Proving Trust Locally

Simulators and analyzers for protocols do not distinguish between the local 'subjective' view of the agents and the global 'objective' view of the analysis perspective. In practice this means that security analysis in general and protocol analysis in particular do neither model the agents local beliefs nor their local deduction power in a satisfactory way. This paper suggests a solution to the problem by proposing a new approach to epistemic logic, explained in terms of term rewriting, in order to make agents capable of performing logical deductions themselves. Local deductions of security properties, like trust, are crucial to assure that the agents can expand their knowledge about the environment they inhabit.