Dirichlet-Based Initial Trust Establishment for Personal Space IoT Systems

Trust has played a crucial role in enhancing the security of IoT systems over their lifecycles from creation to retirement. Particularly, in a personal space IoT system where devices join and leave the system dynamically, it is important to evaluate the device's behavior in the form of trust on its admission to the system to reduce the risk and uncertainty of the overall system. Currently, proposed trust evaluation models primarily rely on the historical knowledge or trusted recommendations. However, in many situations, such information is not available at the first encounter between the system and the device. The challenge tackled by this paper is how to establish whether a device can be trusted to a level that merits further evaluation for admission into an IoT system when it encounters the system for the first time. We propose a Dirichlet-based trust assessment model to establish the initial trust that the system places on a device in a mobile and dynamic environment called personal space IoT. The proposed scheme can also be used to affirm the trust of a device during its operation or when it is being re-admitted to the system after an interruption. We describe and evaluate our proposed model theoretically and by simulation.