Clustering-Based Wireless Sensor Networks Fault Detection Using Recovering Approach

A cluster's efficiency negatively impacts the functionality of wireless Sensor Networks (WSN s). The failure of the clustering head (CH) or a Clustering member (CM) might be attributed to a lack of available energy or a malfunctioning piece of hardware. To fix a clustered system, locate the malfunctioning CH or cluster member. The failure detection and recovery strategy (FDRA) is suggested for cluster-based networks. A backup clustering head (BCH) is selected alongside a clustering head utilizing fuzzy logic methods that consider factors including mobile nodes, remaining energy, demand, and connection quality. The cluster's head, BCH, and a member eventually fall apart. When a cluster head fails, the BCH will replace it. If the BCH can't be elected, any other cluster member will take over. The cluster leader would notice if there were a breakdown in communication between the cluster members. Therefore, the data transmission from the malfunctioning cluster member to the cluster head will need BCH. The proposed and existing solutions for data packet loss, end-to-end latency, energy consumption, and packet delivery are compared. The results show that FDRA for wireless sensor networks based on clustering may improve fault detection and system restoration times, cut down on emissions, and lengthen the lifespan of the network. The experimental results show an F -measure ratio of 92.2%, Detection accuracy ratio of 95.3 % compared to other methods.