Sybil Attack Detection in Iot Environmental Sensors Using Optimized Fuzzy C-Means Clustering Algorithm

Abstract

IoT environmental sensors are devices that use the Internet of Things (IoT) to monitor and measure various environmental conditions like temperature, humidity, air quality, and light levels. A Sybil attack happens when a malicious node in a network falsifies multiple fake identities to look like many distinct nodes. It affects an IoT sensor network where a malicious node generates multiple fake identities to betray and interrupt network operations. The focus of this work is to use of Machine Learning (ML) algorithm to detect Sybil attack instances in IoT wireless networks. Hence an optimized Fuzzy C-Means (FCM) clustering algorithm is proposed for Sybil attack detection in IoT networks. During the training phase, both the signal feature (SF) and frequency offset feature (FOF) from each IoT device are trained to classify the Sybil attack nodes, by means of FCM clustering technique. For optimizing the clustering performance, a metaheuristic Improved Gradient-based Optimizer (IGBO) algorithm is employed. Experimental results show that the optimized FCM clustering algorithm provides higher detection accuracy with reduced affected packets and computational overhead, when compared to the existing techniques.