An Improved Fault Tolerant System Using Clustering for Multiple Object Tracking in Wireless Sensor Networks

In Wireless Sensor Networks (WSNs), Efficient clustering is key for optimal use of available nodes. Fault tolerance to any failure on the network or node level is an essential requirement in this context. Hence, a novel approach towards clustering and multiple object tracking in WSNs is being explored. The Proposed method employs improved K-means clustering algorithm which employs judicious mix of burdening all available nodes including CH (Cluster Head) to earn energy efficiency and fault tolerance. Initially, node with the maximum residual energy and least distance with fewer faults prone in a cluster become cluster head and node with the second maximum residual energy with minimum distance as next in hierarchy ranked accordingly. Later on, selection of cluster head will be based on available residual energy with no permanent fault due to energy drain and hence the object is being tracked with least fault occurrence. We use Matlab software as simulation platform to check energy consumption at cluster by evaluation of proposed algorithm. Eventually we evaluated and compare this proposed method against previous method and we demonstrate our model is better optimization than other method such as Traditional clustering in energy consumption rate.