The Global Navigation Satellite System

I'm not very good at this kind of thing but try this. If it doesn't work I suggest that you join the forum.

Hugh

Phew!

I'm afraid that I just go to pieces when I'm asked to make links; I never expect them to work.

Hugh

Of course, there's no The Global Satellite Navigation System, but there are a number of Global Satellite Navigation Systems, active or planned; the Global Positioning System (GPS), GLONASS, Galileo, Beidou/Compass, etc.

Hmm.. and looking at the link above

i) "In order to provide meaningful accuracy, every GNSS needs SBAS (a Satellite Based Augmentation System)"

Whilst SBAS increases the accuracy of a GNSS, you can still get 'meaningful accuracy' without it (e.g. 20-30m).  I guess it depends on your definition of 'meaningful'.

ii) "The American’s is called WAAS (Wide Area Augmentation System) and as it had to be retro-installed into their GPS (Global Positioning System) and it caused them lots of headaches."

No more of a headache than EGNOS.  Both systems require monitoring stations, and the means to transmit correction signals (WAAS uses geostationary satellites to transmit correction signals, no different from EGNOS), and to have receivers capable of exploiting these signals.

iii) "The European Space Agency learned from the experience of the Americans, and built and commissioned their SBAS first."

WAAS was operational in the US before EGNOS was officially declared operational in Europe. WAAS is not intended to cover Europe.  WAAS was approved for instrument landings before EGNOS (and not long after EGNOS test transmissions started).

iv) "in other words we are already all using Galileo in Europe"

No we aren't; we are using EGNOS.  EGNOS isn't Galileo.

I passed the micronavigation thread under my colleague's nose for a more expert opinion.  Here is his response:

<quote>

> In order to provide meaningful accuracy, every GNSS needs SBAS (a Satellite Based Augmentation System)

Not true, standard-alone GPS gives good accuracy (~8 m) for many applications. The benefit of SBAS is that it allows the receiver to reliably estimate the instantaneous limits of that accuracy – this is essentially for critical applications. A GNSS doesn’t need an SBAS.

> The American’s is called WAAS (Wide Area Augmentation System) and as it had to be retro-installed into their GPS (Global Positioning System) and it caused them lots of headaches.

WAAS (and EGNOS) are standalone systems – there was no retrofitting into GPS (or Galileo). Both make observations of the GNSS SVs and use these observations to estimate error corrections and error bounds for each SV and the GNSS systems as a whole.

All complex systems cause “headaches”. The major verification problem with both WAAS and EGNOS was collecting enough data to statistically prove that the integrity failure rate was low enough for safety-critical applications. Both systems were useful additions in many applications, even if not approved, long before certification was signed-off.

> The European Space Agency learned from the experience of the Americans, and built and commissioned their SBAS first. It is called EGNOS (the European Geostationary Navigation Overlay Service) and, as such, it is the first pan-European satellite navigation system.

WAAS was deployed before EGNOS (July 2003 vs. March 2011). Both meet the same standard (RTCA DO-229), which allowed some of the same equipment to be used in each. The underlying system algorithms and architectures are different, but this is not apparent to the receiver. Both complement the GPS systems but do not replace it: EGNOS is not a satellite navigation system – it is equivalent to WAAS but covers the European region. Note that there are additional equivalent systems planned: MSAS for Japan/Asia, GAGAN for the Indian subcontinent.

DO-229 includes support for broadcasting corrections for non-GPS constellations, however bandwidth limitations on the broadcasted signal mean that the existing WAAS/EGNOS systems will not support Galileo.

> The optimal number of SV’s in an orbital plane is 4 (GPS has six orbital planes, each with 4 satellites in - hence 24 being the number required for complete functionality).

The “optimal” number is a cost/performance trade-off. Low numbers of planes are cheaper to deploy, since the same launcher can deploy multiple satellites. Higher numbers of planes/more SVs per plane improve accuracy.

> In the European system the SBAS SV’s are geostationary, in other words they are in permanently fixed locations above Europe. Currently there are 3 in-situ and operational.

WAAS and EGNOS each use two satellites for redundancy. Both also have an additional satellite for test purposes.

> EGNOS Open Service has actually been available since 1 October 2009 and it augments the US GPS satellite navigation system, as WAAS’s footprint is optimised for the USA.

Correct.

> EGNOS allows the positions reported to be accurate and reliable enough for safety critical applications, such as the safety-critical task of guiding commercial aircraft - vertically as well as horizontally - during landing approaches and navigating ships through narrow channels.

Constraining the vertical error bounds is probably the most critical aspect of SBAS operation. This use case comes from precision vertical guidance on final approach for aircraft.

<contd>

> Last December the German Federal Aviation Authority approved EGNOS landings at major airports throughout Germany for all commercial aircraft, with the UK & France following suit earlier this year – this is importanceof this decision cannot be over-emphasised!

If you are a lawyer, then yes. For everyone else the key date for EGNOS was July 2005 when initial operations started. Even at that point EGNOS was providing significant benefit for some applications.

> What it means for us folk who use the basic eTrex right up to an encrypted military satnavs, is that we are now all using the transmissions from these 3 geostationary satellites, and the computed data corrections from the ground based installations – in other words we are already all using Galileo in Europe – and because the system was bespoke, rather than retro, it allows users in Europe (and beyond) to determine their position to within 1.5 metres.

There are several misleading parts to this statement:

Existing SBAS systems only monitor the public unencrypted L1 signals - the corrections are not directly applicable to the encrypted signals. This is primarily because the SBAS systems provide corrections for ionospheric delay suitable for use by single frequency (i.e. low cost) receivers. Under open sky conditions the residual uncorrected ionospheric delay is the predominant source of position error.

Dual-frequency receivers (i.e. military/professional) can measure the ionospheric delay directly and therefore don’t need the SBAS ionospheric corrections, though there is some benefit in using the clock and orbit corrections provided by the SBAS.

Note that the SBAS signals are just as vulnerable as the open services they support. In a hostile environment the availability of SBAS data cannot be assumed.

Furthermore, the two SBAS satellites in each redundant pair both transmit identical data streams. A receiver therefore only ever uses one SBAS satellite, but may track the other to ensure continuity.

Galileo is not yet operational Europe.

With the GPS L1C modernisation the single frequency L1 performance of both GPS and Galileo will be comparable. Where Galileo will really shine is with the new E5 ALTBOC signal, however it is very unlikely that any consumer-grade equipment will track this signal due to the high level of signal processing (cost) required.

But probably most critical: SBAS does not guarantee any level of accuracy. Under open sky (i.e. aviation/maritime) conditions SBAS will allow a user to accurately estimate the current level of accuracy and help to stabilise some of the extreme error bounds. In urban/rural environments SBAS provides little benefit in either actual accuracy or the estimation of accuracy. This is simply because the SBAS and the receiver know nothing about the local environment.

Reported/estimated accuracy is not the same thing as actual accuracy.

e.g. in many multi-story carpark buildings it is possible to track multiple GPS SVs and an SBAS satellite, however the actual accuracy is likely to be much, much worse than predicted. Why? Because the SBAS data and the receiver are unaware that the signals being received are reflected off the office building 200 m down the street. Similar examples exist for road, rail, general outdoor use, and many maritime use cases around harbours.

<contd>

> Given that Galileo was designed for Europe the earliest satellites will cover this region, again what this means is, that whilst it takes between 24-30 satellites to give global cover, 14 will provide full cover for Europe, and the launch period for these is planned for the next 3 years. So by 2015 we will be using Galileo as our primary GNSS in Europe - and the system will be fully up and running (globally) by 2019.

The Galileo satellites are not geostationary and therefore will provide world-wide coverage from initial launch. However the initial availability will be low due to the low number of satellites causing visibility constraints. To be useful Galileo needs at least 24 satellites (much like GPS – that’s physics for you).

With the advent of Galileo there will be no such thing as a “primary GNSS” – all modern receivers will combine measurements from both GPS and Galileo (and Glonass) to form a optimal solution across all available systems. Even today many professional applications fuse GPS and Glonass observations.

> You will notice that I used the word primary; I used this because the algorithms for new GNSS chipsets will preferentially differentiate between GPS & Galileo signals when operating within Europe as the predicted accuracy for the system is 1.0m compared to the US GPS system 2.7m. (This 'preference' is a contentious subject, for another posting, and not often publically aired).

Not true. Modern navigation algorithms “weight” the measurements from each tracked satellite based on an estimation of the quality of that satellites signal. The weighted measurements are then filtered and combined to generate the navigation solution. This process happens in GPS-only receivers and will happen in all future receivers. There is information in any available signals – the receiver takes a view of the reliability/trustworthiness of each signal based on statistical data for that signal, the signal strength, the SBAS data, the elevation of the satellite above the horizon, and many other factors.

> The wild card, that has come left of field, is from the Russians this month. The Russian Deputy Prime Minister, Igor Shuvalov, announced that all foreign-registered aircraft must use GLONASS in Russian airspace. Therefore, the architecture of all GNSS chipsets in commercial aircraft will need revising and this will quickly spin off into handheld satnavs.

It is not clear if this will actually happen since there are international agreements/treaties to consider.

The real wildcard is the Chinese COMPASS system. You could consider Glonass as being “confined” to Russia, however if Chinese-manufactured goods start to include COMPASS receivers then that will effectively become the de-facto standard. Navigation is largely regarded as a service - most consumers don’t really care if the infrastructure is GPS or Galileo or Compass.

> Some forward companies have anticipated this, in particular Sony Ericsson and Apple who in the past few months installed a new GNSS chipset into their Sony Ericsson Xperia S and Xperia ION phones and Apple’s iPhone 4S, which uses GLONASS.  So these two leading phones now use GPS, GLONASS & Galileo to navigate! And by using all of these systems, the total number of satellites available becomes 58 (31 GPS plus 24 GLONASS plus 3 EGNOS).  GNSS has truly come of age, and I personally believe it affords us all greater safety and security.

Don’t be fooled – your safety and security is no better than it was before. User error is still your greatest threat.

</quote>