Two level data fusion model for data minimization in Periodic WSN
Abstract
Wireless Sensor Network is widely used for environment monitoring application. Sensors nodes used for such applications are densely distributed and senses redundant data due to temporal and spatial correlation. Data fusion and aggregation is commonly introduce to remove redundant data and transferred minimized data to the sink node. Aggregation and data fusion methods which involves single attribute of sensor nodes for monitoring may produce inaccurate information at Sink node. In this paper, two level data fusion model is proposed to forward accurate and minimize data to sink node. At level1 data fusion, single most similar measurement is transferred to Cluster Head. At level2 data fusion, similarities between measurements are determining using multi-attribute Euclidian similarity function. We analyzed the performance of proposed model on real dataset and the obtained result shows that two tier data fusion model transmits accurate data in an energy efficient way.