• After 15+ years, we've made a big change: Android Forums is now Early Bird Club. Learn more here.

All Things GPS

Torq said:
-snip-
Hell, aviation grade GPS systems costing the same as a small car, and equipped with motion adjusting algorithms are about the only devices to even claim <10 feet accuracy (and they don't guarantee that in static situations).
Click to expand...


For aviation, there will be a ground based transmitter that is needed to provide the accuracy needed to fly an approach. As an FYI, in the world of surveying, a ground based reference transmitter is also used, and you can get sub-centimeter accuracy.

Before 5/2/2000, GPS accuracy was +/- 100 METERS, then SA (Selective Availability) was disabled which gave nominal accuracy to +/- 15 METERS. A while later, land based transmitters serving North America were installed - called Wide Area Augmentation System (WAAS). Normally, a WAAS enabled device will get you accuracy in the range of 3 METERS.

GPS is still a MILITARY based endeavor... The common civilian doesn't get the 'good stuff' (high accuracy data) that the .mil folks do. But, if you have a WAAS enabled device, you can get +/- 9 FEET from where you want to be, which is plenty close enough for most folks - unless you're flying that is...
 
MrFlipside said:
For aviation, there will be a ground based transmitter that is needed to provide the accuracy needed to fly an approach. As an FYI, in the world of surveying, a ground based reference transmitter is also used, and you can get sub-centimeter accuracy.

Before 5/2/2000, GPS accuracy was +/- 100 METERS, then SA (Selective Availability) was disabled which gave nominal accuracy to +/- 15 METERS. A while later, land based transmitters serving North America were installed - called Wide Area Augmentation System (WAAS). Normally, a WAAS enabled device will get you accuracy in the range of 3 METERS.

GPS is still a MILITARY based endeavor... The common civilian doesn't get the 'good stuff' (high accuracy data) that the .mil folks do. But, if you have a WAAS enabled device, you can get +/- 9 FEET from where you want to be, which is plenty close enough for most folks - unless you're flying that is...
Click to expand...

On the general behavior of GPS people keep "correcting" me by posting the same numbers I am quoting, which makes no sense.

Use static transmitters in precisely known locations, with very accurate clocks (all running in sync), and yes you can get ultra-precise fixes.
There are many, comparatively, short range systems that do just that and have for years.

This gets much harder with moving signal sources, which is precisely why WAAS is based on a network of fixed references.
 
Torq said:
On the general behavior of GPS people keep "correcting" ....

.
Click to expand...

people keep correcting you because you are making false statements. You said no phone can get under 10 ft accuracy. Every smartphone I've used, with the exception of the Vibrant, gets under 10ft accuracy.

So are you telling me these results are made up, fake, a fragment of my imagination? ALso these are not fluke or lucky results, I get this consistently.

HTC PHONE:
samsungvibrant-albums-gps-picture1949-htcgps.jpg
 
SamsungVibrant said:
people keep correcting you because you are making false statements. You said no phone can get under 10 ft accuracy. Every smartphone I've used, with the exception of the Vibrant, gets under 10ft accuracy.

So are you telling me these results are made up, fake, a fragment of my imagination? ALso these are not fluke or lucky results, I get this consistently.

HTC PHONE:
samsungvibrant-albums-gps-picture1949-htcgps.jpg
Click to expand...

No, people either just aren't reading or don't understand.

Telling me that GPS w/ WAAS has an accuracy of 10 feet when I've just said that GPS w/ WAAS has an accuracy of 10 feet is just repeating what I've said.

I'm not telling you that you're imagining what your phone is telling you.

I'm telling you that your phone is reporting an estimate and that there are many factors that go into how accurate it is. Sometimes those factors align and give you better than the nominal, guaranteeable, performance of the system, over time they'll average out so they don't.

I've built GPS systems from the ground up, hardware and software ... I know a little about what I'm talking about here.

I'm also done with this thread and topic.

It's abundantly clear there is a software issue on the Epic in terms of how it initializes the GPS. It's clear there is an issue with a fixed reporting value (which I'm now thinking is someone failing to replace a piece of template code in the Broadcom GPS SDK). It's also clear that the received signal on the Epic is abnormally weak.

The first two are software issues.

The last is very likely hardware; they may be able to tweak the gain on the signal amplifiers in software but that'll amplify noise just as efficiently and likely won't help.
 
Torq said:
I'm not telling you that you're imagining what your phone is telling you.

I'm telling you that your phone is reporting an estimate and that there are many factors that go into how accurate it is. Sometimes those factors align and give you better than the nominal, guaranteeable, performance of the system, over time they'll average out so they don't.
Click to expand...

Okay now you are getting into symantics. Fine fine fine, lets use your verbage, estimate. My HTC phone estimates my accuracy is 6.6 ft.
My vibrant usually estimates me at over 100 ft.

RAISE YOUR HANDS IF YOU WOULD RATHER OWN A PHONE WITH AN ESTIMATE OF UNDER 10 FT, COMPARED TO ONE THAT ESTIMATES 100FT.

My hand is raised.


WHO CARES if its an estimate, as long as the estimate is fairly close. You make it sound like the phone estimats 6.6 but the real accuracy is 3000 ft.
Maybe the estimate is 6.6, but I doubt if the real number is far off from that. BTW I disagree with you, I trust the 6.6 ACCURACY on my HTC phone.

BTW the dictionary.com definition of Estimate:
1) to form an approximate judgment or opinion regarding the worth, amount, size, weight, etc

I'm sure the apporoximation is closer to the stated accuracy than you claim, you are making it sound like the accuracy posted is soooooo way off.
 
Torq said:
No, people either just aren't reading or don't understand.

Telling me that GPS w/ WAAS has an accuracy of 10 feet when I've just said that GPS w/ WAAS has an accuracy of 10 feet is just repeating what I've said.

- snip -
Click to expand...

Sir,

I feel that it's unfortunate that you interpreted my commentary the way you did. For clarity's sake, the information that I presented 'repeated' what you had already posted. Had it been omitted, the other data that I introduced would have had no meaning. So, if it gets your feathers smoothed back into place, I officially give you credit for stating the WAAS accuracy before I did.

All the best,
MrFlipside
 
It would be fantastic if Samsung couples the GPS optimization with Froyo all together, we might have a little surpise in september, imagine if they did that without telling us, like a little present.
 
I'm wondering...

I can literally death grip my Captivate and lose cell signal. WIFI signal isn't all that great either. Can you guys/gals death grip your Vibrants and Epics? Does your phone seem weaker in reception relative to other phones?

Perhaps Samsung's antenna design isn't optimal for WIFI, cellular and GPS. Hence, the lower SNR and constant switching of locked satellites. Just a theory...
 
Since the imputed accuracy (always exactly 30 meters) is obviously a bug, I continue to wonder whether the other details such as SNR are reporting are not also buggy, and should not be accepted at face value. That is, the actual signal strength may be better than the SNR that is being calculated and passed to the Android listener in the API, and hence to utilities such as GPS Test.

The reported general SNR gap of ~10 Db, relative to other handhelds, is a lot. Suspiciously, I don't think users of the international I9000 variant do report this gross pattern. Rather, their symptoms are SNRs that are mostly acceptable, with up to 8 satellite locks. But the phone will never use birds with an SNR below 20. The U.S. Vibrant has different symptoms: SNRs are not generally as low as those now reported on Epics, as I recall. But the only satellites reported as used for the fix are 1 or 2 with very low strength. Also, even in Standalone mode, the Vibrant will report a lock with only 1 satellite. (Of course, the Epic has a different form factor, so the antenna may be different from its siblings' antennas. It could just have bad reception.)

So the details of satellite performance are just squirrelly in different ways in different cases.

But the other Galaxy platform all have accuracy problems, and the actual tracking performance across platforms is a different story. I still haven't been able to spend time doing my own side-by-side tests, but I have looked at the tracks published online. And so far I have not seen any examples as bad as the worst tracks published in testing the other Galaxy platforms.
 
I noticed an odd bug, I accidentally opened Latitude, pressed the "hide your location" button and then hit the quit button. Even with the program off I couldn't get the gps to locate me. I had all the options turned on and it still wouldn't work, and my gps is very accurate. It kept giving me the location not available notice until I went and turned off hiding my location in Latitude.
 
I'm curious if any of the Epic owners are willing to make an organized list of every single GPS setting that came stock on your Epic. I'm very curious what settings Samsung has changed.

If you don't mind dialing in *#1472365#
Then make a list of all the settings in "Application Settings" and "SUPL/CP" settings.

Very Much Appreciated.
 
Here is a comparative test I recorded today, comparing tracks captured simultaneously on my Epic 4G and my G1. The course is described in the description within the linked map.

The course was a walking track, with the granularity of My Tracks maximized (1 meter minimum distance between points, instead of the default 5 meters).

I would have to say that the performance was close, perhaps within the same margin of error, although the G1's track subjectively seems slightly better to me. However, this Epic's performance is not nearly as bad as that on a Vibrant in a roughly similar test I ran a few weeks ago.

Before starting the track, I rebooted both phones, and started GPS Test first and allowed it to run for several minutes to intialize the fix. (The peak SNRs on the G1 were higher (~40 on the G1, ~36 on the Epic). and the average G1's SNRs seemed to be a few DB higher, but not as much higher as I had understood from anecdotal reports here. The SNRs reported on the phones for particular satellites seemed highly uncorrelated. In general, the two phones were reporting a similar number of satellites locked. If anything the Epic seemed to be locking on slightly more. Hard to tell any of that scientifically, because GPS Test cycles the display every couple of seconds.)

During the walking test, the imputed accuracy reported on the G1 was typically 3.0 meters, fluctuating between 2.0 and 4.0. The corresponding figure on the Epic, as expected, seemed hard-wired at 30.0 meters. It is intuitively apparent from the tracks that the G1's imputed accuracy range is roughly descriptive of both. The 30.0 meter estimate, once again, is obviously bogus.)
 
SamsungVibrant said:
I'm curious if any of the Epic owners are willing to make an organized list of every single GPS setting that came stock on your Epic. I'm very curious what settings Samsung has changed.

If you don't mind dialing in *#1472365#
Then make a list of all the settings in "Application Settings" and "SUPL/CP" settings.
Click to expand...

Following is the complete menu structure and default values for that utility, which is called GpsSetup. It apparently supplants the old LbsTest utility used on other Galaxy S phones.

Code: 
I. Button labeled "Setup"

    A. Button labeled "Position Mode"

        1) Button labeled "Operating Mode"

            Radio buttons:
            * Standalone
            * MSAssisted
            * MSBased (SELECTED AS DEFAULT)

        2) Button labeled "Starting Mode"

            Radio buttons:
            * Cold Start
            * Warm Start
            * Hot Start (SELECTED AS DEFAULT)

        3) Button labeled "Session Type"

            Radio buttons:
            * Single Fix
            * Navigation (SELECTED AS DEFAULT)

    B. Button labeled "XTRA Settings"

        Radio buttons:
        * Enable XTRA
        * Disable XTRA
        * Enable Automatic download
          Text entry box labeled "Set Period of download"
          Button labeled "Save time of Period"
        * Disable Automatic download (SELECTED AS DEFAULT)
          Button labeled "Trigger Download"
          Button labeled "Query Data Validity"
          Button labeled "Inject External Time via SNTP"
  
    C. Button labeled "PDE Settings"

            Button labeled "Set Server PDE Address and port"

            Text label: "Current SUPL Address: supl.google.com:7276"
            Radio buttons:
            * Use Default Server Address (SELECTED AS DEFAULT)
            * Use Local Server Address

            Text label: "Set SSL setting:"
            Radio buttons:
            * Enable
            * Disable (SELECTED AS DEFAULT)

    D. Misc settings

        Text label: "Setup"
        1) Button labeled "Set QOS"

           Popup dialog labeled "QOS parameters":
            Text box labeled "Accuracy Threshold mtrs":
      Value populated at 1000
           Text box labeled "Session timeout(in secs)":
      Value populated at 180
      OK/Cancel buttons
   
        2) Button labeled "GPRS Settings" (GRAYED OUT)

        Text label: "Threshold: 1000 Sessiontimeout:180"


II. Button labeled "Command"

    A. Button labeled "Start GPS"
    This button starts cyle of GPS test and shows lat/lon, satellites seen, locked, SNRs, etc.

    B. Botton labeled "Stop GPS"
    This button stops test above

    C. Button labeled "TTFF Test"
    This button runs Time to First Fix test, after user enters test parameters.

    D. Button labeled "Enable Log"
    This button enables logging, after quering for path of logfile on SD card.

    E. Button labeled "GPS Results"

      1) Button labeled "Position Info"
      This button displays current position and satellite info

      2) Button labeled "Signal and measurement Info"  (GRAYED OUT)

      3) Button labeled "NMEA Strings"
        Just seems to display a text literal, "NMEAInfo"  (an unfinished stub?)

III Button labeled "Help"
Does nothing
Hope I didn't leave anything out.

AFAIK, the changes to defaults, relative to the other Galaxy S phones, are the same as those circulated in the first workaround (wrongly called a "fix") that circulated in the other forums for the other platforms: Enable MS Based aGPS mode, and change the SUPL server to the Google address.
 
boomerbubba said:
Following is the complete menu structure and default values for that utility, which is called GpsSetup. It apparently supplants the old LbsTest utility used on other Galaxy S phones.


AFAIK, the changes to defaults, relative to the other Galaxy S phones, are the same as those circulated in the first workaround (wrongly called a "fix") that circulated in the other forums for the other platforms: Enable MS Based aGPS mode, and change the SUPL server to the Google address.
Click to expand...

Thank you thank you. Wow they changed so much, 70 percent of that stuff I don't even have on my phone. They must have really really really been trying to fix this.
 
Torq said:
Err ... not sure what you're on about here ... my Epic has GPS problems too so I don't need convincing that it's borked.

Once locked mine has acceptable accuracy; my VERY first post on this matter specifically said it wasn't as good as any of my other devices.

A bigger issue for me is the damn things outright failure to acquire a lock most of the time, and it needing to be rebooted even for that!

Yes, the SNR is lousy. It'll see 10 birds and not be able to lock on any of them. Yet an Evo, an iPhone 3GS and iPhone 4, an ancient eTrex GPS, hell an old LowRance 5 channel model, let alone one of my modern receivers are up, locked and solid with easily double the SNR for most birds in seconds.

I built a GPS board using a PCB trace as an antenna recently and IT gets better SNR than the Epic.
Click to expand...
OK I think we are on the same page then. I have just about had it with the drive-bys of people saying their GPS is working fine when they haven't even tested it and this thread is not about whetehr it works at all but how well. the GPS performance on the Epic is objectively very much below what typical smartphones over the past two years do.
 
boomerbubba said:
AFAIK, the changes to defaults, relative to the other Galaxy S phones, are the same as those circulated in the first workaround (wrongly called a "fix") that circulated in the other forums for the other platforms: Enable MS Based aGPS mode, and change the SUPL server to the Google address.
Click to expand...
When I do that it looks like my epic using hybrid of towers and sats under nominal condtions. I don't know what is going on with the api reporting to the testing sopftware but it is showing birds where there are none at elevations that suggester towers.
boomerbubba said:
The peak SNRs on the G1 were higher (~40 on the G1, ~36 on the Epic). and the average G1's SNRs seemed to be a few DB higher, but not as much higher as I had understood from anecdotal reports here.
Click to expand...
My average was 20 lower db for Epic which is a lot and it isn't an anecdote but a test. I believe others are reporting the same numbers.
 
Aero1 said:
When I do that it looks like my epic using hybrid of towers and sats under nominal condtions. I don't know what is going on with the api reporting to the testing sopftware but it is showing birds where there are none at elevations that suggester towers.
Click to expand...

I don't understand your statement about "When I do that..."

The values I reported were the default settings on my Epic. I didn't change anything.
 
Aero1 said:
My average was 20 lower db for Epic which is a lot and it isn't an anecdote but a test. I believe others are reporting the same numbers.
Click to expand...

What software are you using for your test? I am looking for something that would run on any Android platform, and log the results. My problem trying to characterize the SNRs using the GPS Test app's GUI, side-by-side on the two devices, was that GPS Test refreshes its display very rapidly so hard data doesn't get captured. But I'm pretty sure I didn't see an average difference of 20 DB.
 
So have any of you noticed your GPS works better at certain times than others?

For example, a few california residents have made the connection between time and satellite locks. It appears at night, we are able to get more satellites locking on. For example on my phone, during the day I get zero-one satellites locked and in use. At night, I can get three-four.

Anyone else make the connection between time and number of sats locked on?
 
SamsungVibrant said:
So have any of you noticed your GPS works better at certain times than others? ... Anyone else make the connection between time and number of sats locked on?
Click to expand...

I haven't had that experience. But in general, one property of the GPS satellite system is that the birds are in orbits repeating twice daily, almost exactly. So the constellation of satellites you see in a certain spot at a certain time will be the same set you see there 24 hours later (with a difference of about 4 minutes.)
 
Torq said:
What I've seen after a day of usage however is that periodically I have to restart the entire phone to get it to lock on to any satellites at all. Then it'll work fine for a while again, and at some point thereafter it will still see satellites but won't lock on to any and consequently needs to be rebooted before I can use GPS again.

This would seem to be a software/GPS stack issue rather than a hardware fault, but it is very annoying.

Anyone else seeing the same thing?
Click to expand...

I did experience that once this weekend, and also cured it by rebooting. I have read similar anecdotes in another forum.

I agree with you about the GPS software stack, which obviously has changed significantly with the Epic. My suspicion is that it has something to do with the cached data used for the GPS Hot Start initiialization.

If I experience this problem again, I think I will try a new workaround: The very latest version of the GPS Status utility (which I updated yesterday) includes new functionality in its Settings menu: With a couple of clicks, the user can clear the GPS cache and then reload the satellite ephemeris data over the Internet. The functionality, not intended only for the Epic, worked for me and was very fast. But I wasn't testing it during such a problem incident.
 
I had flawless coverage in a car last night in Austin, TX. I was using Google Maps Nav and it was basically in real-time, maybe a 1 or 2 second lag. This was with GPS and Use Wireless Networks on.
 
i compared mine next to a droid 2 today running 2.2. both basically locked the same, but SNR was higher on the droid 2 and accurace was like 2 meters versus 30 meters on gps status. not super but not horrible.

I've tested it driving around and usually i'll lock 1 less satellite than is visible, sometimes 2 less. usually I have like 6-8 or 9 out of 10. when it was pouring out previously today, I locked 6 out of 10 with no problems, so while I firmly believe there is a hardware issue with these phones (as I've stated all along), it does seem decently usable. Also I noticed 2.2 seems to have an extra checkbox aside from the two that we have, something about AGPS, don't know if thats just specific to the droid.
 
You must log in or register to reply here.
Back
Top Bottom