Data prediction Strategy for Sensor Network Clustering Scheme

Sensor network clustering scheme is an efficient method that prolongs network lifetime. However, when it is applied to an environment in which collected data of the sensor nodes easily overlap, sensor node unnecessarily consumes energy. Accordingly, we proposed a data prediction scheme that sensor node can predict current data to exclude redundant data transmission and to minimize data transmission among the cluster head node and member nodes. Our scheme excludes redundant data collection by neighbor nodes. Thus it is possible that energy efficient data transmission. Moreover, to alleviate unnecessary data transmission, we introduce data prediction graph whether transmit or not through analyze between prediction and current data. According to the result of performance analysis, our method consume less energy than the existing clustering method. Nevertheless, transmission efficiency and data accuracy is increased. Consequently, network lifetime is prolonged.