Application of Kalman Filter for positioning precision improvement in UWB localization system

Accuracy and precision of positioning in UWB localization system depends on the quality of clock oscillators. Usually clock frequencies in particular nodes have different offsets what influences system performance. The paper presents a method for nodes synchronization and distribution of reference time interval. Anchor nodes measure time of arrival of packets conveying reference time information. A Kalman Filter based method for measurement result processing leading to positioning precision improvement is described in the paper. Results of simulations and experiments confirming method efficiency are presented and discussed. Accuracy and precision improvement is presented using CEP (Circular Error Probability) values and positioning error's CDFs (Cumulative Distribution Functions).