Trust-based Approach to Secure Low-Power and Lossy Networks Routing Protocol

Abstract

Resource-constrained things can now connect to the Internet through Wireless Personal Area Networks (WPAN) and Low Power Wide Area Networks (LPWAN). Since then, they are vulnerable to both attacks from inside the network and from the Internet as well. Many research has already investigated the security threats of such an environment. In this article, we focus on how securing the RPL routing protocol in such a constrained and vulnerable environment.To tackle these issues, we investigate how trust that is inspired by social human behavior can enhance RPL security. Our trust-based solution enables motes evaluating locally the threat level when interacting with their neighbors. However, evaluating the trust in such environment is a challenge due to various factors to take into account and the motes resource scarcity.To this end, we propose a distributed trust management approach based on machine learning to evaluate trust and detect untrustworthy behaviors. Machine learning-based approach have been widely investigated in IoT and RPL.We evaluate the performance through significant metrics so as to obtain a good assessment of the effectiveness of our method.