Research On Users' Behaviors In Cloud Computing Environment Based On Belief Modification

In recent years,as a new information service model,cloud computing with its large-scale data storage resources and computing power which provides users with low cost and customized services.Cloud computing is fast and convenient services brought a large number of users,because the cloud environment has the characteristics of open,dynamic and resource sharing,attackers and malicious users can turn to take the 'user's' legal identity to attack.Consequently,they steal and damage interest of service providers and other users.The problem of user behavior trust has become the hotspot research in the field of cloud computing security.In this context,this paper takes the security management mechanism of user behavior analysis as the goal,we deeply studies the work of cloud computing users'behavior security and trust mechanism.First of all,an Intuitionistic Fuzzy Analytic Hierarchy Process(IFAHP)is proposed on our model combined with trust-worth computing.And it evaluated the trustworthiness of users when they use cloud services.Considering the user behavior has a certain degree of randomness and fuzziness,the model is decomposed into several factors,which are considered to be more reasonable,and the objective and quantitative characteristics are obtained.The trust value is obtained by using the model,and the degree of trust is given to reduce the security of the cloud environment by using the model.Then we combined with traditional access control technologies and trust management technologies,we proposed users' belief modification method based on game theory.According to the different sensitivity and importance of services,we gave the different degree of modification.We introduce the concept of the game between the user and the service provider cloud.After that update the user's belief(trust value)dynamically during the each stage of game;We also introduce the concept of trust level into the traditional access control.To make access control model more flexible,hierarchical management will have the trust of users,According to the user's trust value as the decision basis,user can obtain the authorization.To evaluate the user trust value,the mechanism of the window is introduced.The combination of historical and timeliness of user's behaviors are used to assess the credibility of the user's behavior and belief modification.Experiment show that the two models can effectively identify malicious users in a cloud environment,they can reduce risk caused by user actions in a cloud environment as well.Two models achieve service classification management:take different trust management services TMS(trust management service)with a different importance and sensitivity of the service.And we take system false positive rate and false negative rate into consideration of practical factors,so the models ensure users to use cloud service with highly fault tolerance defense.