Maximizing the reliability of clustered sensor networks by a fault-tolerant service

Wireless Sensor Networks (WSNs) are usually dispersed in the harsh environments which are prone to different types of faults. So fault-tolerance is an essential characteristic which should be considered in the design level of such networks. On the other hand, WSNs are so energy-constrained and through some solutions, like clustering the nodes, energy should be conserved as much as possible. In this paper, we propose Fault-Tolerant Service (FTS) for the clustered protocols. At the beginning, all the nodes are divided into some groups as clusters via the Energy-Efficient Distance-based Clustering (EEDC) [4] protocol. Then, the FTS along by the main operation of the network, i.e. data gathering, is performed by the cluster-heads . The FTS is composed of three steps: fault detection, fault diagnosis and fault recovery. The main idea of fault detection in the FTS is message exchange. Redundancy is effectively utilized in this paper, where we propose a novel approach to elect the spare cluster-heads in which the elected spare cluster-heads are enough close to the cluster-heads. We have investigated a clustered WSN via simulations which performs the FTS. The simulation results and probabilistic analysis have confirmed the effectiveness of the proposed FTS in maximizing the reliability of the clustered sensor networks. Although the reliability is significantly improved, we show, via simulations, that a clustered WSN should consume about by 30% more energy in order to perform the FTS.