Computing today is becoming pervasive. To provide pervasive computing environments flexible, dynamic and less distractive security mechanisms, trust managements are introduced. However, current trust frameworks are designed for specific scenarios and can’t fulfill requirements in different environments, which leads to that these trust frameworks can’t be deployed in different environments and pervasive devices can’t enjoy pervasive computing anywhere really. To remedy this problem, this paper proposes a universal trust framework. In our framework, the flexible attribute vector model is used to represent trust evidence that are needed to evaluate trust relationships, thus the framework can work in identity-based trust environments as well as context-based trust environments. Besides,
the universal trust framework defines four kinds of trust relationships that are suitable o model trust relationships in
various scenarios. In a word, the proposed universal trust framework is an especially designed trust framework that can unify trust management in different pervasive computing cases.