GNSS Interoperability Enabling the Unknown Future

Transcription

1 GNSS Interoperability Enabling the Unknown Future Tom Stansell Consultant to The Aerospace Corporation CSNC 2014 Nanjing, China The Aerospace Corporation 2014

2 Disclaimer: The views and opinions expressed herein are those of the author and do not necessarily reflect the official policy or position of any agency of the U.S. government. 2 The Aerospace Corporation 2014

3 Introduction The author has been continuously involved in Navigation by Satellite since 1960 more than five decades Experience proves it is impossible to predict GNSS products and services one or two decades in the future ~1963 single channel Transit Receiver 1984 five channel GPS Navigator Impossible then to envision today s GPS products and services Impossible now to envision new GNSS capabilities in The Aerospace Corporation 2014

4 Who Could Have Anticipated GPS in Cell Phones? About a Billion Cell Phone GPS Users Worldwide Sparked by the E911 requirement Use of Location Based Services (LBS) is expanding rapidly Improved by Assisted GPS (A-GPS) Better accuracy Location in seconds Turn-by-turn navigation 4 The Aerospace Corporation 2014

5 Who Could Have Anticipated Precision Agriculture? One to 10 cm accuracy Far better productivity, efficiency, and protection of the environment Enabled, e.g., by MSS signals for the StarFire or OmniSTAR Services Illustrations courtesy of John Deere Automatic Steering Automatic Spray Control 5 The Aerospace Corporation 2014

6 We Can t Envision the Future But We Can Enable It By making good decisions now about GNSS Interoperability Past example: The first GPS satellite was launched in 1978 The types of signals transmitted then are in use today, 36 years later Much time is needed to build a complete constellation of satellites Once in place, it is a long and difficult process to make signal changes in the entire constellation while ensuring backward compatibility GPS is a good example long lasting signal designs, backward compatibility with fielded user equipment, reliable and trustworthy service The GPS Directorate expects the same for Modernized GPS signals L2C the 2 nd open GPS signal after L1 C/A will be available from every new GPS satellite, starting with the first IIR-M launch (26/09/2005) L5 the 3 rd open GPS signal will be available from every new GPS satellite, starting with the first IIF launch (27/05/2010) L1C the 4 th open GPS signal will be available from every new GPS satellite, starting with the first GPS III launch 6 The Aerospace Corporation 2014

7 Enabling the GNSS Future We are at an important point in GNSS history where new and modernized GPS signals have been defined and are being launched The new signals were designed to enable future products and services Similarly, the Galileo program has defined its signals, available now from four satellites, with more expected in 2014 GPS and Galileo worked closely to assure excellent interoperability of GPS L1C and Galileo E1 OS as well as GPS L5 and Galileo E5a signals The default dual-frequency GNSS signals will be centered at and MHz, with matched signal spectra in each of the two bands All modernized GPS and Galileo signals have been defined, documented, and are being or will be transmitted Multi-GNSS receivers for GPS and Galileo are available and in use today The first QZSS satellite also is in use and it is was designed for maximum interoperability with all GPS plus the Galileo E6 signals The urgent question now is about interoperability of signals from BeiDou, GLONASS, and IRNSS 7 The Aerospace Corporation 2014

8 The Goal of Interoperability Ideal interoperability allows navigation with one signal each from four or more systems with no additional receiver cost or complexity Success requires signal designers to think globally while also satisfying national interests Interoperable = Better Together than Separate 8 The Aerospace Corporation 2014

9 GNSS Spectra Will BeiDou deploy a global constellation with Phase III signals? Will GLONASS provide interoperable CDMA signals? What will IRNSS do? 9 The Aerospace Corporation 2014

10 Key Interoperability Issues Increase of noise floor in GNSS receivers More signals from more satellites in the same band Common or offset center frequencies Frequency diversity vs. frequency commonality How many global systems should share spectrum? Common signal spectra in each band or not? Can minimum satellite elevation tracking limits be raised? Reduces Multipath error as well as Ionospheric and Tropospheric refraction error enabled by more satellites International clock and geodesy references, or not ICAO acceptance of new signals for international aviation Transmitter bandwidth to enable better multipath mitigation and code measurement accuracy Another common open signal for wide area, high precision, phasebased navigation Potential to use existing or planned spare capacity in open service or SBAS messages to increase multi-gnss interoperability 10 The Aerospace Corporation 2014

11 Tri-Lane Phase Navigation is Near Over the next decade there will be a dramatic improvement in potential wide area GNSS accuracy Providing rapid access to reliable 10 cm navigation accuracy From wide area differential code and phase corrections Precision agriculture will be the first large scale user Enabled by having three GNSS frequencies Two will be MHz and MHz GPS L1/L5, Galileo E1/E5a, BeiDou Phase III B1-c/B2-a What middle frequency or frequencies will be used? 11 The Aerospace Corporation 2014

12 Interoperability Questions and the Author s Answers At the ICG WG-A interoperability meeting in Honolulu on 26 April 2013, the author posed 30 questions to 12 industry leaders The Honolulu meeting agenda is on the next slide Their responses are available at: The following are the author s own answers to a subset of the questions based on professional experience and industry interaction The three slides following the agenda number all 30 questions and the author s answers correspond to the numbered list Enabling future GNSS products, services, and applications is enhanced by providing the best possible signal interoperability Disclaimer: The views and opinions expressed herein are those of the author and do not necessarily reflect the official policy or position of any agency of the U.S. government. 12 The Aerospace Corporation 2014

13 Agenda from 26 April 2013 Interoperability Meeting Num HST Dur. Topic Organization Speaker How 1a 9:00 0:10 Welcome and Introduction Co-Chair WG-A Dave Turner P 1b 9:10 0:05 Welcome and Introduction Co-Chair WG-A Sergey Revnivykh P 2 9:15 0:05 Welcome and Introduction Nat. Time Service Ctr., CAS Xiaochun Lu P 3 9:20 0:10 Framing the Presentations Stansell Consulting Tom Stansell P 4 9:30 0:20 Quasi-Coherent Processing Tsinghua University Zheng Yao P 5 9:50 0:25 Certified Avionics #1 MITRE/FAA Chris Hegarty P 6 10:15 0:25 Certified Avionics #2 Rockwell Collins Joseph & Wichgers G 7 10:40 0:25 High Precision #1 Septentrio Peter Grognard G 11:05 0:15 Break :20 0:25 High Precision #2 Trimble Stewart Riley G 9 11:45 0:25 High Precision #3 John Deere Ron Hatch P 10 12:10 0:25 High/Medium Precision Hemisphere GPS Brad Badke P 12:35 1:15 Lunch :50 0:25 GNSS Past, Present & Future MITRE John Betz P 12 14:15 0:25 High Precision #4 Topcon Ivan De Federico P 13 14:40 0:10 Consumer Applications #1 CSR plc Greg Turetzky G 14 14:50 0:25 Consumer Applications #2 ST Microelectronics Philip Mattos S 15 15:15 0:25 Consumer Applications #3 Broadcom Charlie Abraham P 16 15:40 0:25 Consumer Applications #4 Qualcomm Doug Rowitch P 16:05 0:15 Break N/A Space Science Applications JPL Dr. Larry Young S 18 16:20 0:10 Summary Stansell Consulting Tom Stansell P 19 16:30 0:10 Summary Nat. Time Service Ctr., CAS Xiaochun Lu P 20 a 16:40 0:10 Summary and Conclusion Co-Chair WG-A Sergey Revnivykh P 20 b 16:50 0:15 Summary and Conclusion Co-Chair WG-A Dave Turner P 17:05 End How P=Present, G=GoToMeeting, S=Submitted 13 The Aerospace Corporation 2014

14 Interoperability Questions (1 of 3) 1. What types of applications do your receivers (or receiver designs) support? 2. Do you see a threat to GNSS receivers due to many more GNSS signals centered at MHz? 3. Whether you see a threat or not, do you prefer all new CDMA signals at L1 to be centered at MHz or have some of them elsewhere, e.g., at 1602 MHz? 4. Given that most GNSS providers plan to transmit a modernized signal at MHz, what is your long term perspective on whether you will continue to use C/A? Why? How? 5. Once there are a large number of good CDMA signals, will there be continuing commercial interest in FDMA signals? Why or Why Not? 6. Do you prefer signals in different L1 frequency bands for interference mitigation rather than at one center frequency for interoperability? Why? 7. If a satellite s signals do not meet quality standards, what should happen (see list in slide)? 8. To assure only good signals, should GNSS providers agree on minimum international signal quality standards and agree to provide only signals meeting the standard? 9. Given that L5/E5a will be transmitted by most GNSS providers, do you intend to use the E5b signal? If so, for what purpose? 10. For your applications, are small satellite frequency steps a problem? 11. If so, what interval between frequency steps and what delta-f magnitude would be excessive? 12. Assuming signal quality is acceptable from every provider, would you limit the number of signals used by provider or by other criteria? What criteria? 13. Is having more signals inherently better or do you think there should be a limit? 14 The Aerospace Corporation 2014

15 Interoperability Questions (2 of 3) 14. Will the marketplace force you to make use of every available signal? 15. For best interoperability, how important is a common center frequency? How important is a common signal spectrum (PSD)? 16. Will you provide tri-lane capability in the future? Why? 17. If so, do you prefer a common middle frequency or the combined use of L2 (1227.6), B3 ( ), and E6 ( ) if B3 and E6 open access is available 18. Would you prefer a common open signal in S Band? In C Band? Why? 19. Does a wider satellite transmitter bandwidth help with multipath mitigation? 20. What minimum transmitter bandwidth would you recommend for future GNSS signals in order to achieve optimum code precision measurements? 21. Would you recommend GNSS or SBAS services provide interoperability parameters (see list in slide)? 22. Should they be provided by other means so as not to compromise TTFF or other navigation capabilities? 23. For your applications and for each signal, what amount of drift between code and carrier over what time frame would be excessive? 24. For your applications and for two or more signals in different frequency bands, e.g., L1 and L5 (when scaled properly), what amount of relative drift in code and carrier between the signals would be excessive? 25. Should the international community strive to protect all GNSS signal bands from terrestrial signal interference? 15 The Aerospace Corporation 2014

16 Interoperability Questions (3 of 3) 26. Do the current differences (~10 cm) in Geodesy pose a problem for your users? Why or why not? 27. If geodesy differences are a problem, what is the preferred method of compensation (see list on slide)? 28. Do you want each system to cross reference the other s time (e.g., with a GGTO type of message) or compare itself to a common international GNSS ensemble time? To what precision? 29. Will your future receivers calculate a time offset between systems based on signal measurements or use only external time offset data? 30. What is the preferred method of receiving time offsets: Satellite messages, Internet messages, or internally calculated? 16 The Aerospace Corporation 2014

17 Author s Answers (1 of 11) Q02, Do you see a threat to GNSS receivers due to many more GNSS signals centered at MHz? For older generation receivers not able to take advantage of the new signals, there may be a noise floor problem (TBD). (This will not be a problem indoors or in urban canyons because many potentially interfering signals are blocked.) However, modernized receivers able to use all available signals at one frequency will achieve cost savings and provide significantly improved performance, so there is a net benefit to more signals centered at MHz. Q03, Whether you see a threat or not, do you prefer all new CDMA signals at L1 to be centered at MHz or have some of them elsewhere, e.g., at 1602 MHz? If they have identically the same center frequency and identically the same spectrum as L1C and E1 OS, that would be the most interoperable result. If these conditions are not met, it may be better to locate the signals elsewhere. 17 The Aerospace Corporation 2014

18 Author s Answers (2 of 11) Q04, Given that most GNSS providers plan to transmit a modernized signal at MHz, what is your long term perspective on whether you will continue to use C/A? Why? How? C/A may continue to be used for rapid acquisition purposes over the next two decades. However, the significant advantages of new signals will cause them to be adopted for all applications. The key advantages of L1C are listed in a backup slide and explained in the following papers: Stansell, T., Hudnut, K.W., Keegan, R.G., "GPS L1C: Enhanced Performance, Receiver Design Suggestions, and Key Contributions," Proceedings of the 23rd International Technical Meeting of The Satellite Division of the Institute of Navigation (ION GNSS 2010), Portland, OR, September 2010, pp Stansell, Thomas A., Hudnut, Kenneth W., Keegan, Richard.G., L1C Signal Performance and Receiver Design, GPS World, April 2011, pp & 41, Q05, Once there are a large number of good CDMA signals, will there be continuing commercial interest in FDMA signals? Why or Why Not? There is near consensus the answer is no. It is more difficult to receive, calibrate, and use FDMA signals. With lots of interoperable CDMA signals, the use of FDMA signals will decline rapidly. 18 The Aerospace Corporation 2014

19 Author s Answers (3 of 11) Q06, Do you prefer signals in different L1 frequency bands for interference mitigation rather than at one center frequency for interoperability? Why? Interoperability is best with one center frequency. However, if GPS, Galileo, and BeiDou each provide full global constellations having signals with a common center frequency and spectrum, there is little need for more, so then it would be desirable to have a frequency offset for additional signals. Q07, If a satellite s signals do not meet quality standards, what should happen? The spreading code for that signal should become a nonstandard code, the health indication should be set unhealthy, and, optimally, the signal power should be reduced substantially. Q08, To assure only good signals, should GNSS Providers agree on minimum international signal quality standards and agree to provide only signals meeting the standard? Yes 19 The Aerospace Corporation 2014

20 Author s Answers (4 of 11) Q09, Given that L5/E5a will be transmitted by most GNSS providers, do you intend to use the E5b signal? If so, for what purpose? A large majority of GNSS receivers never would have used E5b. Because E5b apparently will not be used for integrity messaging, it may have little ongoing value. In the U.S., the RTCA and the FAA have no plan to use E5b for aviation A few may use the E5a/b 50 MHz bandwidth for ultra high resolution code measurements, although the same precision can be achieved with any signal having a 50 MHz transmitter bandwidth. Because the ionosphere-free combination of L1 and L5 pseudoranges is (L1*2.26 L5*1.26), the benefit of extra L5 precision is deweighted by 56% Q10, For your applications, are small satellite frequency steps a problem? Yes, small frequency steps are a major problem, they should be eliminated, and they should be a key quality indicator for any system 20 The Aerospace Corporation 2014

21 Author s Answers (5 of 11) Q12, Assuming signal quality is acceptable from every provider, would you limit the number of signals used by provider or by other criteria? What criteria? If signal quality is the same from all providers or if differential corrections equalize the signal quality, there is no reason to limit by provider the number of signals used. All are the same. Therefore, the particular signals used should be based only on navigation availability, accuracy, and integrity considerations. Q13, Is having more signals (satellites) inherently better or do you think there should be a limit? More satellites is inherently better because it enhances performance in difficult situations, e.g., urban canyons, etc. Any limit should be based on compatibility, i.e., where the benefit of more satellites is undermined by an increase in noise floor. See answer to Q06. Q14, Will the marketplace force you to make use of every available signal? Yes. Whether or not more is better, customers will demand all. 21 The Aerospace Corporation 2014

22 Author s Answers (6 of 11) Q15, For best interoperability, how important is a common center frequency? How important is a common signal spectrum (PSD)? We cannot predict the future need for very low cost products or for very high precision products. Although receivers can be designed to use different center frequencies and different signal spectra, it is clear the best performance at the lowest cost is enabled by the same center frequency and signal spectrum. Such choices best enable the unknown GNSS future. Q16, Will you provide tri-lane capability in the future? Why? Three GNSS frequencies enable wide area navigation and positioning with 10 cm accuracy within about two minutes of tracking, based on ionosphere-free carrier phase measurements. This is achieved with two frequencies now, but the time to first fix is about 15 times longer. It is easy to predict that this capability will be adopted quickly for precision agriculture, but it has promise for lanekeeping car navigation and personal survey of boundaries. 22 The Aerospace Corporation 2014

23 Author s Answers (7 of 11) Q17, If so, do you prefer a common middle frequency or the combined use of L2 (1227.6), B3 ( ), and E6 ( ) if B3 and E6 open access is available Studies have shown that a middle frequency halfway between L5 and L1 would be best, and of the available signals E6 best meets that criterion. For ideal interoperability it would be desirable to have one common middle frequency from all GNSS. Because code and carrier measurements are what is needed, perhaps commercial encryption could be limited to message data only. Better yet, to enable widespread use of this powerful capability, the common middle frequency signal should be open and not encrypted. The benefit to society in terms of safety, commerce, and environmental protection should not be overlooked or underestimated. Q19, Does a wider satellite transmitter bandwidth help with multipath mitigation? Yes. Transmitter bandwidth is the most important factor in enabling high precision pseudorange measurements and multipath mitigation. 23 The Aerospace Corporation 2014

24 Author s Answers (8 of 11) Q20, What minimum transmitter bandwidth would you recommend for future GNSS signals in order to achieve optimum code precision measurements? Not only should the transmitter bandwidth be wide, at least ±18 MHz to the 1 db points, but it should have very low bandpass ripple and as linear a phase versus frequency characteristic as possible. Q21, Would you recommend GNSS or SBAS services provide interoperability parameters? Q22, Should they be provided by other means so as not to compromise TTFF or other navigation capabilities? QZSS should provide as few parameters as possible. Data bits are a precious resource and should be used primarily to minimize time to first fix. GNSS should provide system clock offsets, preferably to a common GNSS ensemble clock, but nothing else. Other interoperability parameters are better provided by communication services, including the Internet. 24 The Aerospace Corporation 2014

25 Author s Answers (9 of 11) Q23, For your applications and for each signal, what amount of drift between code and carrier over what time frame would be excessive? Q24, For your applications and for two or more signals in different frequency bands, e.g., L1 and L5 (when scaled properly), what amount of relative drift in code and carrier between the signals would be excessive? The ionosphere causes a separation and a drift between the received code and carrier. However, to use multiple signals to solve for or to eliminate the ionospheric refraction effects it is important for the transmitted signals to have as little drift as possible between code and carrier on each signal as well as between signals. Since the codes can be clocked synchronously and the carrier signals can be phase locked, such drifts can be and should be undetectable. Q25, Should the international community strive to protect all GNSS signal bands from terrestrial signal interference? Yes! Not only should nations protect their own GNSS signal bands but also those of other nations. Think Globally while also satisfying National interests. 25 The Aerospace Corporation 2014

26 Author s Answers (10 of 11) Q26, Do the current differences (~10 cm) in Geodesy pose a problem for your users? Why or why not? Q27, If geodesy differences are a problem, what is the preferred method of compensation? Such small differences are negligible for consumer and most commercial applications. Real-time high precision applications use differential corrections which eliminate geodetic differences. Non real time applications, e.g., point positioning, can obtain necessary corrections from the Internet or other sources. Over time, GNS systems will continue to converge to a common geodesy so corrections eventually will become unnecessary. 26 The Aerospace Corporation 2014

27 Author s Answers (11 of 11) Q28, Do you want each system to cross reference the other s time (e.g., with a GGTO type of message) or compare itself to a common international GNSS ensemble time? To what precision? It is important for GNSS users to cross reference system clocks. To enable a quick solution, the information is most valuable when a receiver is first turned on. Unfortunately, GGTO messages will have a low priority and be transmitted only occasionally. Therefore it is desirable to minimize the required number of bits by referencing each system to a GNSS ensemble time rather than individually to each other. The accuracy and precision should be 3 ns or better. Q29, Will your future receivers calculate a time offset between systems based on signal measurements or use only external time offset data? Best accuracy is obtained by calculating system time differences within the receiver, so that should be done regardless of corrections. Q30, What is the preferred method of receiving time offsets: Satellite messages, Internet messages, or internally calculated? Internally calculated is best, but GNSS messages also are required. 27 The Aerospace Corporation 2014

28 Summary We cannot predict the GNSS future, but we can enable it by implementing excellent signal designs with optimal interoperability Success requires signal designers to think globally while also satisfying national interests. The GNSS world is anxious to know exactly what signals BeiDou, GLONASS, and IRNSS will transmit, especially on a global basis. This presentation has offered the author s view on interoperability. It seems the use of GNSS is entering a new golden era with many more GNSS satellites and interoperable signals, the benefits being: Much better availability, especially in difficult environments Better accuracy through improved GDOP, the ability to ignore low elevation signals, and by including more measurements in each solution Enabling and improving the use of RAIM and ARAIM integrity monitoring Enabling robust, wide-area, tri-lane navigation with 10 cm accuracy New products, services, and commerce which cannot yet be predicted Helping us all realize we are connected and neighbors in this one World 28 The Aerospace Corporation 2014

29 Backup Slides 29 The Aerospace Corporation 2014

30 L1C Innovations in order of Judged Value Seventy-five percent of power in the dataless or pilot component Use of low density parity check (LDPC) for forward error correction (FEC) CNAV-2 message structure Time of Interval (TOI) message with BCH encoding Pilot overlay code to frame the message and improve correlation properties The spreading and overlay code designs The MBOC (TMBOC) waveform Symbol interleaving 30 The Aerospace Corporation 2014