Cluster Head Recovery Algorithm for Wireless Sensor Networks

A wireless sensor network (WSN) is a set of sensor nodes devoted to sense sizes of physical phenomena of a geographical area of interest. The sensed data so collected are transmitted to a specific node called sink via cluster-heads (CHs). In most wsn's applications, CHs play an important role in aggregation and transmission of sensed data. Indeed, a failed CH may abort the overall mission of the sub-network. Due to their crucial functions, CHs must be designed and maintained to be robust enough in order to face trouble coming from the harsh environment. Therefore, each failed CH should be detected, recovered and a new one takes the relay without data loss. In this paper, we propose a novel algorithm called Cluster-Head Recovery Algorithm (CHRA). In this algorithm, a checkpointing technique is used to save CH's state in order to create a recovery path within each cluster. We compute the node's failure probability through a Bayesian network. The efficiency of the proposal is that can detect and recover the CH's failure online and in a distributed manner. The proposed scheme functionalities have been conducted by simulation and the obtained results demonstrate the effectiveness of the proposal.