Realisation of a cluster-based protocol using fuzzy C-means algorithm for wireless sensor networks

In this study, real-time design and implementation of an energy-efficient cluster-based protocol for wireless sensor networks (WSNs) are presented. The formation of suitable clusters is of prime importance to balance energy usage by sensor nodes within each cluster of the cluster-based WSNs. This leads to energy savings for the sensor nodes resulting in longer network lifetime. To obtain this objective, the fuzzy C-means (FCM) clustering algorithm is incorporated in the protocol. The protocol is realised on a hardware test-bed with the support of the embedded operating system, TinyOS. Experimental results obtained from a scale-down laboratory based test-bed with up to 50 sensor nodes are provided to illustrate the efficacy of the WSNs using the proposed protocol and compared with the well-known cluster-based protocols such as Low Energy Adaptive Clustering Hierarchy. It has been shown that the FCM protocol is able to achieve better organisation of the network and thus can extend its lifetime under varying operating conditions and with different number of nodes.