Machine Learning Approaches To Foster Trust Among Social Iot Devices

Abstract

In today's communication era, physical objects or items have gained the ability to communicate with each other and exchange information through the Internet. The advancement has made these objects smarter and capable of interacting with authenticated devices, leading to the emergence of the term "Internet of Things" (IoT). Social IoT (SIoT) devices take this concept further by incorporating social networking paradigms into interactions among smart devices. Achieving socialization among these devices necessitates a secure and trusted connection between them. Consequently, privacy and security enhancements for SIoT devices have become a significant concern and are now essential in contemporary communication.The primary responsibilities within trust management encompass the design of trust architecture and the assessment of reputation. However, applying existing trust architectures and reputation evaluation methods directly to the Internet of Things (IoT) poses challenges. This is primarily because of the huge volume of physical entities involved, the constrained computational capabilities of these entities, and the highly dynamic nature of the IoT network. In this work, we have presented trust management architecture and its evaluation in IoT. It also presents, how trust management helps in reliable decision making process and we have also discussed about various ML approaches used in building trust management system.