GNSS Signal Observations - Stanford and DLR

Transcription

1 GNSS Signal Observations - Stanford and DLR Christoph Günther, Sherman Lo Contributors: Dennis Akos, Alan Chen, Johann Furthner, Grace Gao, Sebastian Graf, David de Lorenzo, Oliver Montenbruck, Alexander Steingass, Christian Weber Institute for Communications and Navigation Page 1

2 Four GNSS Programs under Development and Evolution GPS Modernization Galileo II-RM with L2C signals, II-F with L5 WAAS broadcast of L5 wideband civil signal Block III with L1C OS and SoL on L1, E5a, E5b Additional signals on L1 and E6 COMPASS GLONASS the great unknown B1, B1-2, B2, B3 thinking heavily of including a CDMA signal compatible to other systems Institute for Communications and Navigation Page 2

3 GNSS Signals, Institute for Communications and Navigation Page 3

4 Signals around 2015 Institute for Communications and Navigation Page 4

5 Benefits of the New Signals Interoperability Noise performance Cramer Rao bound BW [MHz] RX Power [dbw] CRB [cm] Galileo E1bc Galileo E1bc Galileo E1bc Galileo E Galileo E Galileo E5a Galileo E6bc GPS L1C/A GPS L1C/A GPS L1C/A GPS L2C GPS L5P Dual and triple frequency linear combinations of codes and carriers undifferenced Integrity Institute for Communications and Navigation Page 5

6 Unrealized Potential Capability Near & Mid Term Compatibility Decisions GNSS as it stands today GNSS Evolved cooperatively GNSS Evolved individually GNSS Evolved in Conflict Now 5 Years 10 Years Without careful thought, study, and design of the signals, we can potentially be leaving a lot on the table Performance, Signal design, Interoperability all need to be well studied With apologies to the US National PNT Architecture Study Group Institute for Communications and Navigation Page 6

7 Signal Observations Needed to understand real world performance for the verification and validation of systems, e.g. Galileo signals, clocks, for understanding the situation with Beidou/Compass signal compatibility/interoperability for analyzing potential threats interference, signal deformations for studying monitoring and augmentation algorithms LAAS/GBAS, RRAIM, Rapid response and diagnosis of system faults Institute for Communications and Navigation Page 7

8 Analysis Capabilities Data sampling on reflector antennas from 1.8 m to 45 m signal quality and performance Data sampling on hemispherical antennas interference analysis Experimental networks for monitoring and augmentation test of augmentation systems satellite clock analysis Institute for Communications and Navigation Page 8

9 Data sampling with large antennas Stanford DLR Courtesy: SRI Institute for Communications and Navigation Page 9

10 Measurement Setup - Cassegrain hyperbolic sub-reflector 4 m parabolic reflector 30 m antenna gain 52 db: feed sub-reflector LNA1 LNA2 ~60 [db] PC for Analysis approx samples/s Rhode & Schwarz FSIQ26 Agilent E4440 PSA Institute for Communications and Navigation Page 10

11 Spectrum Galileo L1 BOC(1,1) BOC(15,2.5) MHz January 24 th, 2006 Institute for Communications and Navigation Page 11

12 Measured IQ-Diagram Measured ( ) Simulated PRS only transition 1. the PRS transition with I=0 has an eye 2. the two exterior groups of PRS transitions are tilted wrto the central one 3. the orbits of the OS(+PRS) transitions are different Institute for Communications and Navigation Page 12

13 Eye Diagramm OS subcarrier transition OS code chip boundary [=µs] PRS subcarrier transition PRS code chip boundary Institute for Communications and Navigation Page 13

14 Satellite TX Model Details unknown primary candidate for imperfections: power amplifier Institute for Communications and Navigation Page 14

15 Potential Explanations asymmetry in the spectrum eye opening in the IQdiagram tilt in the IQdiagram orbits in the IQdiagram non-linearity never only for extreme values yes no dispersion yes yes never (rotation only) yes Institute for Communications and Navigation Page 15

16 Non-linear Power Amplifier Saleh Model (25MHz), α=0.11 TWTA Model (25MHz) Institute for Communications and Navigation Page 16

17 Dispersive Behavior Find an equivalent Wiener Filter amplitude phase and group delay Institute for Communications and Navigation Page 17

18 Simulation Result IQ Diagram Measured ( ) Simulated using the filter and non-linear Institute for Communications and Navigation Page 18

19 Galileo E5 Signal AltBOC(15,10), or two BPSK(10) with center frequencies separated by 2x15x1.023 MHz GIOVE-A, , Weilheim E5: MHz E5a: MHz E5b: MHz Institute for Communications and Navigation Page 19

20 Signal B2 and B3 of Beidou M Satellite Institute for Communications and Navigation Page 20

21 Signal B1 of Beidou M Satellite 14. May May 2007 Institute for Communications and Navigation Page 21

22 Stanford GNSS Monitor Station Utilizes a 1.8 m parabolic antenna with automated tracking/control On roof of GPS Laboratory allowing for on demand access Institute for Communications and Navigation Page 22

23 Stanford GNSS Monitor Station Institute for Communications and Navigation Page 23

24 Data Collection GIOVE-A E1-L1-E2 BOC(1,1) BOC(15,2.5) OR Dish allowed us to see GIOVE-A signal when transmission was initialized Same set up used in SRI dish Vector Signal Analyzer used to capture data from transmission Institute for Communications and Navigation Page 24

25 Portable Ground Station set up Institute for Communications and Navigation Page 25

26 Observations of New GNSSs Galileo GIOVE-A Decoded E1 BOC(1,1) Signal on Jan 11, 2006 Later discovered the code bits to be 98% correct (using high gain dish data) E5a + E5b codes were determined solely using data from Compass Beidou 2B Used data collected from SGMS to determine codes on multiple frequencies Sole source of data used for code determination WAAS L5, Modernized GLONASS, GPS L2C also observed using SGMS Institute for Communications and Navigation Page 26

27 L2 Data From IIR-M GPS SV (SVN53/PRN17) -120 Frequency Domain x Time Domain of the Baseband Signal magnitude amplitude frequency (MHz) Captured a 200 msecond record of 36 MHz bandwidth about the MHz L1 carrier frequency The first lobe of the L2C code and primary lobes of the L2 P(Y) code spectrum are clearly visible A significant RFI component is clearly visible in either frequency but not in the time domain representation likely a result of local radar activity -4 time (msec) Institute for Communications and Navigation Page 27

28 L5 WAAS Test Data from the Galaxy 15 Geostationary Satellite Frequency Domain Time Domain of the Baseband Signal SNS (Salinas) MHz magnitude amplitude frequency (MHz) time (msec) Captured a 200 msecond record of 36 MHz bandwidth about the MHz L5 carrier frequency The primary lobe of the MHz L5 PRN code spectrum is apparent Multiple significant RFI components are visible in both the frequency and time domain representations Pulsed interferences from the inband DMEs in the surrounding area Expand time domain plots to confirm/verify this Institute for Communications and Navigation Page 28

29 Expanded Time Domain View of L5 Band WAAS Data 0.06 Time Domain of the Baseband Signal Zoomed View 0.06 Time Domain of the Baseband Signal amplitude time (msec) The GNSS L5 signals will experience pulsed interference as a result of the inband DME transmissions Data collections at Stanford show the presence of a number of different DME, identifiable by their underlying frequencies, with varying signal strength within the collected data amplitude time (msec) L5 (E5A/B) GNSS receivers will utilize pulse blanking to minimize the impact of the inband DME broadcasts Stanford GNSS monitor station utilizes extended dynamic range to receive/process both DME and GNSS L5 signals Institute for Communications and Navigation Page 29

30 Not Just a GPS Problem! E5a ( MHz) PSD E5b ( MHz) PSD Institute for Communications and Navigation Page 30

31 Mobile Interference Measurement Setup Spectral analysis E5 : MHz (including L2) E6 : MHz E1-L1-E2 : MHz Potential interferer Narrowband (single carriers etc.) Broadband (DVB-T, UMTS etc.) Very broadband (UWB, etc.) + pulsed interference Planned campaign next year near Frankfurt with up to 48 DME Institute for Communications and Navigation Page 31

32 Measurement Equipment Spectrum Vector analyzer R&S FSH6 R&S FSP3 Agilent E4443A GPS-Receivers Ashtech G12 Leica GSP1200 Antennas Hemispherical GPS-patch directional Institute for Communications and Navigation Page 32

33 Examples of interference patterns E6: MHz L1-Band L1: MHz kt=-174 [dbm-hz] E6-Band Institute for Communications and Navigation Page 33

34 Wideband Interference in L1 7 MHz wide dbm average Power Interference-to-Signal-Ratio (ISR) 42.6 db GPS carrier Galileo BOC(1,1) main lobes Institute for Communications and Navigation Page 34

35 Jammer Mitigation with Adaptive Beamforming Combination of ESPRIT + constrained minimum variance (LCMV) beamforming Antenna array processing Time of interference occurrence Estimated C/N0 value Satellite signal has been detected and enhanced Beamforming gain SW Receiver tracks GPS L1 signal with acceptable C/N0 Single antenna (common case) Estimated C/N0 value SW receiver loses tracking C/N0 tracking threshold Interferers have been mitigated by producing spatial nulls 2X2 Antenna array Institute for Communications and Navigation Page 35

36 EVNet Architecture GNSS Sensor Station 1 GNSS Sensor Station 2... GNSS Sensor Station N EVnet Administrator EVnet Operator C&C* Central Processing and Control Facility User Configuration Processing Centre C&C RT Data Broadcaster Data Archive RT* Data Archive Data C&C External Processing Facility 1... External Processing Facility M C&C RT Data User Component 1 User Component 2... User Component L Institute for Communications and Navigation Page 36

37 EVNet Sensor Station Network Kiruna Stanford University Quebec Toulouse Canaries Island Africa Neustrelitz OP Japan Bandung Brazil Galileo Sensor stations Operational EVnet-Stations intended Stations Institute for Communications and Navigation Page 37

38 Conclusion New signals in new bands: GPS, Galileo, GLONASS, Compass New opportunities for improved navigation performance Analysis of the performance of the satellites: signals characteristics, clock stability, biases, Understanding and mitigating the interference situation in the new bands and even L1 Stanford and DLR have developed and are developing tools and methods for these tasks, with a particular focus on aeronautics We have started a very promising cooperation in this and other fields relevant to aeronautical navigation Institute for Communications and Navigation Page 38