Research On Trust Mechanism Based On User Behavior And User Relationship In Cloud Environment

With its powerful computing advantages,cloud computing has become the mainstream of current computing services.With the promotion of the market and the increase of users,security issues have gradually emerged.If the security of cloud computing is not effectively solved,the cloud computing model will be difficult to gain recognition from cloud users,which will slow down the development of cloud computing technology and business.Many experts and scholars have enhanced the security of cloud computing platforms through technological innovation and mode changes.Although they have achieved certain results,they still cannot completely solve the security problems they face today.At this time,many experts and scholars found that applying trust to cloud computing can solve the current problem.This paper fully studies the previous trust algorithms and models.By analyzing the advantages and disadvantages of each classic model,it is found that user behavior data and user relationship data can reflect the credibility of users in the cloud environment to a certain extent.In view of the above problems,this paper focuses on the trust of users in the cloud environment from two aspects.On the one hand,the calculation method of user trust vector based on AHP and Markov chain is proposed.Faced with the subjective characteristics of trust,an improved analytic hierarchy process AHP model is proposed.The criterion layer is manually set up,and the criterion layer is scaled.In the evidence layer,the user behavior is quantified by the method of user behavior quantification,which can quickly calculate the score of user behavior under the criterion layer.Finally,according to the characteristics of the hierarchical analysis model,the vectorization method of calculation results is proposed,which is used to vectorize the calculation results and satisfy the three-dimensional structure of the user trust vector.At the same time,in order to avoid the impact of user-relational spoofing on trust delivery,a Markov chain-based trust delivery method is proposed.The past global trust delivery method is changed into local trust delivery,and finally the behavior trust value is based on the user behavior and user relationship activity.And the trust transfer value is weighted to calculate the user's comprehensive trust vector.On the other hand,the user risk assessment model based on Bayesian network is studied.For the user's deceptive behavior and multi-user collaborative malicious behavior,a Bayesian risk assessment model is established,and the Bayesian network is structured with improved greedy algorithm.According to the user behavior sample data,the parameter learning are established by maximum likelihood estimation.And a Bayesian network with high classification accuracy rate is constructed,and the users are rated according to their trust.Finally,the user trust vector is modified according to the trust level to make it reflect the user's real situation as much as possible.Then the user trust vector is used as the reference standard to manage the rights of the user to reduce the security threat of the cloud platform.Finally,the simulation experiment is carried out in the simulated cloud environment.The experimental results show that the proposed calculation method and model can carry out the trust analysis area for cloud users in different scenarios and divide different trust levels.The experimental results show that the model improves the security of the cloud computing platform to some extent.