Data prediction in Wireless Sensor Networks using Kalman Filter

The use of Wireless Sensor Networks to monitor the environmental parameters is on the rise. Wireless Sensor Networks can be employed in various applications ranging from personal health-care, indoor temperature monitoring to military theater. The data from each sensor is required to be collected at the central node for further processing. In spite of special designs and routing protocols for improving efficiency, data can often be delayed or go missing due to node failures. This problem can be overcome by predicting the data beforehand and using the predicted data when there is a delay in the arrival of the data or when the data goes missing. The prediction of such data is a very challenging problem. Network performance experiences a bottle neck with node failure and a mechanism should be employed for predicting the lost data so as to maintain network resilience. In this paper, the usage of Kalman Filter in Wireless Sensor Network is shown and the performance of the same is evaluated.