GPS Chip ?

Discussion in 'iPhone' started by Bob, Dec 24, 2014.

  1. Bob

    Alan Browne Guest

    If that BSSID is not already cached, then send the BSSID to the server,
    which returns the estimated location of the WiFi (if previously seen and
    stored in the database). Of course you need a datalink (Apple's own
    Messages service could serve this - out of sight of the user) via the
    data plan on the iPhone.

    As to "accurately", if the various reporting iPhones get a reasonably
    dispersed set of locations around the WiFi station, then the location
    can be estimated as somewhere in the middle. Knowledge of signal
    strength for each sampling iPhone could hone that somewhat. With a half
    dozen such locations around a WiFi I'd WAGtimate 10 - 20 m accuracy for
    the location of the WiFi. Of course that is not the location of the
    iPhone using the WiFi as an initial source location.

    IAC, knowing the position of the WiFi to within 200 metres is enough for
    a good first guess and will help the GPS/GLONASS receiver correlate much
    quicker.
     
    Alan Browne, Dec 31, 2014
    1. Advertisements

  2. Bob

    Guest Guest

    true but the data is still tiny. even if it's bloated by 10x, it would
    be 130 bytes.
    apple says it is in response to a congressional inquiry (locationgate).
    they're not lying.

    there really isn't a need for a unique identifier since the same user
    picking up the same bssid at the same location at another time would be
    useful information. that doesn't mean there isn't but there probably
    isn't.
     
    Guest, Dec 31, 2014
    1. Advertisements

  3. Bob

    Guest Guest

    airplane mode does nothing to the antennas.

    airplane mode turns off all radios, some of which can be turned back on
    while still in airplane mode, such as wifi and bluetooth.
     
    Guest, Dec 31, 2014
  4. Bob

    Guest Guest

    it's exactly for that situation (as well as others).
     
    Guest, Dec 31, 2014
  5. Bob

    Guest Guest

    and it's trivially easy to ignore them.
     
    Guest, Dec 31, 2014
  6. Bob

    john james Guest

    Trouble is that if that has come from just another idevice user,
    that idevice had no way of determining the accurate location
    of the bssid either. All it knows is its own location and the
    fact that it can see the bssid and some rough idea of the
    signal level.
    And would need to ensure that that is only used very
    minimally in case the idevice is currently roaming or
    has no data plan at all.
    Yes, but that is nothing like accurately locating it.
    Only somewhat tho.
    It would in fact be nothing like that unless the idevice
    did a lot of sampling and calculation that wouldn’t
    even be possible as you drive by it in your car.
    You don’t need anything like that accuracy for that.

    All you are doing is calculating what satellites should
    be visible and even 1KM doesn’t change that significantly.
     
    john james, Dec 31, 2014
  7. Bob

    Rod Speed Guest

    Wrong, as always.

    That situation uses what the idevice itself collected
    when it did have good GPS is available.
     
    Rod Speed, Dec 31, 2014
  8. Bob

    Alan Browne Guest

    You're missing the purpose. A first fix within seconds that is within
    100 metres (or 200 for that matter) of where you are is better than no
    fix at all. Further, it aids the GPS/GLONASS receivers in acquiring
    satellites more quickly (the correlators can be near perfectly set
    resulting in far less search time). See bottom.
    Perhaps. On the other hand the amount of data involved is very little.

    No data plan? Okay, keep the phone where it has a good view of the sky
    until it eventually locks onto enough sats to track.
    Depends on the value of accurate. I used to work with Omega/VLF
    navigation systems. 1-2 NM was pretty accurate (and sufficient for long
    range navigation). At night it could be as bad as 4 NM. Still - more
    than sufficient for long range.

    Also worked with Doppler nav. Accuracy was a function of distance flown
    and depended on how good a heading source one had.

    Also worked with RNAV (DME-DME) where accuracies were typically better
    than 1/10 NM (185 metres).

    Worked with unaided GPS before S/A was disabled. That 100 m circle
    seemed damned accurate at the time.

    Accuracy? What's that? Has no meaning w/o application.

    Knowing where I am based on a WiFi station means I'm probably within 200
    metres of the truth. Damned good first estimate for most daily needs.
    More than you'd suspect. You don't need survey accuracy - the whole
    point of the WiFi aiding is a) faster acquisition of the more accurate
    GPS/GLONASS and b) usable accuracy when you can't get sat signals
    (indoor, in canyon, etc.)
    You clearly don't understand what is happening. The iPhone is a
    "sampler" and reports it's samples to the Apple server. The server
    receives many reports on a given BSSID from many iPhones at different
    times (dates/times). Each iPhone is reporting from a different position
    and reports:
    -position of the iPhone (accurately: using its GPS/GLONASS)
    -BSSID of the WiFi station
    -Signal strength received

    So the server (given many samples over a lot of time from a distribution
    around the WiFi) can estimate:

    -the dispersion (was it a circle with somewhat even power
    distribution from the centre of the circle? Then the WiFi is
    at the centre; High accuracy.
    Was it a long obloid? Then the WiFi is
    perpendicular to the centre of mass of the obloid, but no
    way of knowing which side of the obloid.

    And again, this is most useful for a first estimate of where a phone is
    when using that position. _Accuracy_ thence comes from GPS/GLONASS ...
    or even trilateration from local cell towers.
    It's not about "which are visible" but where they are wrt to the
    receiver and most importantly (for quick acquisition) the Doppler shift
    due to relative motion between the receiver and the satellite. That
    allows you to align the correletor (which is a direct function of
    Doppler shift) more accurately and thence be able to track the satellite
    more quickly.

    So, the better you know your position, velocity and time, the more
    accurately you can position the correlators and the more quickly you get
    a fix, thence tracking.
     
    Alan Browne, Dec 31, 2014
  9. Bob

    john james Guest

    You don’t need anything like that for the quicker first fix.

    Even 1km is plenty. Even 10km is plenty for that.

    Its different when there are no GPS satellites visible at all because
    you are in the basement or deep within a big building.
    You don’t need anything like that accuracy for that.
    But if you don’t have a data plan, the data you do use is often
    charged for in 1MB chunks so it doesn’t matter what size the
    data packet actually is.
    In fact even with the data turned off it gets a very fast
    GPS fix and that only matters when the phone has
    been turned off or the GPS has been turned off.
    But we are talking about the accuracy you see with GPS and
    with the claimed better accuracy than that if you leave the
    wifi and bluetooth on.
    But not as good as GPS.
    Yes, but you won't get b) by signal strength sampling as you
    drive past in your car.
    I do actually.
    I don’t believe that it does actually see many reports at all.

    I know that mine doesn’t report anything when I do the
    yard sale run because I can see it doesn’t use any data then.
    But the phone only uses that when its been moved
    a long way turned off or with GPS disabled.
    No, because all that matters is what comes in thru the receiver.

    And even 1km doesn’t change where the
    satellites are wrt to the receiver anyway.
    That doesn’t change significantly in 1km either.
    Not when that doesn’t change significantly in 1km.

    And that only matters when the phone has moved a
    long way when turned off or with the GPS off too
    because otherwise that stuff can just be calculated.
    That is just plain wrong. 1km accuracy is plenty for that.
     
    john james, Dec 31, 2014
  10. Bob

    JF Mezei Guest


    Current location accuracy is not important for the GPS to get a fix
    since you can be 10s of kilometres away and expect to see the exact same
    satellites over you. (hundreds of kilometres off and some of the
    satellites closer to horizon may no longer be visible).
     
    JF Mezei, Dec 31, 2014
  11. Bob

    Alan Browne Guest

    1. You need it to know where you _are_ for your own purposes.

    2. The tighter that estimate, the quicker the GPS/GLONASS are tracked.
    Most data plans (these days) are just how much data you use and often
    are 250 MB, 500 MB, etc. It's 2014 in case you weren't looking and
    that's chaning in a few hours too.
    Depends where you are (urban or real canyon's for example).
    Who said that. The point of WiFi is for faster location of the phone
    and to aid the GPS in faster acquisition.
    So what? It'
    Not all samples get to Apple that way. Indeed they may filter WiFi
    "samples" based on the speed of the sampler to reduce error.
    You're proving you don't.
    The data is likely cached and uploaded when you access a WiFi or USB
    connection to your computer.
    Sure. But the more accurately known the position of the WiFi, then the
    more accurately known the position of the phone for its first fix.
    Bzzt. See below.
    Enough to change where the correlator is positioned.
    It's plenty accurate for iPhone's, that's true. But then so is the
    first fix from the WiFi location or cell tower trilateration in most use.
     
    Alan Browne, Dec 31, 2014
  12. Bob

    Alan Browne Guest

    It's true that in the context of a phone where a few seconds don't
    matter that much that high accuracy won't help fast (really fast)
    acquisition.

    But high accuracy location of the WiFi's definitely helps in urban
    canyons with the position provided by the iPhone.
     
    Alan Browne, Dec 31, 2014
  13. Bob

    JF Mezei Guest


    What would really help get faster GPS location is if the phone gave the
    GPS an indication of current speed. When traveling (even at bicycle
    speed) the GPS takes longer to get a fix because it need to test wider
    range of Doppler shift for each satelite (some will have frequency shift
    down others go up as you more in one direction towards some and away
    from others).

    Obviously, the phone probably couldn't give direction of travel.
    The "default" charges when you don't have a data plan are still often in
    the per kilobyte range and end up in the tens of thousands of dollars
    per gig. This is one reason some carriers such as AT&T will routinely
    actually disable data when they spot a user with an iPhone without a
    data plan.

    In an era where a plan gives you 1gb or more per month, using
    inefficient XML to send data is not an issue. But for users who are
    still billed by the kilobyte, then it starts to make a difference. This
    is especially true since you don't even know when such data exchanges occur.


    This is why when an iPhone doesn't have a data plan, you should always
    disable Cellular Data, and also disable individual application's use of
    data. This way, if you even need to use data to , say, send an email,
    you can enable cellular data temporarily for email and be fairly
    confident that a gazillion other apps won't starts to exhange data all
    of a sudden.

    There is no increased accuracy with wi-fi. Assisted GPS only gives you
    faster access to *a* location, with no garantee that that location is
    accurate until the GPS gets a location.

    And even GPS locations can be inaccurate in urban canyons due to signals
    bounding off buildings. My bike ride into downdown sydney has a most
    interest track that skips from one block to the next (left/right travel
    on straight street).


    Accuracy matters if you are in a city and pull up google maps to figure
    out where you are and whether you should walk left or right.

    If you have been driving around, then chances are your phone has been
    GPS active for quite some time, so acquisition time is not an issue.

    Proper car GPS systems have software tweaked for urban canyons and
    tunnels (likely with accelerometres) so they can pretend to know the
    location after having lost GPS signal and can paliate for urban canyon
    artifacts (it knows the car can't suddently shift from one street to the
    next). I have no idea how a smalrtphoen behaves compared to caer GPS
    systems in that regard.
     
    JF Mezei, Dec 31, 2014
  14. Bob

    john james Guest

    That comment was about your previous purpose, acquiring
    satellites more quickly. That is what the 'for that' was there for.
    That is just plain wrong when you know where the device is to
    within a few km, more accurate than that doesn’t help at all.
    That was talking about how its charge when there is no data plan.
    No, in both of those situations the device already knows
    where the satellites are and just can't see them currently.

    It is completely trivial to calculate where they have moved
    to during the time that they were not visible unless the
    phone has moved a long way turned off or with the GPS
    turned off.
    The phone when you turn those off.
    Its also for location when the satellites are no longer
    visible because you are in the basement etc.
    I know it isn't because I know when my wifi talks to apple.
    That isnt true when it knows the location to within a km or few.

    More accuracy than that has no effect whatever on the speed
    of the first fix because none of that changes when the phone
    moves by a km or few.
    Doesn’t help.
    No. Basic geometry.
    So there isnt any point in knowing where the phone is more
    accurately than that to get a faster first fix because basic
    geometry means that nothing changes wrt the satellites.
     
    john james, Dec 31, 2014
  15. Bob

    Alan Browne Guest

    I'll grant that the accuracy doesn't have to be that great for a quicker
    fix. But - it does to get you located nearby. Got off a long flight
    recently?
    Most current smartphone users have even basic plans with gobs of data.
    (it'll be 2015 here soon, not sure for you).
    Odd then when I'm downtown that it can take several minutes to track.
    Watch it constantly? Keep a log of everything?

    It does, in fact. But I'll grant that for the time scales in question
    it's meaningless.
    Sure. Gotten off a long flight of late?
     
    Alan Browne, Dec 31, 2014
  16. Bob

    john james Guest

    I don’t see anything like that with the iphone, even tho I routinely
    turn the phone off so I don’t have to charge it so often. The first
    fix is very fast even when I am driving around in the car.

    I used to get a significant delay for first fix with the tomtom
    when I turned it on and headed out in the car, but don’t get
    that with the iphone at all, its navigating as soon as the phone
    is turned on.
    It can actually.
    That is only seen much in north america.
    And much more when they are charged by the megabyte
    even if the packet is only a few tens of bytes.
    I don’t do that and I don’t have a data plan.
    That doesn’t in fact work with 8.1.2, it uses
    the data anyway with app store updates.
    I don’t bother and I know that none of the apps start
    to exchange data as soon as I turn the cellular data on.
    There is actually when the GPS satellites can no longer be seen.
    It's still better than no GPS satellites visible than just dead
    reckoning from the last fix when the GPS satellites were visible.
    No, GPS doesn’t work like that.
    I don’t get that myself. There must have been something
    wrong with how you were carrying the phone that prevented
    it from seeing the satellites so it was guessing instead using
    the wifis it could see.
    Yes, but it works fine in the car and is even able
    to work out where you are lane wise too.
    And that is true even when you haven't been driving around.
    Smartphones even know which lane you are in.
     
    john james, Dec 31, 2014
  17. Bob

    Alan Browne Guest

    Look up "Multipath" and get back to us. It is a serious issue with GPS
    receivers in urban canyons.
     
    Alan Browne, Dec 31, 2014
  18. Bob

    Alan Browne Guest

    Unless you've disabled location services then you're missing something.
    You're part of the "crowd":
    ------------------------------------------------------------------------------
    https://www.apple.com/pr/library/2011/04/27Apple-Q-A-on-Location-Data.html

    quote:

    3. Why is my iPhone logging my location?
    The iPhone is not logging your location. Rather, it’s maintaining a
    database of Wi-Fi hotspots and cell towers around your current location,
    some of which may be located more than one hundred miles away from your
    iPhone, to help your iPhone rapidly and accurately calculate its
    location when requested. Calculating a phone’s location using just GPS
    satellite data can take up to several minutes. iPhone can reduce this
    time to just a few seconds by using Wi-Fi hotspot and cell tower data to
    quickly find GPS satellites, and even triangulate its location using
    just Wi-Fi hotspot and cell tower data when GPS is not available (such
    as indoors or in basements). These calculations are performed live on
    the iPhone using a crowd-sourced database of Wi-Fi hotspot and cell
    tower data that is generated by tens of millions of iPhones sending the
    geo-tagged locations of nearby Wi-Fi hotspots and cell towers in an
    anonymous and encrypted form to Apple.
     
    Alan Browne, Dec 31, 2014
    1. Advertisements

Ask a Question

Want to reply to this thread or ask your own question?

You'll need to choose a username for the site, which only take a couple of moments (here). After that, you can post your question and our members will help you out.