A performance evaluation of the internet based static GPS computation services

Abstract

In conventional relative GPS surveying, users must use at least two sets of GPS receivers, as well as the associated processing software packages. This is not only a hardware or software constraint, but also a logistical problem. Furthermore, in many instances the data processing steps may be more difficult for users with limited experience in GPS technology. These problems can be resolved with an Internet GPS Service, whereby the user can finish all of the processes with only one GPS receiver without the need to use a base-station or to buy and maintain any data processing software. This study explores the usage of internet-based GPS services. Two fixed stations, NCTU and DONS established by National Chiao Tung University and National Cheng Kung University respectively, were used. The observation period was from 2001/3/1 to 2001/3/31. To establish the reference data set, the IGS station WUHN and PIMO were used as reference stations in order to compute the NCTU and DONS positions with daily observation from 2001/3/1 to 2001/3/7. The average of the seven daily positions is then used as the reference data set. The subdivided set of data sets, taken at intervals of 24, 6, 4 and 2 hours, were then processed with the Internet GPS Services. The results are evaluated using mean absolute error, maximum absolute error and minimum absolute error. In this paper, with the situation of an observation interval of 24 hours, the differences of horizontal coordinates computed from all Internet GPS Services are in the cut range, but the resultant height values vary, some within the 2 similar to 3 cm range, and some within the 10 similar to 20 cm range. For the baseline of 170286.0442m, the error is +/- 29.63mm.