Analysis of GPS and DGPS Performance at Sea

Lu’s Sardinha Monteiro, Terry Moore, Chris Hill

Abstract:	"This trial was conducted onboard a survey vessel with the aim of assessing the accuracy of GPS and of DGPS in a maritime environment, comparing GPS and DGPS positions with ÒtrueÓ positions, obtained by differential carrier phase post-processing. Two geodetic receivers were used, with separate antennas installed on a support structure of known length. The coordinates of the geodetic antennas on the vessel were calculated independently. The baseline between the antennas was then computed from the coordinates and this, when compared to the known length (2.00 m), was used as a simple check on the quality of the GPS carrier phase data and processing. The positions of the two geodetic antennas, at each epoch, allowed determining the reference positions for the other two antennas, which were placed very close to the middle of the support structure. By recording also data on a static location, it was possible to compare the performance of similar receivers, at the same epochs, one on a dynamic environment (at sea) and the other on a static location. This trial allowed concluding that there are two factors which degrade a little bit the performance of a shipboard receiver, when compared to that of a static one: - The shipÕs roll and pitch cause a continuous motion of the antenna, which is forced to loose track of above horizon satellites during its inclinations. The frequent loss of satellites causes a degradation of satelliteÕs geometry, with corresponding increase in Horizontal Dilution Of Precision (HDOP) and also on positioning errors (accentuated by the close correlation between HDOP and positioning errors). - Shipboard receivers performance is slightly degraded by frequent accelerations and decelerations (caused by roll, pitch and yaw) which make signal tracking more difficult, by increased noise floor (due to noise provoked by machines and communications equipment, which may affect the reception of the GPS signals) and by increased multipath susceptibility (particularly if the GPS antenna inclines significantly). The analysis of the results enabled also devising the existence of a much higher correlation between HDOP and DGPS errors than between HDOP and GPS errors. The reason behind this is that on GPS the impact of satellites geometry on the errors is superseded by the other error sources (atmospheric errors, clock errors, ephemeris errors, receiver noise and multipath), which are high enough so as to mask the DOP impact on positioning accuracy. On differential systems, most of those errors (with the exception of multipath and receiverÕs noise) are eliminated (at least on zero baselines) and users experience very low errors. However, there is something that the differential technique cannot change, which is the number of visible satellites and their distribution on the sky. Therefore, when the geometry is bad, its impact on the positioning accuracy is higher if there are no remaining major sources of error, as happens in differential positioning. Therefore, in this paper it is recommended that all DGPS receivers incorporate a capability of displaying a written warning message or an audible alarm every time HDOP exceeds a pre-defined value, thus indicating (in real-time) probable worse positions. As HDOP is a function only of receiver and satellite coordinates and may be predicted ahead of time, receivers should also display the expected duration of the high HDOP condition. For instance, if a buoy is being positioned with DGPS, such alarm could warn the user to wait a few minutes until the quality of the geometry improved, instead of placing the buoy in a less propitious moment. Another example: if a ship is navigating inside a harbour, such alarm would warn the navigator to reinforce the care and attention paid to alternative means of navigation, such as visual aids and radar. This paper discusses also the HDOP threshold that users should set in their equipment. Summarizing, this paper will discuss not only the final results of this trial, but also the reasons behind the differences on accuracy from shipboard to static receivers and also the correlation between HDOP and positioning errors."
Published in:	Proceedings of the 61st Annual Meeting of The Institute of Navigation (2005) June 27 - 29, 2005 Royal Sonesta Hotel Cambridge, MA
Pages:	783 - 794
Cite this article:	Monteiro, Lu’s Sardinha, Moore, Terry, Hill, Chris, "Analysis of GPS and DGPS Performance at Sea," Proceedings of the 61st Annual Meeting of The Institute of Navigation (2005), Cambridge, MA, June 2005, pp. 783-794.
Full Paper:	ION Members/Non-Members: 1 Download Credit Sign In