Inferring Threatening IoT Dependencies using Semantic Digital Twins Toward Collaborative IoT Device Management

IoT Device Management (DM) refers to registering, configuring, monitoring, and updating IoT devices. DM is facing new challenges as dependencies between IoT devices generate various threats, such as update breaks and cascading failures. Dependencies-related threats are exacerbated by the fragmentation of the DM market, where multiple actors, e.g., operators and device manufacturers, are uncoordinately ensuring DM on interdependent devices, each using its DM solution. Identifying the topology of threatening dependencies is key in developing dependency-aware DM capabilities for legacy DM solutions to tackle dependencies-related threats efficiently. In this work, we apply Semantic Web and Digital Twin technologies to build a decision-support framework that automatically infers the topology of threatening dependencies in IoT systems. We integrate the proposed framework into the in-use Digital Twin platform Thing in the future and demonstrate its effectiveness by inferring threatening dependencies in smart home scenarios managed by ground-truth DM solutions, such as Orange's implementation of the USP Controller and Samsung's SmartThings Platform.