复杂时序云异质服务资源的问责机制研究

The opacity of cloud service resource security QoS has become a bottleneck for the rapid development of cloud services, The accountability about safety and quality of cloud service is an important measure to protect the rights and interests of customers， and also promote the quality of cloud services. The accountability mechanism of cloud service is an important content of cloud service security. This project focuses on two scientific problems, which are the spatial structure description of heterogeneous service resources for cloud services, verifying and tracing SLA compliance problem based on cloud service network. Based on the complex network theory and time automata model, this project presents the basic model of cloud service behavior model, security SLA model and cloud service network, in order to construct complex heterogeneous cloud service resource network model, and establish dynamic timing generation cloud service Network model. From the above theoretical model, this project gives the cloud service compliance decision technology based on secure SLA, breaks the difficulty of traceability of multi-source node in cloud service network, and finally forms an accountability mechanism framework involves "modeling - judgment - traceability" of the heterogeneous cloud service network,improve the ability to determine the security responsibility of heterogeneous cloud service resources and to trace the source of the responsibility node, and increase the security transparency of cloud service resources.The research and development of our project is objective to provide powerful theory and technical support in the field of cloud safety problem, and also provide effective support for our future research on reliable services and service optimization of cloud computing.

云服务资源安全服务质量的不透明性已成为阻碍云服务快速发展的瓶颈性问题，保障客户权益、增强云服务资源透明性等问题的解决亟需研究和构建一种有效和可靠的云服务安全质量问责机制。本项目针对面向云服务的异质服务资源空间结构描述和基于云服务网络的SLA合规性问题判定与归属两个科学问题展开。基于复杂网络理论、时间自动机模型等的云服务网络空间结构研究，构建复杂异质云服务资源网络模型以及研究建立面向动态云服务网络时序生成模型；并基于上述模型，研究基于安全SLA的云服务合规性判定技术和云服务网多源责任节点溯源技术，形成一个集“建网-判定-溯源”为一体的复杂时序云服务网络异质服务资源的问责机制框架，提高云服务异质服务资源的安全责任判定能力及责任节点溯源能力，增加云服务资源的安全透明性，为建立自主有效的面向云服务安全领域的问责机制提供有力的理论和技术支持，能为我国未来云计算的可靠服务及服务优化提供有效的支撑。

本项目始终围绕研究内容和目标，按计划对复杂云服务问责当前存在的异质服务资源空间结构描述、云服务的合规性判定及其责任溯源和机制开放框架等挑战性问题展开科学探索，执行期间未做出调整。.项目取得以下研究成果，覆盖了计划的各项内容。.1）发现云服务运行状态的失效变点具有小波形态，揭示了云服务失效的资源欠缺本质及其与运行状态的相关性和因果性，提出云服务事件网络的时变结构生成技术，为后续的失责判定与溯源提供有力支撑。.2）针对SLA约束指标引发灰色责任区的现象，提出“未尽责”概念，并据此新认识构造了基于SLA合规性的责任判定，以及基于云服务事件网络的多源责任溯源等的新模型和新技术，以更细粒度的方式判断云服务失责或未尽责，并追溯责任归属。.3）提出一个集“构网-判定-溯源”为一体的开放问责框架，它将上述新认识和新技术构建出一个有效可靠的云服务问责新机制，其可交付更细粒度度量力和更强质证效力的多源失责链。.4）此外，提出基于变点小波云贮存服务失效预测新模型及其技术，为客商各方规避失效风险提供有力支撑。.上述成果可为国内云计算基础设施建立开放问责机制提供纯国内知识产权的全套问责技术和相关解释理论，为责任纠纷的仲裁给出有质证效力的责任解释，使云服务问责过程更具公正性，从而增强云服务资源透明性、支撑并推动我国云计算的可靠服务及服务优化。.论文成果方面，在国内外重要学术期刊和学术会议上，发表标注基金资助的学术论文共计40篇（国际刊物36篇，中文核心论文4篇）。其中，SCI检索论文17篇，EI检索论文7篇；获国家授权发明专利3项，申请国家发明专利1项，获计算机软件著作2项。.在人才培养方面，项目共培养研究生27名（7名博士生和20名硕士生）。超额完成了预期目标。