Abstract—In wireless sensor networks, the data periodically sensed by sensor nodes are usually of high temporal correlation. Therefore, prediction based data aggregation scheme can be designed to reduce the amount of data transmission and save sensor nodes’ limited energy and bandwidth. We proposed a temporal correlation based data aggregation scheme in this paper which utilizes time series model to predict the data of next several periods at both ordinary sensors and aggregators based on the same amount of recent sensed values. We show through experiments using real sensor data that our proposed scheme can provide considerable aggregation ratio while maintaining a low prediction error rate.
Index Terms—Wireless sensor networks, data aggregation, data correlation, ARIMA model.
Guorui Li is with the Electronic Information Department, Northeastern University at Qinhuangdao, Qinhuangdao, Hebei 066004 China (e-mail: email@example.com).
Ying Wang is with the Department of Information Engineering, Qinhuangdao Institute of Technology, Qinhuangdao, Hebei 066100 China (e-mail: firstname.lastname@example.org).
Cite: Guorui Li and Ying Wang, "An Efficient Data Aggregation Scheme Leveraging Time Series Prediction in Wireless Sensor Networks," International Journal of Machine Learning and Computing vol. 1, no. 4, pp. 372-377, 2011.