Research on Efficient Top- k Query Based on ARIMA Time Series Model

Abstract

In wireless sensor networks, Top- $k$ query is often used to query the first $k$ values which can satisfy the user’s criteria. In the process of query, in order to reduce the transmission of redundant data, different filtering windows are set for each sensor node. However, the filter windows often update because of the dynamic change of the data which will cause the huge consumption of the node energy. In order to solve this problem, a new algorithm based on ARIMA is proposed called AAFU (ARIMA approach based on the filter updating) in this paper. With this algorithm which is based on the FILA algorithm and ARIMA model, the base node can use time series model to predict the future data according to the collected historical data. This algorithm can reduce energy consumption that can make the sensor network effectively deal with the node window updating through comparison to decide whether update the filtering windows. The simulation results show that AAFU algorithm is superior FAPU algorithm with ensuring the accuracy of query.