Le 21/01/2020 à 08:15, Stewart Bryant a écrit :
Given that when I turn off Wifi, my phone asks me to turn it back on to
improve location accuracy there must me something similar already deployed.
That is a problem of the GUI. The GUI gives a wrong sense.
The WiFi improves accuracy of location to the extent it helps the phone
to access the google database which contains a list of locations of
Access Points (by MAC) address. Locally measuring the power it is
possible to determine a rather fuzzy location based on WiFi AP
positions, but which is more than nothing (nothing is present when GPS
is absent).
In a sense yes, 'improve location accuracy' when GPS is absent or very
bad (e.g. DOP higher 3 or 5), can be obtained with WiFi.
Similar to WiFi there is also location based on GSM positioning. It's
still very fuzzy, but better when there is no GPS coverage.
Absence of GPS coverage also has multiple meanings. In some places GPS
satellite dont reach because of the place being wrong (e.g. underground
and without GPS repeaters); but other times GPS is absent because of
human desire (check exercise military Florida recent nanog post).
Moreover, GPS coverage might be less preferred to be used because of
other Layer-8 reasons, where one might prefer Galileo, Glonass, Beidou,
others. These reasons are even further fuzzied by a possibility where
in _some_ of these places, GPS might be preferred in complimentarity,
whereas in others no. For example, use Galileo complimented with GPS
might give more reliable results, yet nobody talks about complimenting
Galileo with Glonass, for example, or complimenting them all into one,
despite being often all available in one place.
GUI expression is hard to tell in just one small screen with all these
factors.
I think that the current system is based on using the SSIDs which have
the advantage that they are a broadcast signal and do not require that
you have permission to log on to the WiFi.
YEs.
In the underground case, providing the map on the phone with the
stations SSID would be a passive alternative which I suspect is already
deployed, and is more in the spirit of the GPS approach.
In this, you mean 'GPS' as the device my car carries. The car also has
odometer (dented wheel that counts turns of bigger wheel). Smartphones
dont have odometers. Cars with GPSs and good maps, like submarines, can
maintain very accurate positioning when GPS satellite are invisible for
longer time and distance. Cars and submarines also use additional
sensing of their environment, like in 'ranging', which helps them build
maps, and then orient, build again, until a near perfect localisation is
obtained.
Smartphones might use their inertial central to act when GPS is
invisible but the trials of in-building movements dont exhibit too high
accuracy. Smartphones are carried in pockets - cant see, can't
'ranging'. They might use the watches on wrists, but rarely can these
see either.
Alex
- Stewart
Sent from my iPad
On 20 Jan 2020, at 17:22, Rob Brew <[email protected]> wrote:
Dear all.
I have submittetd the following draft RFC for approval.
Obviously whilst underground mobile phone users at public locations,
such as the London underground, will not be able to know their GPS
location, as the GPS signals cannot penetrate the earth. I have
submitted a RFC draft at
https://tools.ietf.org/html/draft-brew-intarea-underground-01 to
bypass this problem, as well as a server application in Java at
https://github.com/rydal/underground .
As we know the IP Address contacting the server the server contains a
map containing the GPS location of each underground site, as
referenced by IP address. The returns the phsycial GPS location and
name of the site as a JSON array.
To prevent false servers reporting inaccurate information https can be
used to verify the authenticity of the server.
Providing this service to mobile phone user's without a current WIfi
connection i propose going one stage further, create a hidden wireless
network wih the name ".location". In cases where no GPS location can
be found a mobile phone's GPS system can be programmed to seek such a
network providing the same service.
This would require alterations to th servics as provided by mobile
phone providers such as Apple and Android.
If you are interested in this concept please review it (and my code),
and maybe it will become something.
For a video demonstration of this service please check the link at the
github site.
Thanks,
Rob Brew.
_______________________________________________
Int-area mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/int-area
_______________________________________________
Int-area mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/int-area
_______________________________________________
Int-area mailing list
[email protected]
https://www.ietf.org/mailman/listinfo/int-area
