Galileo and EGNOS User Satisfaction Survey

[Emilio González]

Hello friends,

The EUSPA has opened the Galileo and EGNOS User Satisfaction Survey:

https://ec.europa.eu/eusurvey/runner/GALILEO-EGNOS-User-Satisfaction-Survey-2023

I think it is a good oportunity to give your opinion about these European programmes.

Thanks for sharing.

[Sean Barbeau]

Thanks Emilio!

Sean

--
You received this message because you are subscribed to the Google Groups "GPSTest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpstest_andro...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/04bf7a88-f94d-4e33-a015-93bab782ba1an%40googlegroups.com.

[Jim bell]

Myself, I am highly DISsatisfied with:

The failure of SOMEBODY to activate Galileo HAS IDD, which I consider truly astonishing.  I understand that once they committed the BONEHEADED STUNT  of putting the E6 signal "out of hand" making it entirely impractical to install an E6-resonant antenna, meaning that NOBODY was going to support E6 in smartphones, that left IDD (Internet Data Delivery) to be the sole way HAS corrections would be supplied.

Nevertheless, seemingly nobody has done that!

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/CAEkVq%2BijOp0nG-on%2BvMAm16fe7uSUrcbuaEm8riijd1x-wSkpA%40mail.gmail.com.

[Louis Philippe]

I agree with you Jim…… HAS is great….but they should provide a Youtube link with step-by-step instructions!

On Jan 31, 2024, at 15:53, Jim bell <jimd...@gmail.com> wrote:

le wa

[Emilio González]

That is under preparation...

@Jim: Please think about who is responsible of what... Is it the Galileo system or is it someone in the downstream industry to blame?

Cheers,

Emilio

[Jim bell]

Thank you for the caution. While I have been involved in electronics, mostly as a hobby for the last 54 years or so, I realize that this is more "software" than "electronics". 

   Everything is done by microprocessors, and there has to be a mechanism to turn the gnss data including corrections into a final value that the phone can use.

What I am concerned about is that the people who were designing Galileo at some point knew that they would be defining the methods for supplying those corrections, either by E6 signals or by IDD corrections. They must have known that just about no smartphones would be incorporating an E6-capable resonant antenna, so they should have delivered these specs to the various manufacturers of the gnss chips that would be eventually installed in the smartphones. 

I noticed also from the smartphone gnss database that there is virtually no references to the year of gnss hardware past 2020.  Yet, there are numerous references to 2019, 2018, etc.  when I see this, I wonder if they simply "missed the cut": somehow, they didn't get to the door of those various IC manufacturers to tell them how HAS IDD would be implemented.

If I knew more of the internal operations and architecture of smartphones and gnss chips, I would probably have the answer.     

I hope somebody can direct me to some literature that describes how Galileo intended, and presumably still intends, that these manufacturers implement the IDD correction system.

I have in my head somewhat of a cartoon view of a veritable flood of modern gnss chips that will somehow start being incorporated in the newest smartphones.  

[Emilio González]

Interesting thoughts, Jim.

I agree with you and I hope the Galileo programme can keep up with the adoption pace needed for a satisfactory rollout of high-accuracy applications in the mass-market environment.

I will try to know more about those adoption plans and then share my findings with you.

Best regards,

Emilio

--

Emilio González

[Jim bell]

Since I am a beginner on this, I should mention that there's something else I do not understand: I do not know why the GALILEO HAS correction factors cannot be delivered by EGNOS, in an "in-band" signal that can be received with any device that can received E1 or E5a.  

[Emilio González]

Dear Jim,

First, it is simply because EGNOS is a different system to Galileo: different network of stations, different computing centres, even different service providers... Second, because the aim of both services is different: EGNOS (SoL) is focused on providing integrity (confidence bounds to the errors committed), Galileo HAS only focuses on accuracy; in general, their mission requirements are completely different. And third, because adding messages to satellite signals needs bandwidth, which is a scarce and expensive resource.

Best regards.

--

Emilio González

[Louis Philippe]

Thanks Emilio for the clarifications…..

I liked your comment : {{ I will try to know more about those adoption plans and then share my findings with you. }}

We look forward to use HAS in the near future….

Kind Regards,

[gpsfan]

Yes, EGNOS (or WAAS), aka SBAS, is about "integrity" for mission critical use like flying planes ! You don't want to be using a failing GPS satellite in these conditions.
Now SBAS satellites do transmit corrections but it's DGPS "style", i.e. in meters (you can see them with uBlox's uCenter software for instance), which is no way near "high accuracy" (i.e. decimeter range) like you would get with RTK solutions. I haven't looked into HAAS but I'm guessing it's similar to the latter.

[Jim bell]

Okay, what about PPP, or Precise Point Positioning?    I seem to recall it is at least conceptually related to RTK:  maybe it is 'in a different solution-space'??

Wasn't WAAS supposed to be a kind of DGPS that brought brought GPS error down to a meter or so?

--

You received this message because you are subscribed to the Google Groups "GPSTest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpstest_andro...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/18721ba6-1b1c-4bcb-8d8b-f87d77cfc261n%40googlegroups.com.

[Jim bell]

You  said, "that is under preparation".   I  assume this means that  there is a limited amount you can say on this subject, due to some obligations?

Okay, but can you say something like, 'I believe that eventually, most modern existing smartphones that can receive E1 and E5a will be able to install Galileo HAS IDD corrections using a relatively simple app" ???

I suggest that, because clearly without E6 antennas or reception, smartphones WON'T be able to employ HAS SIS corrections.    But to read much of the propaganda from early 2023, it sounded like any Galileo-receiving device (including smartphones) would have no problem utilizing HAS in some form or another, yet they obviously knew that HAS SIS was not going to be available.  It seems to me like there are at least 1000x as many smartphones as survey-grade devices, which means that the Galileo people should be working extra-hard on this project.  

[Emilio González]

Dear Jim, dear all,

There is a brief video in preparation at the GSC that will show how to use the HAS IDD on a practical basis. We are waiting for a module which is under procurement to have the guidelines displayed in a real environment. This is all I can say at this moment.

And "I believe that, eventually, [some*] modern existing smartphones that can receive E1 and E5a will be able to use Galileo HAS IDD corrections using a relatively simple app". There are, to my knowledge, no HAS PPP libraries ready yet and most RTCM decoding libraries out there are not compatible with the TLS security feature needed to decode the HAS IDD corrections, but I can't identify anything major that avoids getting there. So here is my call to the developer's community to have this ready soon!

* To get the most out of HAS, we would need dual frequency AND phase measurement decoding. This is not commonly found. I identified only around 40 capable models in the GPSTest DB (and this does not mean that they will finally be capable of providing the right outputs). We want to investigate further on these capabilities.

Besides, even if we had all the inputs needed, smartphones are not survey-grade devices (antennas etc), so, in the end, you surely won't have decimetre accuracy. But it will be definitely better than standalone GNSS :-)

Best regards.

[Louis Philippe]

Thanks Emilio !!!

--
You received this message because you are subscribed to the Google Groups "GPSTest" group.
To unsubscribe from this group and stop receiving emails from it, send an email to gpstest_andro...@googlegroups.com.

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/1bba4dbf-1d9d-4f22-add1-59f95ce3645fn%40googlegroups.com.

[Jim bell]

This certainly looks like major progress. If they could do as well as 60 cm accuracy that would be great.

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/D2FF9C5C-C7B1-4E74-BD70-A215C26EA08C%40gmail.com.

[Louis Philippe]

Agreed! 

--

Best Regards,

Louis P Roy

[Jim bell]

These days, when a GNSS receiver can receive 25 or more separate satellites, from at least 4 constellations,  I would think that reliability (i.e. failure of a single satellite) isn't a problem.  

[Jim bell]

Last year, I bought a OnePlus Nord2 5g smartphone, DN2103.  It turns out that seems to be an excellent choice because not only does it support dual frequency, it also has B1 and b2a support, E1 and e5a support, and L1 and L5 support.  It does not however support b1c or l1c.  And, of course, like everything else, it does not have E6.  

And it supports SBAS, and since I am in North America it shows WAAS.

 Further, it supports raw measurements, and carrier phase 'get accumulated Delta range'.

So this makes me hopeful that it will support Galileo HAS when it becomes available.  

The only problem I see is that it does not have a micro SD cards slot.  I detest the idea of smartphone manufacturers removing support for MicroSD card slots: they are trying to get customers to pay 2x, 3x, or even 4x the amount of money for their extra data storage.

[Jim bell]

My Nord2 5g, when exposed to an open sky, routinely uses as many as 34 USED satellites in a fix, according to GPSTest. (and sees a handful more.) This makes me wonder how the GNSS chip actually employs so  many signals.  

I remember the 'bad old days' where a Magellan device had ONLY ONE receiver!

Further, it reports a "E H/V" value as low as 1.2 meters, although the V value is typically 10 meters.  This doesn't seem logical:  I don't think the V value shoud be nearly so high. I don't expect to see a vertical accuracy value as low as H, but...

Does this represent some sort of likely accuracy value?  

[Sean Barbeau]

Hey Jim,

re: explanation of estimated accuracy on Android, see this article I wrote:

https://barbeau.medium.com/measuring-gnss-accuracy-on-android-devices-6824492a1389

Sean

[Jim bell]

That's a very useful explanation.

But not surprisingly I have another question.  Upon working with my nord2 5G and GPSTest, it reported that it used 34 gnss satellites and achieved an estimated accuracy of 1.2 meters, wish I considered excellent.  This, I understand, means that 68% of the time, the displayed value is within 1.2 m of being correct.

What what I wonder is this: suppose I use this device with an averaging software, for example Locus GIS.  Suppose I let it average for one minute, or 5 minutes, or 15 minutes, for 30 minutes, or an hour.  How accurate will the resulting value be?

I understand this is based on some obscure statistical mathematics that I would at least be able to appreciate, although not immediately understand.  Implicitly, I expect that we would recognize that the longer the device averaged, the more accurate the resulting measurement will be.  

I also realize that this is not especially useful for a person who is relying on dynamic location: let's say I had a car driving around. But once that car has stopped, if it nos measurements is stopped and reset,, the average value ought to get more accurate over time.

In fact, I expect that there is a graph that can be drawn, time on the x-axis, and estimated inaccuracy on the y-axis, which will show how much better the average value will be.  If this sounds familiar to you, and you could figure out how to find that, I would very much appreciate it.  

To view this discussion on the web visit https://groups.google.com/d/msgid/gpstest_android/c100baca-9a6a-420c-aa4a-cc5530be2e3cn%40googlegroups.com.

[Sean Barbeau]

Hey Jim,

This paper that I co-wrote with Paul Zandbergen evaluating accuracy of AGPS on Java Micro Edition flip phones might help to see our methodology:

https://www.researchgate.net/publication/231849997_Positional_Accuracy_of_Assisted_GPS_Data_from_High-Sensitivity_GPS-enabled_Mobile_Phones

A major takeaway here is that on Java ME phones we found that the number the devices were reporting as "estimated horizontal accuracy" didn't actually meet the 68th percentile spec - meaning that the phones were reporting that they were MORE accurate than they actually were. I haven't repeated this experiment on Android phones.

Sean

[Jim bell]

"In the upcoming years, once chipset receivers able to make use of those innovations will be integrated into smartphones, users will benefit from both those Galileo services for free. This is expected to fuel a massive development of innovative applications. Within consumer apps, indeed, those new Galileo services will enable innovative solutions and/ or functions. The HAS increased accuracy will have a direct impact on the performance of a wide range of consumer application categories such as Gaming, Healthcare, Robotics, AR and Geo-marketing. Mobile payments and enterprise workforce management and tracking solutions are some of the areas that will benefit from OSNMA."

I'd say, this is misleading.  There may be TWO impediments to the use of Galileo HAS:  Presumably, the lack of a designed chipset, but also the lack of the designed-in antenna physically-sized for the E6 signal.  Because of the (mis-)decision by the designers of Galileo HAS to put the E6 signal into an area of the band which is already in use, this means that "nobody" will design in such an antenna.  And so, "nobody" will be inclined to design into a new chipset a feature which will receive E6.

I suggest that smartphone designers will conclude that in a smartphone already designed to include dual-band L1-L5 GPS, E1-E5a Galileo, and B1a-B2 Beidou, "Who would bother to think they need an extra Galileo HAS E6 signal?  ESPECIALLY because it will need a NEW ANTENNA, and ESPECIALLY since Galileo HAS IDD (Internet Data Delivery) has every prospect of being supplied soon?"

And I'd be hard-pressed to debate the point with them?

I'd say the designers of Galileo should immediately figure out if they can insert the Galileo HAS E6 signals somewhere else, that can be transmitted by the EXISTING Galileo satellites, and received by the EXISTING smartphones, ideally so that they can be received by the EXISTING chipsets.

Otherwise these fools have screwed themselves over.  

[Jim bell]

"we would need dual frequency AND phase measurement decoding."

Does this mean "raw measurements" and "Carrier phase", columns M and N of the database?

If so, I feel rather fortunate:  The OnePlus Nord2 5g that I bought months ago has "supported" in both those columns, in addition to "dual frequency".

But watch out!  No doubt you've seen the news about the implementation of the cm-level RTK in the recent ION conference that Sean has pointed to.  Soon enough, we will be seeing things from "Institute of Navigation" to "Institute for Surveying" to "Institute for Slug-trail Following"!!!

On Wednesday, February 21, 2024 at 5:21:25 AM UTC-8 erlang...@gmail.com wrote: