Re: nearest router
Mark Jeffcoat wrote:
stephane.chouinard.ptelebel@gmail.com writes:
Hi,
I will try with a network newsgroup, but I sent my question here
because i'am doing an homework in java. The reason I would like to know
the first router (nearest one) is because i want to be able to know
where is the person with the pocket pc. I was supposed to do something
with infrared access point but the project aborded. So i was thinking
that maybe it was possible to do something with the ip data from the
network. Like, my pocket pc can send a packet in a specified interval
of time to a server and the server or the pocket pc could determine the
position with the info. from the nearest router. In the university we
are using wireless network...
Wow. Are you trying to triangulate the position of the
handheld computer by timing the response from multiple
wireless routers? Because that would be a sexy project,
though extremely ambitious if you don't already have a
rough idea of what you're doing.
I don't think signal travel times would be very good. As well as the
issues you mention, any reflections would really mess things up. Most of
the systems I've read about or been involved in use either signal
strengths, or the set of access points with signals above a threshold.
Of course, GPS does use signal travel time, but the beacons are very
accurate clocks, and transmit timing signals so that a GPS receiver can
measure the differences in travel times extremely precisely.
Seems unlikely that the the timing difference would
map very well to physical differences, though, since
the actual travel time of the signals is negligible.
Maybe you'd end up with some sort of interference map?
Ping estimates timings with ICMP, which Java definitely
doesn't natively support. You'd need to go through JNI,
or find some library that wraps the JNI for you.
[Sigh. On my fourth reading of the post I'm replying
to, I've decided that the original poster probably
just wants to figure out, say, which building the
pocket pc is in based on which router it's connected
to. I just can't throw away the (hopeless) triangulation
idea; I live in a house with two wireless routers and
a Treo I lose all the time.]
There has been a lot of research on finding location from wireless
access point measurements. The Placelab publications page,
http://placelab.org/publications/, is a good starting point for seeing
the current state of the art.
Usually, location is driven off signal strengths, or the set of visible
beacons. Millions of access points have already been mapped - see
http://wigle.net/
Indoor location in office buildings does have problems with reflections
and variations in signal strength due to walls, floors, metal beams, etc.
The ActiveCampus project, which does indoor location using 802.11
wireless access points, has arrangements for user-entered corrections to
deal with some of those problems.
For a general description of ActiveCampus, see W. G. Griswold, P.
Shanahan, S. W. Brown, R. Boyer, M. Ratto, R. B. Shapiro, and T. M.
Truong, "ActiveCampus - Experiments in Community-Oriented Ubiquitous
Computing", IEEE Computer, Vol. 37, No. 10., pp. 73-81, October 2004.
Generally, 802.11 works much better than cell phones, but cell phones
are rapidly displacing 802.11 equipped PDAs as the common handheld
device. There is some hope for combined cell and 802.11 devices being
able to find themselves accurately.
Patricia
[Who should be getting on with her ubiquitous computing research instead
of reading newsgroups.]