Modelling of Selective Availability for Latency Effect Compensation in WAAS

Abstract

Differential Global Positioning System (DGPS) referenced stations at well-surveyed position can provide accurate pseudorange corrections that will reduce the DGPS users position error to 5 meters. However, these corrections are valid over an area of roughly 50km around the DGPS reference stations. To alleviate this difficulty, Wide Area Augmentation System (WAAS) has been suggested. However, the time delay or latency effect, of the correction signals will raise the error level because users have to apply the pseudorange corrections generated some moments earlier by the reference station. The latency effect gets further aggravated in case of WAAS where Wide area Master Stations (WMSs) have to collect information from various Wide area Reference Stations (WRS) and compute the necessary correctional information and broadcast them to users through geostationary satellites. Typically the time taken for this activity is of the order of 6 seconds. During this period the user is without the services of any corrections. Thus if pseudoranges are to be corrected each second, then one needs correction values for the intervening five seconds times, which somehow need to be generated before the correct values are received from the broadcasts. The error caused by the delay depends on the changing rate of the error source, which is mainly dominated by Selective Availability (SA). This paper suggests the use of a Recursive Least Squares (RLS) filter to model the SA and concludes by showing its effectiveness in compensating the latency effect. The value of this work lays the fact that the GPS signal measurements were obtained from a commercial grade dual frequency receiver and therefore of considerable practical significance. It is also observed from the pseudorange measurements that still there exists a part of SA even after May 2000.