Self-Healing for Self-Organizing Cluster Sensor Networks

Abstract

Self-healing approach enables sensor network configuration robust to node failures. Fault-tolerance of k-connected network can be achieved using threshold of critical transmission range and node density. A self-healing scheme for clusterhead failure in k-connected cluster sensor network is presented in this work. Clusterhead failure triggers re-clustering in existing clustering schemes for wireless sensor networks (WSN). Re-clustering being a global operation in sensor network, incurs high energy overhead in constrained battery powered WSN. An alternative framework for WSN is based on localized scheme, called 'clusterhead-clique', taken in this work for self-healing. The clusterhead-clique is nearest neighborhood of clusterhead within half transmission radius, results a localized scheme for clusterhead position transfer. The simulation of the protocol based on our approach self-healing achieves 35% energy-efficiency in reconfiguring clusterheads over the existing re-clustering scheme