在云环境下基于多轨道图灵机的服务部署研究

发布时间：2018-04-20 12:46

本文选题：云计算 + 云服务　；参考：《南京邮电大学》2013年硕士论文

【摘要】：云计算可以把企业提供的服务通过网络提供给用户。因为现有的云架构缺乏云服务组合和部署的中间件层，所以本文提出基于轨道图灵机的云服务部署模型。这个云服务部署模型使按需云服务部署更加方便和快速，并实现了一个基于云计算的中间件动态服务部署（CMCD）原型。云服务部署模型包括云服务搜索引擎模块、云服务组合模块和图灵机构造模块。云服务搜索引擎模块是一个支持匹配算法多标准的云服务的搜索引擎。它能够搜索满足以下3方面要求的云服务：（1）功能要求，（2）技术要求，（3）费用要求，为了进一步提高搜索结果的准确性，我们设计了一个云本体。云本体包含了一组云的概念，和这些概念的个体以及这些个体之间的关系。云本体是用于确定相似云服务的依据，基于云本体的相似云服务推理方法包括三种（1）概念相似度推理（2）对象属性相似的推理（3）数据类型的属性相似推理。在现实中，云本体论有效的解决了我们如何度量云服务提供商提供的云服务之间的关系即相似度的难题。在云服务的组合上，我们使用有限状态机（FSM）规定其调用序列，，并且基于一种改进的修剪算法来创建云服务组合树。在其云服务组合树中可以找出所有的可执行的路径，然后用一个简单的加权技术用于选择最佳的路径即所要部署的云服务。图灵机构造模块主要是从云服务提供商宏观层面进一步对云服务的部署进行控制。CMCD说明了动态云服务的部署可行性。CMCD中间件使用图灵机的数学模型，实现动态云服务部署。实验说明基于图灵机的云服务部署算法在普通云服务功能方面有较好的性能，并且部署成功率非常高。
[Abstract]:Cloud computing can provide services provided by enterprises to users through the network. Because the existing cloud architecture lacks the middleware layer of cloud service composition and deployment, this paper proposes a cloud service deployment model based on track Turing machine. This cloud service deployment model makes on-demand cloud service deployment more convenient and faster and implements a cloud-based middleware dynamic service deployment (CMCD) prototype. Cloud service deployment model includes cloud service search engine module, cloud service composition module and Turing machine construction module. The cloud service search engine module is a search engine that supports multi-standard matching algorithms for cloud services. In order to improve the accuracy of search results, we design a cloud ontology. Cloud Noumenon contains a set of cloud concepts, the individuals of these concepts and their relationships. Cloud ontology is the basis for determining similar cloud services. The reasoning method of similar cloud services based on cloud ontology includes three kinds of concept similarity reasoning (2) object attribute similarity reasoning / 3) attribute similarity reasoning. In reality, cloud ontology effectively solves the problem of how to measure the relationship between cloud services provided by cloud service providers, that is, similarity. In the composition of cloud services, we use finite state machine (FSM) to specify its calling sequence, and create a cloud service composition tree based on an improved pruning algorithm. All the executable paths can be found in its cloud service composition tree, and then a simple weighting technique is used to select the best path, that is, the cloud service to be deployed. The construction module of Turing machine mainly controls the deployment of cloud service from the macro level of cloud service provider. CMCD illustrates the feasibility of dynamic cloud service deployment. CMCD middleware uses the mathematical model of Turing machine to implement dynamic cloud service deployment. The experiments show that the cloud service deployment algorithm based on Turing machine has better performance in the common cloud service function, and the success rate of deployment is very high.
【学位授予单位】：南京邮电大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：TP393.09

【参考文献】