Titelmaster. Antenna properties

Transcription

1 Titelmaster On the Accuracy of Absolute GNSS Antenna Calibration in Context of Near Field Effects Barbara Görres, Philipp Zeimetz, Heiner Kuhlmann Institute of Geodesy and Geoinformation University of Bonn Antenna properties E PCO ARP 2 1.: PCO: phase centre offset (up to 15 cm) the exact position of the phase centre is unknown Antenna Effect: s ARP = r + PCO e0 + ε

2 Antenna properties 3 1.: PCO: phase centre offset (up to 15 cm) the exact position of the phase centre is unknown 2.: PCV: phase centre variations (up to 2 cm) the phase centre is not a point in a mathematical sense Antenna Effect: s ARP = r + PCO e 0 + PCV, ( α β ) + ε Examples 4 Leica AX 1202 k.a. Leica AT 504 Trimble Zephyr Geodetic

3 Varying Signal Direction 5 Calibration Setup 6 NetworkAnalyser (NWA): measurement of the phase variations 2-axis-Positioner: rotation of the GNSS-antenna in order to change the direction of the incoming signal

4 Anechoic chamber Bonn absorber walkable absorber Grundriss eigene Kammer test area transmitter 7 Partnership: University of Bonn (IGG) & Bezirksregierung Köln Geobasis NRW (surveying and mapping agency of North-Rhine Westphalia) Anechoic chamber Bonn absorber walkable absorber Grundriss eigene Kammer transmitter 8 end of 2007: Agreement between Uni Bonn and LVermA Aug. & Sep. 2008: Construction phase Oct. Dec. 2008: Testing phase ~ January 2009: Start of operation

5 Not so very long ago,... 9 AMK Bonn 10

6 GNSS-Frequencies calibration of a wide frequency spectrum: GHz Leica AX 1202GG Antenna 11 λ ~ cm λ ~ cm λ ~ cm GNSS-Frequencies calibration of a wide frequency spectrum: GHz Leica AX 1202 Antenna 12 λ ~ cm λ ~ cm λ ~ cm

7 Accuracy of Calibration Important aspects concerning accuracy: Accuracy of the NetworkAnalyser Noise of the calibration setup Effects of temperature Effects of varying signal level Repeatability of the calibration results (Precision) Comparison with field calibration procedures (e.g. Geo++) Near-field effects, Sensitivity of the Setup Repeatability differences between two measurements with identical setups: with different setups: 14 Leica AX 1202GG antenna Typical differences for identical setups: +/- 0.3mm Possible differences in the setup are: cable placement mounting of the antenna orientation of the transmitter

8 Near-Field-Effects: one nice Example 15 Cable effect and Repeatability Effect of cable placement Repeatability (different setups) max. diff.: 1.0 mm 16 min. diff.: -0.4 mm Observation: very small changes in the antenna near field produces significant variations in the antenna pattern Near fiels effects limits the accuracy of calibration AND GNSSobservations

9 Effect of cable placement: L1 Up [mm] N,E [mm] 17 Zero-baseline L1 5 Elevation Effect of cable placement: L3 Up [mm] N,E [mm] 18 Zero-baseline L3 5 Elevation

10 A very old idea... Groundplane Foto: Erik Schönemann, Darmstadt 19 The calibration of near field effects leads to an old solution Comparison with Field Procedure (here: Geo++) Graphic source: Geo++ vs. 20 Field vs. Laboratory procedure: Differences: source signal (satellite vs. helical beam) signal level (less noise in case of laboratory procedure) signal structure (no code in case of laboratory test signal) environment (Multipath, Near Field) completely independent procedures absolute accuracy?

11 Comparison with Field Procedure (here: Geo++) Example: Trimble Zephyr Geodetic antenna (L1) Elevation Max. diff ± 1.8 mm ± 1.0 mm ± 0.5 mm 21 deviations are only slightly higher than the effects of the near field near field effects seems to be the main reason for the deviations Both procedures correspond well Remark: For antennas without a good shielding, the max. deviations increase Summary Advantages of the laboratory procedure: - calibration of a wide frequency spectrum (no satellites necessary) - the calibration procedure is very efficient (40-60 min) - constant environments can be achieved The main disadvantage: - the test signal differs from the real GNSS-signal 22 Laboratory and field procedures complement each other Main problem: - Near-field effects are currently the main problem

12 Thank You! 23