Given the high sample rate, it sounds like you are trying to track
a flying object (rocket, R/C airplane or helicopter, etc.)

Is the transmitter CW (continuous wave)? I suspect not since you are using
the ISM band and I'm guessing that your target has a Wi-Fi node hanging
on it as transmitter. If so, you will have to be able to handle interference
from other ISM devices (bluetooth, Wi-Fi (802.11x), leaky microwave ovens,
wireless phones and video links), so some kinematic filtering of the incoming
data should be done.

How far apart are your receivers? Given your operating
frequency you can probably get away with phase detectors to get
a bearing and count slipped cycles to get distance.
Lambda is 12.5cm at 2.4Ghz so typical 5 degree phase accuracy
will get you about 0.2mm ranging.
You can use the chinese remainder theorem to resolve ambiguities
if you can transmit and process a second frequency simultaneously.

What velocity of motion are you tracking?
Is shadowing of the emitting antenna likely?
Is the area likely to be free of RF reflections (which would affect
phase measurements but not time of flight)?

I have looked at doing this for model rocketry tracking, but it is
a non-trivial problem.

Robert

Marcel van Lieshout wrote:
>
> I believed my question was detailed enough, but I now do agree with you.
>
> I am trying to figure out if it is feasible to design a local positioning system (3D).
> Because I cannot set/measure the clock on the transmitter, I thought I best use the TDOA principle.
> I would therefore like to measure time of flight differences with an accuracy of as little mm as possible
> (max 10 mm).
> The RF frequency is 2.4 Ghz
> The maximum distance is 50 mtrs.
> A single transmitter at a time is active
> At least four receivers are present
> RealTime results are needed
> Several thousand (2500) measurements per second are needed.
> The acual location calculations are no problem, just a metter of enough cpu-cycles ;-)
>
> As I am basically a software-guy, I am not sure what information is relevant. Please, keep asking.
>
> Thank you for thinking with me!
>
> Marcel
>
> Robert Rolf wrote:
> > Over what distance range? At what carrier frequency?
> > Phase difference measurement can get you there as can various forms of
> > interferometry if the signal is CW.
> >
> > Look at how geodetic GPS can get down to mm with relatively cheap silicon.
> >
> > Does the result have to be supplied in real time?
> >
> > The CORRECT way to access the list's knowledge is to pose
> > the full problem so that our replies fit your constraints.
> >
> > Robert
> >
> > Marcel van Lieshout wrote:
> >> I'm trying to measures RF time-of-flight up to a few mm accuracy ( 10mm would be fine, 1 mm
> >> would be fantastic). Well, because I have no influence on the transmitter, I actually want to
> >> measure Time-Differences-Of-Arrival at several receivers.
> >>
> >> What do you (or anybody else) think?
> >>
> >> Marcel
> >>
> >> Jinx wrote:
> >>>> Hmm, not fast enough. Looking for something in the low picoseconds
> >>>> switching times :-0
> >>>
> >>> ECL 100K is probably the fastest readily-available family, and 25GHz
> >>> (40ps) operation is the exception rather than the norm
> >>>
> >>> Googling around, it appears anything over that is very specialised, not
> >>> what you'd call a "family". Logic at up to 350GHz (2.8ps) appears to be
> >>> possible but in the experimental domain
> >>>
> >>> <pulling teeth question time> ;-)
> >>>
> >>> Do you actually need ps speed, or is there a way that what you're trying
> >>> do can be done by inference ? For example, in the way that high speed
> >>> oscilloscopes sample to build a waveform ? Can you improve what you
> >>> have or can get with liquid nitrogen for example ?
> >>>
> >>> </pulling teeth question time>
>
> --
> http://www.piclist.com#nomail Going offline? Don't AutoReply us!
> email listserv@mitvma.mit.edu with SET PICList DIGEST in the body

--
http://www.piclist.com#nomail Going offline? Don't AutoReply us!
email listserv@mitvma.mit.edu with SET PICList DIGEST in the body