Data Similarity Based Dynamic Node Clustering Using Bio-Inspired Algorithm for Self-Organized Wireless Sensor Networks

Wireless sensor networks have been used by various environmental sensing applications as an interface between physical and computational environments. The wireless data gathering techniques commonly use the node clustering approaches to reduce energy consumption in the network. Moreover, it can also organize the network to be a stable network, improve the scalability, and prolong the lifetime of the cluster head through load balancing. However, few node clustering methods consider the data similarity based cluster formation referred on the spatial correlation to reduce the size of the large amounts of data and to detect the anomalous events. Furthermore, it can also manage the network through a self-organized approach to reduce the fault in the network. In this article, we propose an approach for data similarity based node clustering through a bio-inspired algorithm as a self-organizing mechanism. The simulation results show that the firefly-inspired synchronicity can synchronize the nodes on the network in order to perform together actions simultaneously. The clustering technique can establish the clusters based on the spatial similarity readings and can detect the anomalous data robustly.