“The general advantages of multi-GNSS are that one can obtain more accurate and more robust parameter solutions”

Professor Peter Teunissen shares his views on a wide range of subjects related to GNSS

Professor Peter Teunissen has recently received Johannes Kepler Award from ION for his influential and groundbreaking contributions to the algorithmic foundations of satellite navigation and sustained dedication to the global education of next generation of navigation engineers (https://www.ion.org/ awards/2019- Kepler.cfm ). He has been recognized by the national newspaper ‘The Australian’ as the world-best in the field of ‘radar, positioning and navigation’ (https://specialreports.theaustralian.com. au/1540291/top-of- the-world/)

He is currently a Professor of Satellite Navigation at Delft University of Technology, The Netherlands and Curtin University, Australia. Prof. Teunissen invented the Least Squares Ambiguity Decorrelation Adjustment (LAMBDA) method, the worldwide standard for ambiguity resolution, which revolutionized high precision GNSS positioning capabilities. His findings are particularly important for multi-GNSS processing, which require a proper understanding of individual system characteristics and their respective contributions to achieve navigation solutions of the highest precision and integrity. He holds several honorary professorships and fellowships of numerous international organizations, including Australia’s prestigious Federation Fellowship of the Australian Research Council. He has published over 300 papers, seven books, is co-editor and author of the Handbook of Global Navigation Satellite Systems, and is a member of 13 editorial boards. He is a regular contributor to ION and ION programs. He is a Fellow of the ION, the RIN and the Royal Netherlands Academy of Sciences.

Would you like to share your present research works and research priorities with our readers?

Recently I developed and introduced a new GLONASS FDMA model with my colleague Dr Amir Khodabandeh from the University of Melbourne. One feature of the model is that it guarantees, independent of the actual satellite channel number entries, the integer- estimability of the GLONASS ambiguities. This means that existing methods of integer ambiguity resolution can now be directly applied and that the GLONASS FDMA model can thus now be used and combined in a standard way with all other CDMA-based GNSS systems.

Next to the GNSS-modelling, we are also working on some of the more theoretical integrity challenges that exist. With the varying and challenging environments in which GNSS often has to operate, there is a multitude of different threat models that need to be considered simultaneously. This therefore requires a multivariate and collaborative integrity approach to tackle the fundamental complexities that are brought forward by the high dimensionality and dynamics of the problem.

You have invented the Least Squares Ambiguity Decorrelation Adjustment (LAMBDA) method. Could you explain the significance of this invention?

GNSS receivers can measure fractions of the signal waveforms that they receive from each GNSS satellite with millimetre precision, but they cannot measure the total number of signal waveforms that exist in each of the distances between receiver and the satellites. The LAMBDA method is a statistically-based algorithm that helps the receiver to determine these whole number of waveforms. As a result of the combination of the very precisely GNSS- measured fractions of waveforms and the LAMBDA-determined whole number of waveforms, each of the distances from receiver to the satellites can be very precisely determined and thus used for a very-precise positioning and/ or navigation of the receiver location.

This is a world of multi-GNSS systems. Having worked on joint use of new GNSSs and also in setting up multi-GNSS receiver test beds, what advantages do you see about this scenario especially in the precise positioning?

The general advantages of multi-GNSS are that one can obtain more accurate and more robust parameter solutions. In the positioning domain this translates into faster executions of PPP and PPP-RTK due to the shorter convergence times and for RTK to the ability of using longer baseline lengths for which instantaneous ambiguity resolution is still possible.

Multi-GNSS and the availability of new signals with higher power and better tracking performance also improves positioning in adverse environments. With more satellites available and the resulting stronger receiversatellite geometry, one can work with higher cut-off elevations and thus be less sensitive to (low-elevation) multipath and the masking that happens in urban canyons.

The increase in number of satellites and signals also improves the spatial and temporal sampling of the atmosphere, thus offering opportunities for more precise ionospheric modelling, which in its turn is beneficial again for, for instance, long baseline RTK.

Many countries plan GNSS systems primarily because of defence and security needs. Do you think that it would lead to ‘more the merrier situation” or there has to be a limit?

Yes, as a researcher and a user, I would love to see more satellites and signals. Not so much because of the accuracy improvements that this will bring, but more because of the integrity potential and improved atmosphere sensing. After all, if you think of the 1-over-k rule, then at a certain point the gain in accuracy will become marginal, i.e. if k gets larger, 1-over-k gets smaller, but the reduction or gain gets smaller the larger k is.

The real advantages of more satellites and signals lie in the fact that the redundancy in satellites and signals would help to further robustify parameter solutions and thus improve integrity, and that the improved atmospheric sampling would allow for a better ionospheric and tropospheric modelling.

With increasing dependence on GNSS, how do you perceive the threats like interference, jamming and spoofing?

These threats are real and serious, especially in safety-critical and liability-critical applications. Fortunately, excellent research by my colleagues is already ongoing to tackle these problems. However, one would wish that also the governing authorities would treat this topic with the seriousness that it deserves.

Given this, what’s your opinion on GNSS back ups?

Back-ups are important. Back-ups introduce redundancy and redundancy has the potential to improve Integrity. What is important though is that the back-ups are independent and chosen such that they are indeed able to strengthen the vulnerable parts of the system.

How do you think the GNSS positioning technology can take the advantages of other positioning technologies cell phones, Bluetooth and WiFi, etc?

These positioning techniques can indeed benefit from each other, conceptually as well as through integration. GNSS is predominantly an outdoor positioning technique, but through integration with terrestrial techniques could contribute to indoor positioning as well. And on the conceptual level, I believe there are various GNSS positioning concepts that lend themselves for the terrestrial positioning techniques as well. One important example is in my opinion carrier-phase tracking combined with integer ambiguity resolution.

UAV is a new technology being widely talked about. How do you see its integration in GNSS technology? What kind of innovative applications it may result in?

Positioning and navigation of unmanned aerial vehicles (UAVs) already relies to a large extent on GNSS. But what I see as an exciting development is the potential that signals of opportunity bring (e.g., signals from LEO satellites, cellular, AM/FM radio). Such navigation has been demonstrated with a positional accuracy in the several metres range, but this could improve dramatically if the highprecision carrier-phase positioning concepts from GNSS can be incorporated. For instance, the fact that the relative receiver-transmitter geometry changes more rapidly with the terrestrial signals of opportunity than with the GNSS satellites in MEO-orbits, is very beneficial for carrier-phase integer ambiguity resolution. With carrier- phase integer ambiguity resolution included, one would then be able to benefit from the high precision with which the carrier-phases can usually be observed.

What influences you envisage in satellite navigation in the near future given the advancements in the field of AI, Autonomous Vehicles, etc.?

I believe – since in an ever more connected world, society’s reliance on high integrity positional, navigational and timing (PNT) data is rapidly growing – that the demand for satellite navigation’s delivery of high-integrity products will strongly increase. The development of a proper integrity theory capable of covering the various challenging applications is therefore paramount.

New concepts of satellite navigation will also have their impact. I think we will be on the brink of another PNT-revolution, if we really can take navigational advantage of the various new LEO-constellations with their hundreds to thousands small satellites. Also new concepts, like the Kepler satellite system, are very exciting as a next-generation GNSS. This concept builds on some of the technological breakthroughs of recent years such as the introduction of optical clocks and the development of optical ranging and optical communication.

You have made significant contributions to educating future generations in different part of the world. Would you like to share your experience in GNSS education? What challenges you see before the academic community and GNSS education?

The beauty of GNSS education is that students with very different educational backgrounds can participate and excel in it. This is true for students from the engineering disciplines, like electrical, mechanical, aeronautical, geodetic, remote sensing and civil engineering, but also true for students from mathematics and physics. I have therefore been very fortunate to work at the different universities with their bright and hardworking students having these various backgrounds.

One of the educational challenges I see (and this is not restricted to GNSS education only) is that with today’s ease of data processing, students run the risk of acquiring a certain ‘analytical laziness’. Such would however be detrimental to their development and general problem solving skills. It is therefore up to us educators to make sure that we keep the analytical skills of the students well trained. And this requires, also in current times when ‘solutions’ can be computed by a simple click of a button, that students remain well-trained (next to basic mathematics and physics) in such foundational subjects as linear algebra, probability and statistics, multivariate calculus, and numerical analysis.