Distributed Fault Node Detection and Classification using Fuzzy Logic and Management Scheme for Wireless Sensor Networks
Kathiroli Raja, Rajnita Leichombam · 2018
A number of sensor nodes varying from hundreds to thousands are deployed forming Wireless Sensor Network (WSN) is deployed over an area to monitor harsh and unpredictable environmental conditions like forest fires, floods, man-made disasters etc. Due to the unavoidable harsh conditions of the surroundings and resource constraints like battery power, the probability of network failure rises. The components of a sensor node like transmitter circuit, receiver circuit, sensor interface might cease to work. Several consequences like loss of communication links between the nodes, disturbance to the network topology might occur due to the occurrence of these faults. Identifying the faults and recovering from the failures are crucial to increase the quality of service of the network. The traditional methods of fault detection adopt centralized approach where fault detection might not happen if the centralized fault detector fails. The fault nodes are rather discarded in the traditional methods. To address these issues, we propose a Distributed Fault node Detection and Classification using Fuzzy logic and Management scheme (DFCDFM) for Wireless Sensor Network (WSN). The nodes are classified using fuzzy logic according to the fault status. The type of node is determined by the fuzzy inference system according to designated membership functions. Further, a management scheme that involves the reuse of some type of fault nodes are provided. A routing scheme based on bee inspired algorithms is adopted in the management scheme. An optimized routing path is obtained using the routing scheme based on bee inspired algorithm.