Dynamic Trust-Based Device Legitimacy Assessment Towards Secure IoT Interactions

—Establishing trust-based interactions in heteroge- neously connected devices appears to be the prominent mechanism in addressing the prevailing concerns of confidence, re- liability and privacy relevant in establishing secure interactions among connected devices in the network. Trust-based assessment of device legitimacy is evolving given IoT devices’ dynamic and heterogeneous nature and emerging adversaries. However, computation and application of trust level in establishing secure communications, access control and privacy domain are rarely discussed in the literature. To compute trust, based on the quality of service, direct interactions, and the relationship between devices, we introduce a multi-factor trust computation model that considers the multiple attributes of interactions in an IoT network of heterogeneous devices providing a wide range of data and services. Direct trust is estimated for quality of service considering the response time, reliability, consistency, and integrity attributes of devices. The time decay factor influences the credibility of computed trust over time. The policy-driven mechanism is employed to sift the devices and isolate the mali- cious ones. Extensive simulations validate the proposed model’s effectiveness using Contiki’s Cooja simulator for IoT networks.