ANFIS AND FUZZY BASED FAULTY NODE DETECTION FOR WIRELESS SENSOR NETWORK

Abstract

One of the main challenges in WSN is detection and management of faulty nodes. Currently, the great majority of fault detection approaches rely on surrounding nodes' views of data, which ignores the happenings of the event or issues coverage. In this paper, we describe a novel distributed fuzzy logic-based defective node identification approach for heterogeneous WSNs. The algorithm, it is hypothesized, weights the observed data depending on criteria such as distance, coverage, and the difference between them. When the proposed distributed technique is employed, each sensor node can accurately estimate its own state even in the presence of events such as sudden and temporary failures. According to rigorous simulations, the proposed technique decreases false positives while also improving problem detection accuracy.