Trust Model And Multi-objective Selection In Cloud Service

Under the background of the rapid development globalization,enterprises are facing the increasingly pressure of external resources environment,and the lack of innovative products or resource consumption internally.The system framework of cloud computing and other new information technologies provide a good opportunity for enterprises to improve production efficiency,integration of resources and collaborative value innovation from a wider scope,build more effective low-carbon operation system and service mode.However,the new cloud service model in the integration of chain and the cooperation of value chain also brings new security problem risks.The remote service mode of virtualization and multi-agents collaborative service features continue to broaden the traditional boundary of information security,causing more security risks.Traditional information security technology cannot solve this problem very well.However,the trust based system management provides a new idea for it.Trust model can help enterprise users effectively identify the trust of cloud service resources.And it can make enterprise users select appropriate service resources to deploy their tasks in accordance with their own business process requirements,which can maintain the security and trust of services on the basis of low cost and high efficiency.It provides an important guarantee for enterprises to improve production efficiency and competitiveness in the emerging wave of information technology.This dissertation proposes a trust representation model for the single cloud service resource.Then it builds the trust network for multiple cloud services resources,to dynamically update and maintain the trust status of cloud service resources.On the basis of trust network,it designs the framework of business process driven trust-based cloud service system for cloud enterprise user.From the service demand side,it proposes the task combination model,which decomposes the business process into related tasks with considering the process structure for cloud service supporting.Finally,from the service supply side,it builds the multi-objective cloud service selection model and algorithm,to help enterprise users configure appropriate resources and select cloud services based on their objectives.The main contributions of this dissertation are as follows:(1)Trust representation method of cloud services.For the trust representation problem of the single cloud service recourse,the description of trust factors in cloud service resources are usually characterized by uncertainty and fuzziness,that is,the trust factor of cloud services is imprecise and difficult to quantify.Fuzzy theory is introduced to describe uncertain or fuzzy trust attributes in cloud service resources which are difficult to be directly quantified.At the same time,using the Bayesian Network integrates fuzzy trust factors,to build the fuzzy trust recognition method for cloud services resources.It can help the system recognize the nodes that have high reliability and trust from massive cloud service resources,to ensure the safety and credibility of cloud service.(2)Building the trust network of cloud service.For the dynamic updating and maintenance of trust values for multiple cloud service resources,a trust network of cloud service based on historical interaction information is proposed.The trust situation in the trust network is constantly changed by the feedback of interaction information between users and services.Based on the trust network,it designs a dynamic trust mechanism to update and maintain the trust value of the user node or the service node to ensure the real-time dynamic trust of resources in cloud service system,providing effective support to solve the security problems of cloud services.(3)Architecture of trusted cloud service system and task combination model.On the basis of cloud services resources trust network,for business users to use cloud services resources more efficiently,this dissertation designs and builds a trust-based cloud service system framework which faces enterprise business process.It guides the user to upgrade its information system from the original system to cloud-based one,through access to external public cloud services and integrating the internal private cloud services.From the service demand side,defining the enterprise business process as the main carrier of service request,enterprise users should consider the specific structure of their business processes,and then select appropriate cloud services to support the completion of tasks in its business process.Therefore,this dissertation establishes the business process driven task combination model.It proposes 6 structures: sequence,parallel,parallel-AND,parallel-OR,parallel-XOR,simple loop and composite loop as the typical structure of task combination model.And it defines the execution time,cost and trust as optimization objectives,to improve efficiency and reduce cost on the basis of keeping security and reliability.(4)Multi-objective service selection model and algorithm.For business users selecting appropriate cloud services to support their business process,this dissertation proposes a multi-objective service selection model under the architecture of trust-based cloud service system from the perspective of cloud service supply side.Trust mechanism is introduced in the system to describe the trust requirement of tasks and trust degree of cloud resources.Meanwhile,the trust constraint and trust objective are also set into the service selection problem.Therefore,it presents a trust-based heuristic algorithm(improved SPGA2)to solve this problem.By introducing the local search strategy improves the search efficiency of solution space.Eventually,some simulation experiments are conducted to verify the superiority of model and the availability of algorithm.