A cloud-user behavior assessment based dynamic access control model

Abstract

In traditional role-based access control (RBAC) model, the permission is bound with identity statically, without being dynamically adjusted by user behavior. Cloud users distribute widely and constitute complex and have legitimate identity whose behavior may be incredible, but any attack is achieved through malicious behavior. The cloud-user behavior assessment based dynamic access control model was proposed by introducing user behavior risk value, user trust degree and other factors into RBAC. First, the times of threat behavior was introduced into the information security risk equation to improve the accuracy of user behavior risk value. Then, both the times of threat behavior and the uneven interval of risk threshold were introduced the trust model based on behavior risk evolution to improve the accuracy of user trust degree. Finally, the dynamic authorization was achieved by mapping trust level and permissions. By the simulation experiment in a small campus cloud system, it can be shown that the change of user behavior risk value and user trust degree is more rational under different times and frequencies of threat behavior, and dynamic authorization is flexible by mapping the risk level and the user permissions.