联系我们

论文查询结果	返回搜索

论文编号：	8399
作者编号：	2120142413
上传时间：	2016/6/9 14:59:20
中文题目：	基于QoS历史的云服务选择问题研究
英文题目：	The Research on Cloud Service Selection Based on QoS History
指导老师：	严建援
中文关键字：	QoS；云服务选择；逻辑衰减函数；熵权法
英文关键字：	QoS; Cloud Service Selection; logistic decay function; entropy method
中文摘要：	近年来，云服务作为一种新兴的IT服务模式，逐渐被众多的企业用户所接受。放眼整个IT市场，云服务依然是一个高增长的领域。随着我国综合实力的提升和市场经济发展进入新常态，加之云服务供应链上下游的共同努力，云服务应用也逐渐发展起来。越来越多的企业，尤其是中小企业用户，更加乐意选择把信息系统等建立在云端。但是，当前国内云服务市场仍然处于发展初期，云服务提供商鱼龙混杂；与此同时，云服务中断事件时有发生，并且造成了严重的后果。所以，为了充分利用云服务，用户应根据自身应用的需求和特点，选择并匹配低成本、高服务质量的云服务。 云服务选择问题自然而然地引起了学术界和商业界的重视。许多IT从业者结合自己经验提出了云服务选择的若干标准和注意事项，但相对分散，偏向概念化，难成一家之言。另一方面，研究人员通过各种各样的学术研究，基于QoS（Quality of Service，服务质量）提出了云服务选择的各种方法。但是，当前一些云服务选择方法往往只考虑实时的QoS表现或者某一时间段内平均的QoS表现，这些方法可能会推荐到一种特定的服务，但并不一定是最合适的，前者会忽略QoS的历史表现，后者则忽视了QoS的阶段性差异。 所以本文在文献研究和理论模型的基础之上，结合云服务选择实践，提出云服务选择的框架，并且采用MCDM（多准则决策）的方法，综合考虑QoS的历史表现情况及其变化和价格因素，利用逻辑衰减函数等，建立了云服务选择的过程模型，提出了一套较为完整的操作方案，并利用实际数据进行实证检验，对仿真结果进行展开讨论。我们发现仅考虑QoS平均值使我们的最终决策偏离最优决策，考虑QoS历史很有必要，同时，逻辑衰减函数和熵权法的采用虽然在本实验中未能影响最终的结果，但使我们的结果更加明显。最后我们进行总结和相关研究展望。
英文摘要：	Cloud services, as a new mode of IT services, is clearly a high-growth area in the IT market. Along with our country economic development into the new normal and industry value chain, more mature cloud growth is expected to come and more and more companies will choose to put the information technology system in the cloud. To take maximum advantage of the full potential of cloud computing, a key issue for cloud services users is to ensure that the specific requirements and characteristics of their applications can be met by cloud service providers. But the current cloud services market is at the early stage of development, and highly mixed. Cloud outage events have occurred, and cause serious consequences. Thus, cloud service selection caused the attention of the academia and the business community. Existing approaches in the literature that assist service users in the decision making process of selecting a cloud service provider only consider the real-time QoS performance or average historical QoS performance of services. Such mechanisms may recommend a particular service, but that service may not be the most appropriate. The former approach (considering the real-time QoS performance) may lead to the selection of a service at local maxima because it ignores past QoS performance, while the latter method (considering the average historical QoS performance) does not capture the frequent variation in the QoS performance of cloud services. In this paper, on the basis of the literature and theory study, combining with the practice, we put forward the framework of cloud service selection, and adopt MCDM (multi-criteria decision making) method, considering the history of QoS performance situation and its change and price factors, using the method of attenuation function and logic of entropy. Finally, we summarize the prospect of research and related.
查看全文：	预览  下载(下载需要进行登录)