A Data Transmission Scheme Using K-Means and Fuzzy Logic for IOT Sensor Based Forest Fire Detection System

Abstract

Forest fires are natural disasters and effective mitigation of these fires require early warning systems. Deployment of sensors that are robust against environmental conditions, collection of sensor data in an energy efficient manner and at real time at the central server, and accurate detection of fire’s existence are all critical elements for implementing such systems. In this paper, we focus on second issue which is energy efficient routing of sensor data in a Internet-of-Things (IoT) sensor network that is deployed for forest fire detection. A Fuzzy-based Cluster Head selection technique for WSN in detecting forest fire is presented. In the proposed scheme, the nodes are divided into clusters using K-means clustering and then the cluster heads are determined by using a fuzzy logic scheme. Unlike traditional parameters, distance from the centroid and the remaining energy levels are used as parameters to select the cluster head. Simulation tool is used to implement the proposed technique. The simulation results suggest that the proposed cluster head selection approach outperforms the existing schemes