مدل خودپنداره برای شبکه های سنسور بی سیم

Wireless Sensor Networks have wide variety of applications and their nodes are prone to failure due to a hardware failure or malicious attacks. The self-healing mechanism is used for fault detection, diagnosis and healing. However, implementing the self-healing procedures at the cluster head affects the network performance. In this paper, we present a distributed self-healing approach for both node and cluster head levels. At node level, battery, sensor and receiver faults can be diagnosed while, at cluster head level, transmitter and mal-functional nodes can be detected and recovered. Compared to the state-of-the art methods, our model tolerates up to 67.3% of different hardware faults at node level. Moreover, it realized a detection accuracy of sensor circuit fault tolerate up to 76.9%, 52% of battery fault and 71.96% of receiver faults. At head class level, 75.7% of transmitter fault and 60% of microcontroller circuit fault are realized.