Improving Network Performance in Wireless Sensor Network Using Self-Maintenance Clustering Model

In Wireless Sensor Networks (WSNs), fault tolerance of a sensor node is a demanding issue since sensors are usually deployed in unattended environments. Limited memory, processing power, and communication range of sensor nodes make conventional fault tolerance schemes infeasible for WSNs. This work introduces a distributed self-healing methodology in which the detection, diagnosis and healing processes were performed at both node and cluster head levels. At node level, battery, sensor and receiver faults were diagnosed. At cluster head level, transmitter and mal-functional nodes were detected and recovered. The simulation results showed that the proposed methodology is precise in locating malfunctioning nodes and fast in finding a cover for such nodes.