Trust2Vec: Large-Scale IoT Trust Management System Based on Signed Network Embeddings

A trust management system (TMS) is an integral component of any Internet Things (IoT) network. reliable TMS must guarantee the network security, data integrity, and act as a referee that promotes legitimate devices, punishes malicious activities. Trust scores assigned by TMSs reflect devices’ reputations, which can help predict future behaviors entities subsequently judge reliability different in IoT networks. Many have been proposed literature, these systems are designed for small-scale attacks deal with where device tries to undermine spreading fake reports. However, prone large-scale attacks. To address this problem, article, we propose called Trust2Vec, manage relationships mitigate performed hundreds devices. Trust2Vec leverages random-walk exploration algorithm navigates relationship among devices computes embeddings, enables it analyze latent structure relationships, even if there no direct rating between two detect attacks, such self-promoting bad-mouthing, embeddings community detection detects blocks communities nodes. The effectiveness validated through simulation. results show achieve up 94% mitigation rate various settings.