Dynamic prediction clustering scheme for mobile sensor node of sensor network environment

Sensor network is being used in a variety of areas. As sensor network nodes are evolving into a mobile environment, we should seek an appropriate method to set up clusters and select cluster headers. In this article, we suggest a dynamic prediction clustering algorithm, which uses directions, angles and hops among the dynamic skyline query properties. This proposed algorithm is one that builds clusters and selects cluster headers on the basis of mobile sensor nodes. It can reduce the waste of unnecessary energy of mobile sensor nodes which are caused by the occurrence of ldquoAdvrdquo message. And in proportion to the density of sensor nodes for efficient clustering, this algorithm builds dynamic clusters, and extends the network life cycle by reducing the average energy of sensor nodes by 2.4 times. Also keeping dynamic clusters and maximizing the hop counts within clusters, the algorithm has reduced the average energy consumption of a node by 14%.