A novel cluster-based self-organization algorithm for wireless sensor networks

Wireless sensor networks (WSNs) consist of a large number of tiny sensor nodes. Hence, a cluster-based architecture can be used to deal the self-organization issues of large networks. This cluster-based organization can prolong network lifetime and reduce broadcast overhead. In this paper, we propose an efficient self- organization algorithm for clustering (ESAC), which uses a weight-based criterion for cluster-head's election. This weight relies on the combination of k-density, residual energy and mobility. In ESAC, the node having greatest weight in its 2-hop neighborhood is chosen as cluster-head for a fixed period. ESAC enables to generate a low number of stable and balanced clusters. Simulation results show that ESAC provides better results when compared with WCA (weight clustering algorithm), and with the algorithms proposed respectively by Lin et al., and Chu et al. in terms of the number of clusters formed. On the other hand, it outperforms LCC (least cluster- head changes) algorithm in terms of the number of cluster-heads changes.