云计算任务调度策略分析以及语义应用构建研究

发布时间：2018-05-14 04:37

本文选题：云计算 + 任务调度　；参考：《华东师范大学》2014年硕士论文

【摘要】：云计算出现被誉为IT产业第四次革命,使计算机相关的产业模式、程序应用环境及用户服务模式等都产生了全新的变化,由于云计算分布式系统本身具有优良的计算性能和强大的可扩展性以及对于使用云计算用户具有透明性,使得云计算作为一种先进的计算机资源组织形式,已经在实际的生产和生活中发挥着越来越重要的作用,而云计算作为一种分布式的计算机组织形式,其效率以及性能与计算硬件平台的架构以及其任务调度算法以及在云平台上的应用构建方式有着密切的关系,如何提高云计算环境下任务调度算法的性能以及优化其上的应用架构方式,从而提高云计算系统整体的性能成为了一个值得研究的问题。本文中分析了云计算环境下任务调度的特性,比较了经典的分布式任务调度算法在云计算环境下的优缺点,并在此基础上结合博弈论的相关理论,引入兼顾了系统最优效率以及每一个任务最优执行效率的纳什平衡和推导以及冲突任务组的概念。说明了纳什平衡代价与全局最优值的比值,即无政府代价,是一个可以接受的常数前提下,提出了一种适合云计算环境基于非合作博弈论的任务调度算法,通过理论分析以及数学模型仿真验证了在云计算环境下提出的基于非合作博弈论的任务调度算法的可行性,并且比较了本文中的任务调度算法与经典的任务调度算法在云计算环境下的优缺点。在分析影响云计算应用性能因素基础上,并结合云计算分布式环境不同的实际应用的固有特点,对于云计算分布式环境下和经典集中式环境下用户应用构建方案及不同的集中式文本应用构建方法进行对比,分析了影响云计算环境下文本应用的构建的因素。通过理论分析这些文本应用构建方法在云计算应用环境下的优缺点,从而提出了一种云计算环境下文本语义提取算法,进而在此文本语义提取算法的基础上构建了相应文本应用,包括文本分类应用和课件知识点提取应用。理论和实验证明提出了文本语义提取算法证明可以在保证提取语义质量的前提下,减少云计算对于网络传输的开销,有效提高了在云计算环境下对于文本语义提取效率,实验证明了本文中提出的云计算环境下的语义提取算法可行性以及理论推导的正确性。
[Abstract]:Cloud computing has been praised as the fourth revolution in IT industry, which has brought new changes to computer related industry model, program application environment and user service mode, etc. Because cloud computing distributed system has excellent computing performance, powerful expansibility and transparency to users of cloud computing, cloud computing is an advanced organization form of computer resources. Cloud computing has been playing a more and more important role in actual production and life, and cloud computing as a distributed computer organization, Its efficiency and performance are closely related to the architecture of the computing hardware platform, its task scheduling algorithm and the application construction method on the cloud platform. How to improve the performance of task scheduling algorithm and optimize the application architecture in cloud computing environment, so as to improve the overall performance of cloud computing system has become a problem worth studying. This paper analyzes the characteristics of task scheduling in cloud computing environment, compares the advantages and disadvantages of classical distributed task scheduling algorithms in cloud computing environment, and combines the relevant theory of game theory. This paper introduces the concepts of Nash equilibrium and derivation and conflict task group which take into account the optimal efficiency of the system and the optimal execution efficiency of each task. It is shown that the ratio of Nash equilibrium cost to global optimal value, that is, anarchy cost, is an acceptable constant, and a task scheduling algorithm based on non-cooperative game theory is proposed for cloud computing environment. The feasibility of the task scheduling algorithm based on non-cooperative game theory in cloud computing environment is verified by theoretical analysis and mathematical model simulation. And compare the advantages and disadvantages of the task scheduling algorithm and the classical task scheduling algorithm in cloud computing environment. On the basis of analyzing the factors that affect the performance of cloud computing applications, and combining with the inherent characteristics of different practical applications in cloud computing distributed environment, In this paper, the author compares the user application construction schemes and different centralized text application construction methods between the cloud computing distributed environment and the classical centralized environment, and analyzes the factors that affect the construction of the text application in the cloud computing environment. Through theoretical analysis of the advantages and disadvantages of these text application construction methods in cloud computing application environment, a text semantic extraction algorithm in cloud computing environment is proposed. On the basis of this text semantic extraction algorithm, the corresponding text applications are constructed, including text classification application and courseware knowledge point extraction application. Theoretical and experimental results show that text semantic extraction algorithm can reduce the overhead of cloud computing for network transmission and improve the efficiency of text semantic extraction in cloud computing environment. The experiment proves the feasibility of the semantic extraction algorithm and the correctness of the theoretical derivation in the cloud computing environment.
【学位授予单位】：华东师范大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP393.01;TP391.1

【参考文献】