It's actually impossible for GPS to show your current position.
The signals have to be received from each satellite, then processed to yield a position valid at the time of transmission.
Every GPS fix you get is delayed by AT LEAST that processing time. Any filtering adds more lag.
Most navigation systems look at the T(fix) -> T(now) difference and project your now position from the prior fixes. Especially if you're following driving directions as opposed to free movement, then programs like Maps etc project how much further you've moved on the route, not just along your velocity vector.
After a few seconds, though, that projection will stop moving, too, when the gap between last fix and now gets too large.
> getting a GPS fix through network assistance still sends data to Google servers
Well, that's not a GPS fix.
If you want a traceless position fix, use GPS Navstar / Glonass / Beidu / Galileo. Yes it takes longer from cold start, but that's the trade-off for privacy.
I wonder if filtering just based on the GPS previous positions (thus averaging different errors to kinda cancel them) works almost equally well in this case, since cars tend to move in a quite smooth way.
That is not true, A-GPS just speeds up the time to fix by supplying ephemeral data via a sidechannel. It just takes longer to receive via GPS if it is out of date or you have moved position considerately since last fix
The parent post already noted that the GPS doesn't know your instantaneous speed, just the average speed since your last reading. If your signal was blocked by trees, buildings, bridges, (not to mention your car itself), you can go at least several seconds without a reading, and what happened in between those points is unknown to the GPS, except for total ground covered. And that's even before you account for GPS inaccuracy, which is ~3m in optimal conditions.
> and my phone has no problems finding GPS and I have navigation instructions pulled up in about 5 seconds.
That's because your phone isn't actually using GPS ( e.g. Navstar or Glonass ) at that point, it's using cell triangulation or wifi-based location.
For example the fastest possible first-fix with Navstar from a warm-start, using cached prior-location and ephemerides, is 30 seconds[0]. From a cold ( position-unknown ) start, such as a car nav-system powered-down in a parking-lot for an extended period, the first-fix will take longer than 12 minutes. There is no technological way around that, it is an artifact of the system architecture.
[0] and that's for a top-end receiver than can sync to four satellites 'simultaneously' using time-division.
Another problem with GPS is the accuracy (or lack thereof). Sometimes you get a good GPS signal (e.g. outside, clear view of the sky etc.), sometimes you don't, and you position could be off by 10m. Some GPS store the accuracy of each track point (e.g. VDOP and HDOP), some don't. If your GPS doesn't store the accuracy, then someone could claim your GPS is off by 20m, and hence you were speding (or the data is unreliable).
> The GPS issues have nothing to do with getting a GPS fix. It's just a plain ol' bug.
GPS on phones aren't that accurate, especially downtown, but most applications will filter the data because of that. Ever use Google Maps for directions and make a turn different from what the app is telling you to do? Your car icon will continue down the "correct" path for 15-30 seconds before re-adjusting to the new route. This actually happens all the time, it's just more likely that your GPS is misreporting you as 20m east instead of you taking a different turn, so the app holds off on updating the display until it is more sure.
So it's a combination of the two factors. But most apps handle it muuuuch better than P:GO does.
Here's a link I was able to find on running apps specifically. Apps that deal with you being on foot will obviously have a harder time determining exactly where you are, as opposed to apps for driving where you're moving faster, in a straight line, and down a grid of streets.
"On a two-dimensional surface, GPS position errors introduce a systematic bias, tending to exaggerate the length of a trajectory."
It's not that simple as it really depends on the sampling rate. For example, if the sampling delta is the same time as the lap time, the GPS will think that it is not moving.
And it's not just theoretical: you can see from the map that the GPS thinks he's always cutting corners.
That's crazy... I think I'll contain my frustration more now when my GPS takes a few minutes for a cold lock. It's even more crazy because with the latest GPS systems in phones, I'm able to get a fix with only a very small portion or even no sky visible.
> While modern GPS can get a position fix pretty quick
Can they? I have popular usb gps receivers and it still takes many minutes to get a lock on enough satellites to give accurate positioning. Given that it is dependent on things flying overhead in space, I don't think the receivers are where any "boot speed/signal lock" improvements can happen.
Mobile phones are no exception, they've just employed several tactics to make the perception seem as if it's instant.
They start with "course" location which is based of the geo data of the cell's ip address and vicinity to known wifi access ssids. They also know where you were at when the phone went into standby or was powered down, and can reasonably assume you haven't gone far away from that location without the accelerometer noticing movement of any kind (keep in mind most phones do not fully shutdown even when powered off). That's all a very good and pretty accurate starting point for when you fire up maps on your phone, and by the time it is a problem a "fine" gps signal is locked. And usually due to other system services (like network config, find my device, etc) firing up gps when you unlock your phone or some other privacy invading app running gps as a background task, there's usually a fine grained gps lock by the time you even open your maps app.
The signals have to be received from each satellite, then processed to yield a position valid at the time of transmission.
Every GPS fix you get is delayed by AT LEAST that processing time. Any filtering adds more lag.
Most navigation systems look at the T(fix) -> T(now) difference and project your now position from the prior fixes. Especially if you're following driving directions as opposed to free movement, then programs like Maps etc project how much further you've moved on the route, not just along your velocity vector.
After a few seconds, though, that projection will stop moving, too, when the gap between last fix and now gets too large.
reply