Adaptive Kalman Filter for target tracking in the UWB networks

The aim of this paper is to emphasize the benefits of using the information provided by the UWB network for the Adaptive Kalman Filtering (AKF). The standard 802.15.4 specifies in the section PHY UWB the field of parameters Ranging FoM, parameters that show the trust level of the receiver in determining the reception moment of leading edge of the ranging pulse. Using the specified parameters, Confidence Interval and Confidence Level, and based on statistical theory the current standard deviation is determined for every anchor of the network. The standard deviations of the anchors are combined in order to get the observation covariance for the Kalman filtering. In the steady state the innovation should be within the Confidence Region ellipsoid and if the innovation does not fall in that region the filter diverges or it is biased and it should be online tuned. The difference between the innovation and the Confidence Interval will constitute the information for adjustment of the state covariance in the Kalman Filter.