基于Web缓存和预取技术的性能优化研究

发布时间：2018-03-15 05:18

本文选题：Web缓存　切入点：缓存替换算法　出处：《江西理工大学》2014年硕士论文　论文类型：学位论文

【摘要】：随着Internet网络的迅速发展，Internet应用已遍及工业、军事、医疗、教育等领域，给人们的生活带来了极大的便利。与此同时，网络用户接入数量急剧膨胀，而网络带宽等硬件设施的升级非常有限，用户在浏览网络资源时具有较大的访问延迟。如何减少用户访问延迟、提高用户访问速度是Web领域面临的实际问题。Web缓存通过保存用户访问过的内容，，有效减少了用户获取网络资源等待时间，但其无法对用户即将访问的内容进行主动地缓存。预取技术弥补了Web缓存技术的不足，已成为优化缓存系统性能、降低用户访问延迟的重要方法。本文充分利用预取技术能预测用户的访问对象的优势，提出了一种基于预测模型的缓存替换算法。通过仿真实验表明了算法的优越性，在一定程度上提高了缓存系统的性能。本文的研究内容主要包含以下几点： 1、本文介绍了Web缓存系统结构的主要组成部分和预取技术的基本理论，对不同的缓存和预取方式分别进行了阐述，详细讲解了Web访问的四个相关特性，对缓存替换算法和预取算法的研究成果进行了研究和总结。 2、为了使日志中的数据正确、可用，介绍了Web日志挖掘的概念，以及日志数据预处理的四个主要步骤。简要介绍了Markov链的概念，对多Markov链模型进行了详细阐述。为了构造多Markov链模型，对类Markov链合并、准则函数和聚类相似度的相关定义进行了讨论。 3、在分析了影响缓存替换因素的基础上，构造了一个目标函数来计算缓存中对象的键值，用于衡量对象在缓存中的存储价值。利用预取技术能预测用户访问对象的优势，为所有用户构造了多Markov链模型，用于访问对象的预测。为了提高缓存的命中率，使缓存保留具有较大存储价值的对象，以及通过模型预测得到的访问对象，提出了基于多Markov链预测模型的缓存替换算法。使用日志驱动的方法对算法进行了仿真实验，并使用不同的性能评价指标将该算法与多种算法进行了比较。实验结果表明，提出的算法优于其它算法，更能减少用户的访问延迟，进而优化缓存的性能。
[Abstract]:With the rapid development of Internet network, Internet application has been widely used in the fields of industry, military, medical treatment, education and so on, which brings great convenience to people's life. At the same time, the number of Internet users has expanded rapidly. However, the upgrade of network bandwidth and other hardware facilities is very limited, and users have greater access latency when browsing network resources. Improving user access speed is a practical problem in the field of Web. The web cache can effectively reduce the waiting time for users to obtain network resources by saving the contents that users have visited. But it can't cache the content that the user is about to access actively. Prefetching technology has made up the deficiency of Web caching technology, and has become the optimization of caching system performance. This paper takes full advantage of the prefetching technology to predict the users' access objects, and proposes a cache replacement algorithm based on the prediction model. The simulation results show the superiority of the algorithm. To a certain extent, the performance of cache system is improved. The research content of this paper mainly includes the following points:. 1. This paper introduces the main components of Web cache system structure and the basic theory of prefetching technology, expounds the different caching and prefetching methods, and explains in detail the four related characteristics of Web access. The research results of cache replacement algorithm and pre-fetching algorithm are studied and summarized. 2. In order to make the data in log correct and usable, the concept of Web log mining and four main steps of log data preprocessing are introduced. The concept of Markov chain is briefly introduced. In order to construct the multiple Markov chain model, the definition of Markov like chain merging, criterion function and clustering similarity are discussed in detail. 3. Based on the analysis of the factors affecting cache substitution, an objective function is constructed to calculate the key value of the object in the cache, which is used to evaluate the storage value of the object in the cache. A multiple Markov chain model is constructed for all users to predict access objects. In order to improve the hit rate of cache, keep objects with large storage value, and access objects predicted by the model, A cache replacement algorithm based on multiple Markov chain prediction model is proposed. The algorithm is simulated by log driven method, and the algorithm is compared with many algorithms by using different performance evaluation indexes. The proposed algorithm is superior to other algorithms, which can reduce the user's access delay and optimize the performance of cache.
【学位授予单位】：江西理工大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TP333;O211.62

【参考文献】