Studienarbeit 185

Performance Evaluation of the Starfire Correction Service.

von Michael Mink

Task Definition: Prof. Dr.-Ing. B. Heck
Supervisor at EADS Astrium: Dipl.-Ing. M. Kirchner
Interne Bibliotheksnummer: 185

The „Studienarbeit“ has been performed in the period from 18.11.2009 – 18.12.2009 at EADS Astrium GmbH in Ottobrunn (near Munich) and concentrated on the evaluation of the performance of the Starfire Correction Service.

Starfire is a global service which provides correction data for non-differential Global Positioning System (GPS) measurements. The system is based on a global network of reference stations to compute and provide GPS satellite orbit and clock corrections. After the data have been processed they are sent via geostationary satellites to the end user to improve the positioning accuracy of the receiver. Starfire is a licensed service of NavCom Technology whose performance bond with respect to accuracy and applications is very high (~10cm horizontal, ~20cm vertical). Because it is a proprietary service and usable only with NavCom receivers, independent performance tests are barely available. This study is aimed at testing the performance of the Starfire service and at assessing the impact of the correction data on the position.

First of all some, evaluation scenarios have been carried out in which the applications could be reasonably tested to give an answer to the basic questions for performance evaluation. Although the NavCom SF-3050 receiver is capable of receiving GLONASS signals it is intentionally abdicated to make use of them because the Starfire service is limited to provide correction data for GPS satellites only. In using only GPS satellites the influence of Starfire can be made more transparent. All the measurements are performed with a static antenna with free sight to every direction above horizon. The measured coordinates refer to precalculated coordinates of the antenna whose position accuracy is about 1 cm – much better than the expected Starfire accuracy. Some of the tests that are accomplished shall be alluded here:

Long-term measurements (~ 3 days)

In a first scenario long-term measurements (~3 days) have been made to demonstrate the impact of Starfire correction data in the accuracy of the topocentric coordinate differences (figure 1). While the coordinate differences vary between -4m and 2m, the results improve using Starfire (-0.1m to 0.3m) (figure 2).

Convergence Period

The global accuracy according to NavCom is supposed to be 10cm horizontally and 20cm vertically globally for a dual frequency equipment after a static convergence period that is necessary to reach the maximum accuracy level.

It should be noticed, that NavCom does not provide a definition of the convergence period and therefore to analyse it an “own” definition has to be introduced. The convergence period is defined as the period between switch on until the position error falls below a certain level for the first time, namely 10cm horizontal (PE2D) , 20cm vertical (PE1D) and 20cm in the 3D position (PE3D).

Figure 3 shows a typical convergence period. After switching on the accuracy is about 1-2 meters. During the convergence period the accuracy improves to decimeter level.

RTK-Extend

RTK Extend is a StarFire specific feature that can be obtained optionally for NavCom RTK receivers (figure 4). If a rover loses the radio link to the base station for instance due to obstruction, it receives no correction data from the base station resulting in reduced accuracy in the position after a few seconds (figure 5). Until re-establishing the connection to the base station differential GPS cannot be performed to improve accuracy. Like the feature name indicates, the RTK Extend feature extends the RTK mode during loss of RTK signal up to 15 minutes. This is possible since the StarFire signal contains clock and orbit corrections. The remaining errors are assumed to be a static difference between the last RTK corrections and StarFire corrections at the epoch of RTK loss. In fact, the remaining errors change with time, however. Within a few minutes the errors grow slowly so that the RTK Extend option is supposed to be a useful feature to bridge short interruptions in RTK base data availability. RTK Extend improves the working conditions for the user as short communication losses to the base station have less impact on the precision of the position.

Applications

According to the results in this study Starfire is a multi-purpose service which provides an accuracy of the position of about 10 to 20cm. The primary purpose of Starfire is the use for agricultural purposes, to automate harvester-thresher for example. The requirements in this application is of course the knowledge of the position at any time when decimetre accuracy is absolutely sufficient.

The basic conditions over the wide fields are very good concerning satellite visibility. Meanwhile the Starfire service comes in operation in offshore as well as aerial applications. A maritime application is for example the port departure where continuous static measurements are possible. So the receiver on the ship can converge to maximum accuracy before departure. Another application can be found in the unmanned flight guidance. In this application a free sight to the satellites is guaranteed. The vehicle can be positioned with high accuracy at any time. Land survey and geographic information applications can be well realized by Starfire. Starfire is not recommended for use in urban environments where the problems with obstructions are too serious. For example using Starfire in cadastral applications is not recommended for the reasons that a loss of GPS satellites cause a loss of accuracy and besides the decimetre accuracy is not suffcient.

In principle the Starfire service is applicable when a continuous and free sight to a suffcient number GPS- and Starfire satellites is guaranteed. Nevertheless it always should be kept in mind that a convergence period is needed to achieve the promised accuracy. Working with Starfire in urban environment is not reasonable as obstructions always cause massive loss of accuracy.

List of References

Dr. Peter Williams (2004): Satellite Precise Positioning for Agriculture and Farming, FCC Rural Satellite Forum

Tenny Sharpe, Ron Hatch, NavCom Technology Inc.; Dr. Fred Nelson, John Deere Co. (2002): StarFire and Real-Time GIPSY: A Global High-Accuracy Differential GPS System, p. 1-6, available at www.navcomtech.com.