GPS Data/Accuracy Improving?

[kalnel#WN]

For the past few days, I've been leaving my GPS radio on, mainly to get a sense of how much it actually affects battery life. I've also been using either Navigator and VZNav a lot.

As I've driven around, I've noticed an interesting thing: Both programs seem to be getting more accurate the longer I run GPS and use them.

For example, when I've used VZNav on the Fascinate in the past, it has put my house and several other places I go frequently on the wrong block, even when I can see/am connected to several satellites. Now, it's consistently more accurate to within about 10 feet.

I'm curious if either of these programs "learns" or has a way to correct data as you run them, because the accuracy difference lately is amazing.

 

[jack22182]

For the past few days, I've been leaving my GPS radio on, mainly to get a sense of how much it actually affects battery life. I've also been using either Navigator and VZNav a lot.

As I've driven around, I've noticed an interesting thing: Both programs seem to be getting more accurate the longer I run GPS and use them.

For example, when I've used VZNav on the Fascinate in the past, it has put my house and several other places I go frequently on the wrong block, even when I can see/am connected to several satellites. Now, it's consistently more accurate to within about 10 feet.

I'm curious if either of these programs "learns" or has a way to correct data as you run them, because the accuracy difference lately is amazing.

I've never seen GPS data get better over time once the initial accurate fix is established. Outdoors with a good view of the sky you should be getting around ~10' +/- depending on how many satellites are in view and what the geometry is, shortly after a GPS 3D fix is established.

I don't think the apps do a very good job of telling the user where the location fix is coming from. I see folks that comment in their posts that "the GPS is giving them a 30 meter fix", etc. If someone is getting a 30 meter fix (e.g. "wrong house"), it's probably coming from a WiFi based location service, where the phone determines its location based on the WiFi hotspots it can hear and their locations based on someone driving around collecting hotspot signal vs. location data. If someone is getting a 200-1800 meter (or greater) fix (e.g., "wrong block"), its coming from the cellphone tower location service.

A really good way to learn more about how GPS behaves in the phone are apps like GPS Status or GPS Tools. Both will let you watch your phone's GPS acquire satellites in real time, so you can see exactly how long the phone is taking to acquire a fix, what the overhead geometry of the spaceships is, and what kind of signal levels you are getting from each satellite.

 

[worwig]

It will improve over a period of many minutes. But after 10 to 15 minutes of good satellite signal, it should have all of the satellite catalog data downloaded.

As mentioned, try something like GPS status and see how many satellites you are getting.

 

[muzhik]

i use GPS Status (free on the market) it works magic, great add on for google maps

 

[kalnel#WN]

That's the interesting thing -- I've always had good satellite coverage and relatively quick locks. I use GPS Test to see the lock-ons and generally get 9-12 satellites and locks. According to that program, my accuracy ranges anywhere from 3 feet to 45 feet.

But, even have locks on 10 satellites and good accuracy (according to GPS Test), the map accuracy has often been far off. In the case of VZNav, it has had buildings on the wrong blocks, etc. Same with Google Navigator. That's what seems to be working better as I've kept the GPS on.

I can see a difference in accuracy between when it uses cell towers for a location v. when it uses GPS, so I'm pretty confident that the past inaccuracies weren't because of that.

Interestingly, I've also noticed that since I flashed DL30, my compass apps are off. (One is 180 degrees off.) It makes me wonder if something "fixed" the mapping and "broke" the compasses.

 

[jack22182]

That's the interesting thing -- I've always had good satellite coverage and relatively quick locks. I use GPS Test to see the lock-ons and generally get 9-12 satellites and locks. According to that program, my accuracy ranges anywhere from 3 feet to 45 feet.

But, even have locks on 10 satellites and good accuracy (according to GPS Test), the map accuracy has often been far off. In the case of VZNav, it has had buildings on the wrong blocks, etc. Same with Google Navigator. That's what seems to be working better as I've kept the GPS on.

I can see a difference in accuracy between when it uses cell towers for a location v. when it uses GPS, so I'm pretty confident that the past inaccuracies weren't because of that.

Interestingly, I've also noticed that since I flashed DL30, my compass apps are off. (One is 180 degrees off.) It makes me wonder if something "fixed" the mapping and "broke" the compasses.

I got ya now. I've never seen this (after a fix is established, one app gets good fix data and others are way off). I wonder if there can sometimes be problems when multiple mapping/GPS apps need to access the GPS data stream at the same time. For grins you might try shutting down any background GPS Status/mapping type apps and see if that makes any difference with the one you want to use.

For the compass issue, have you every tried the "Wave the phone in a Figure 8 pattern a few times" calibration?

To be quite candid, I have no idea what they mean when they say "Wave the phone in a Figure 8 pattern a few times", but if my compass reading is way off, when I do what I think they seem to be suggesting, it fixes the calibration right away. I wish someone would post an official video of the Figure 8 procedure though. It seems like there could be thousands of possible ways to "wave the phone in a Figure 8 pattern".

For what it's worth, I just hold it straight in front of me and make a series of wide Figure 8s. Sideways and vertical seem to work equally well.