A density-based clustering paradigm to detect faults in wireless sensor network

Abstract

Event detection using wireless sensor network is an emerging area of research nowadays in distributed environment. In geographical regions, it is a great area of research to set the sensors for event (volcanic eruptions) detection by taking local decisions. But due to failure of nodes in these regions, it is difficult to detect the event. In this paper, we have proposed an approach of detecting the fault by using Density-Based Clustering method. Our main idea is to form a density-based cluster in which the nodes within the cluster have same behavior (faulty or active). The cluster is formed by using ε-Neighborhood, in which the Density-Reachability and Density-Connectivity concepts are used to get the Density-Based Cluster. By this method, the faults are detected as the nodes which are not in the cluster. Our observation shows better results in modeling a Fault-Detection Paradigm to detect the faults in the network. © 2013 Springer.