数据预取前置处理,采用Monte Carlo熵权决策方法,把云计算预取信号从缓存域变换到波束域,构建置换函数,实现了对算法的改进。仿真实验研究得出,该算法通过熵权特征提取,进行云计算数据预取决策,提高了云计算数据预取性能,大数据访问延迟率降低,云计算数据存取和调度效率提高,保真率较好。Cloud computing data prefetching algorithm design is the basis of technology to realize the cloud platform communication link optimization and task scheduling equilibrium. Traditionally, the user of each query request must be established a database connection, lead to prefetch data congestion, performance is not good. A cloud computing data prefetching algorithm is proposed based on Monte Carlo entropy weight decision-making method. The cloud computing model of data query is constructed, Hybrid cache replacement prefetch data pre-processing is taken, Monte Carlo entropy weight decision-making method is used, the cloud computing prefetching signal is converted from the cache domain to the beam domain, the replacement function is constructed, the improvement of the algorithm is realized. Simulation results show that the algorithm improves the cloud data prefetching performance, delay of data access rate is reduced, data access and scheduling efficiency are improved, and it has better fidelity rate.

关 键 词：熵权决策 云计算 链路 数据 

分 类 号：TP391.3[自动化与计算机技术—计算机应用技术]