Dynamic sensor data stream estimation method based on Kalman filtering

In the application system of wireless sensor network, the sensor nodes acquire the parameters of the monitored objective constantly at a certain interval. For the data center, the sensor data form a wireless sensor data stream, as the data model of the sensor data stream may vary with time, so the sensor data stream is dynamic. Aiming at the problems of low estimation precision, high calculation complexity and that the estimation model does not update timely in current estimation methods of the wireless sensor data stream, a dynamic sensor data stream estimation method based on Kalman filtering is proposed, which is called KF-CAMLR (Kalman filter-correlation analysis-based multiple linear regression). In this method, the Kalman filter is adopted to realize the dynamic adjustment of the estimation model; in order to decrease the calculation complexity of data stream estimation, multiple linear regression estimation method based on correlation analysis is adopted, which combines the Kalman filtering with the multiple linear regression model to achieve the accurate estimation of the dynamic wireless sensor data stream. The experiments on real sensor data were carried out; and the experiment results show that the proposed dynamic sensor data stream estimation method based on Kalman filtering can estimate the dynamic sensor data in the sensor data stream efficiently with high estimation accuracy.