云计算环境下基于虚拟数据中心的资源优化分配技术研究

Cloud computing platform provides infrastructure resource to various cloud applications through data center networks, in which how to obtain the maximum resource utilization under the premise of satisfying the quality of services (QoS) for the applications has become one of the core technical challenge in the related international research areas. However, due to its lack of effective management of the network resources of the main resource allocation approaches, the problem of the QoS of the network intensive applications cannot be guaranteed arises. Thus, this project tries to solve this problem based on virtual data center technology in data center networks. Pointing to the dynamic resource allocation with the users’ requests arrive and leave dynamically, the various topologies, the balance of the different resource, the efficiency and the dynamic problem will be studied. Four kinds of key approaches will be proposed for solving the above problems, which involves the topology awareness of the virtual data center based resource allocation approach, multidimensional resource node ranking based resource allocation approach, online resource allocation approach based on approximation algorithm, and dynamic resource allocation approach based on minimizing the cost. This project targets to construct a set of theories and mechanism for the resource allocation in data center networks, and promote the research level as well as the independent innovational ability of our country in the related areas.

云计算通过数据中心网络为各类云应用提供基础设施资源。如何在满足应用服务质量的前提下，实现数据中心网络下资源的最大化利用已成为当前国际相关领域面临的核心技术挑战。针对现有的资源分配方式缺乏对网络资源的有效管理，导致对网络资源敏感应用的服务质量难以保证的问题，本项目提出了云计算环境下基于虚拟数据中心的资源分配模型。针对请求动态到达与离开环境下的动态资源分配过程，将着重研究拓扑结构多样性、资源均衡性、高效性以及动态性的问题。具体研究包括基于虚拟数据中心拓扑结构感知的动态资源分配方法、基于多资源节点排名的动态资源分配方法、基于近似算法的动态资源分配方法和基于代价最小化的动态资源分配方法等四个方面。本项目旨在建立一套切实有效的数据中心网络下资源分配的理论机制，提升我国在相关领域的研究水平和自主创新能。

云计算通过数据中心网络为各类云应用提供基础设施资源。如何在满足应用服务质量的前提下，实现数据中心网络下资源的最大化利用已成为当前国际相关领域面临的核心技术挑战。针对现有的资源分配方式缺乏对网络资源的有效管理，导致对网络资源敏感应用的服务质量难以保证的问题，本项目提出了云计算环境下基于虚拟数据中心的资源分配模型。针对请求动态到达与离开环境下的动态资源分配过程，重点研究了拓扑结构多样性、资源均衡性、高效性以及动态性的问题。提出了一种基于虚拟数据中心拓扑结构感知的动态资源分配方法，一种基于多资源节点排名的动态资源分配方法，一种基于近似算法的动态资源分配方法，以及一种基于代价最小化的动态资源分配方法。本项目建立了一套切实有效的数据中心网络下资源分配的理论机制，有效地提升了我国在相关领域的研究水平和自主创新能。