Test and calibration field for GPS receiving antennae

Sources of error in the GNSS GPS example

In recent years, the development and expansion of high-precision local, regional and global GPS networks has been rapidly increasing. Moreover, the production of low-cost GNSS equipment has made the processing of heterogeneous antenna arrays (GPS receivers) mostly inevitable in the everyday life of modern geodesists.

This fact stems from not only the use of different types of antenna in the building of single parent networks, but also the use of GNSS antennas specifically. If the differences between the various antenna series are disregarded, one must no longer expect a negligible deterioration of the intermediate and final results (e.g. phase ambiguities, coordinates, accuracy measurements) in the course of the GNSS signal processing. For several years, users of global GNSS have known that even within a model range, for example, production-related deviations (e.g. construction tolerances) can be made, which must be supported by the use of improved and advanced models, and grown on the basis of demands.

Thus, this station-specific interference effect potentially causes a limitation in the accuracy. In previous years, this limitation was diluted by the noise of the signal, "the signal's noise disappeared". Today, because of neutral atmospheric areas, the ionosphere, satellite orbit data or multipath effects, the limitation in the accuracy plays an increasingly important role in the highly precise point positioning as well as in geodynamic applications like observing static concepts as well as in real-time applications. As research shows, the effects of non-comprehensive or incorrect handling of this error have amounted to an influence of up to several centimeters for the coordinates under unfavorable conditions.

Using the example of GPS receivers, the current research of the GIK will be described below. Similarly, reference is made to an existing test and calibration field for GPS receivers.

Principles of the modelling and calibration of antennae

Receiving antenna model: phase centre offset

Receiving antenna model: phase centre offset and elevation-dependent phase centre variations

The model for the GPS receiver antenna is primarily composed of the constant phase centre offset (PCO) and is dependent on the direction of the incident signal's phase centre variation (PCV) values.

In this context, the satellite’s antenna phase centre can be termed as either mechanical or electrical, both of which are of particular importance.

With respect to the mechanical structure, its phase centre corresponds to a theoretical (e.g. production-related) reference position of the phase centre, and is typically part of the symmetry of the antenna's axis. The position of the fixed antenna is described in a specific coordinate system. It originates at what is called the antenna reference point (ARP), which is located at the centre of the antenna's base. The coordinate axis is coincident with the axis of a local topocentric system of coordinates. It should, however, be stated that the mechanical phase centre (for example) is composed of several individual components.

The electromagnetic signals are received by the GPS antenna, though these interactions are offset due to variations (e.g. elevation angle and azimuth). The result is a function of the position of the electrical phase centre, where the GPS signals can be received and calibrated. The electrical phase centre of the two GPS carrier frequencies do not usually coincide (frequency dependence). This continues to result in different phase centres as a function of the linear combination for GPS evaluations.

The PCO is located at a central position of the electrical phase centre which is indicated in the antenna-specific coordinate system frequency metric. The PCV directionality takes directionality into account, and thus represents improvements of the phase measurements with respect to the average electrical phase centre. PCV are also frequency dependent.

In the context of the antenna calibration method, individual values for PCO and PCV can be determined using both carrier frequencies.

A classification of the principle procedures for the receiving antenna calibration is possible using location (lab, field), signal type (artificial, real), or the existing /non-existing reference to a reference antenna (relative, absolute). Relative field calibrations, such as absolute laboratory calibration and traditional calibration methods, are applied at the GIK, and have been used successfully since the early 90s.

Objectives

The reception performance of GPS antennas depends on various factors. In particular, the influence of the location and atmospheric factors that affect the signal propagation, are currently of special scientific interest. Results of un-comprehensive modelling, as well as the utter elimination of these factors in the context of calibration methods, offer inaccurate and erroneous correction values. Therefore, a calibration is made for use by the GIK, and it is characterised as highly robust against these influential factors. The results provided also give a good insight into the range of scatter effects of the calibration results, and help to draw conclusions about the behaviour of the GPS receiving antennas, and the validity of correction values that can be taken in different environments.

The correction values are provided in all standard formats (e.g. ANTEX, Bernese, IGS). The results obtained are also set in relation to the known calibration type values, and possibly with antennas of the same type. They are calibrated by the GIK so that deviations from the determined individual calibration (e.g. from manufacturer's data) can be quantified and statements regarding the relevance of considering the determined correction values (for example) can be made in the context of static post-processing applications. Furthermore, there is a visualisation of the behaviour of the frequency-dependent receiving antenna.

Methods

Calibration column

The project uses a relative field method. The reference represents the reference station KARL. These calibration pillars are stationed about 15 meters away from the GNSS-permanent station, currently individually placed there by the company Geo++. The specific values describe the behaviour of the receiving antenna with high precision. Thus individual absolute calibration values can be determined by the relative method used for each sample. For this purpose, a precise measurement of the height difference shall be calibrated and the reference antenna is necessary.

Information will be collected and evaluated over 24 hours of observations for each test piece and location. In this case, the example per minute is automatically rotated in four different directions (NSEW) whereby correction values can be determined for the entire antenna hemisphere. Raw data are currently recorded at a rate of 60s for each antenna.

There are at least two locations used per antenna from the top (objectives), and can be determined in the range of the scattered calibration results.

In addition, regular monitoring (every 2 years) of the antenna seems to make sense, so that changes over time can be detected. We also recommend an antenna check when there is suspicion of corruption.

Components

Background: reference antenna KARL; foreground: reference height bolts for more precise determination of the level difference

Levelling to the reference antenna KARL

Levelling the test item

Equipment to build and automate rotation (DRB, made at the Geodetic Institute of TU Dresden)

Laboratory room in the observatory; analysis software: WaSoft/Kalib (engineering office Wanninger)

Presentation of the results of selected studies

Future outlook

For minimizing the potential effects, calibration columns are used at the surface of the calibration to form multipath effects. Next, a special material is used which is applied to the smooth surface finish. This so-called absorber material prevents the reflection of electromagnetic signals from the L-band, resulting in the subsequent overlay of an original and reflected GPS signal. It follows an increase in the reliability of the process and higher quality.

References

Landesvermessungsamt Baden-Württemberg, Department of Geodesy in Karlsruhe, Unit 31, Team SAPOS and special calculations
City of Mannheim, Department of Geoinformation and Surveying, Department of Surveying
State Bureau of Surveying and Geospatial Information, Rheinland-Pfalz (Contact: Reichert)

Expenses

The Geodetic Institute offers services in the context of antenna calibration and testing for external clients.

Details upon request

Contact

Dr.-Ing. Michael Mayer

Dipl.-Ing. Andreas Knöpfler

Publications/Presentations/Posters

Mayer, M.
GPS-Antennenkalibrierung.
In: Heck, B./Illner, M. (Hrsg.): GPS 2002: Antennen, Höhenbestimmungen und RTK-Anwendungen. 57. DVW-Seminar, Karlsruhe, 16.-17. September 2002, DVW-Schriftenreihe, Band 44 (2002), Wittwer, S.118-134.

Freiberger Jr., J. / Seitz, K. / Mayer, M. / Nuckelt, A. / Heck, B. / Pereira Krüger, C.
Ein Kalibrierungsverfahren für GPS-Referenzstationsantennen des Bundesstaates Parana/Brasilien.
Geodätische Woche 2004, 12. - 15. Oktober 2004, Stuttgart.

Rozsa, S. / Mayer, M. / Westerhaus, M. / Seitz, K. / Heck, B.
Towards the determination of displacements in the Upper Rhine Graben area using GPS measurements and precise antenna modelling.
Quaternary Science Reviews, (24)2005, S.425-438.

Freiberger Jr., J. / Mayer, M. / Seitz, K. / Heck, B. / Pereira Krueger, C.
Calibracao de antenas GPS em diferentes estacoes.
Boletim de Ciencias Geodesicas, Vol. 11, 2/2005, S. 157-178.

Freiberger Jr., J. / Heck, B. / Pereira Krueger, C. / Mayer, M. / Seitz, K.
Estimacao do centro de fase medio de antennas gps.
Serie em Ciencias Geodesicas, As Ciencias Geodesicas nas Politicas de Desenvolvimento 2005, Vol. 5, S. 36-48.

Knöpfler, A. / Mayer, M. / Nuckelt, A. / Heck, B. / Schmitt, G.
Untersuchungen zum Einfluss von Antennenkalibrierwerten auf die Prozessierung regionaler GPS-Netze.
Universität Karlsruhe, Schriftenreihe des Studiengangs Geodäsie und Geoinformatik, Heft-Nr. 2007/1, 2007.
zur Internetveröffentlichung ...