Enhanced Belief Function-Based Decision Blending for Detecting Fault in Wireless Sensor Networks

Abstract

In wireless sensor networks (WSN), various software and hardware issues can lead to various fault types. The issue can be found using many forms of fault detection. The diverse obstacles determine the distinct fault kinds and need to find out effective fault detection and problem-solving are required. This chapter discusses four main types of faults: gain fault, offset fault, stuck-at fault. In this work, the authors use the notion of decision blending to categorize the blending outcomes and to assess the accuracy in order to save energy and make better use of the available bandwidth for data transmission. Three performances are assessed by the decision blending function: detection accuracy (DA), sensitivity, and rate of error. Different methods, such as k-nearest neighbor (KNN), enhanced extreme learning machine (EELM), enhanced support vector machine (ESVM), and enhanced recurrent extreme learning machine, are used in the belief function approach (ERELM). Here, the authors applied decision blending approaches in WSNs to emulate these techniques for improving belief function.