Wouldn't a "leaky wire" solution magnify the multipath probs? Especially
for a moving target?
At 03:51 PM 10/4/00 -0500, Brooker, Ralph wrote:
>It's not exactly using the tracks, but Andrew has provided many
>communications systems for rail networks using Radiax (R) radiating coaxial
>cable laid alongside the rails. The cable has slots tuned for specific
>bands such as VHF, cellular, GSM, or PCS, and has a tapered radiation loss
>characteristic to give even signal strength along the line. Every so often
>you put a repeater or network access point. Most of these systems have been
>in underground metro tunnels, such as in Hong Kong where commuters talk on
>their GSM phones through this network. However, it does get rather
>expensive for a network hundreds of miles long, and the unusual propagation
>and fade charateristics can limit the bandwidth.
>
>Ralph
>
>-----Original Message-----
>From: Charlie Younghusband [mailto:[email protected]]
>Sent: Wednesday, October 04, 2000 10:26 AM
>To: Andreas Timm-Giel; [email protected]
>Subject: Re: AW: Moving Targets
>
>
>At $7 US/minute @ 64kbps, that's $420 /hr @ 64kbps or 6720 US $ / hr @
>1Mbps.
>Plus hw & ISP costs for down under. Yikes. Your research project could end
>pretty quickly with an Inmarsat type solution.
>
>Want a real research project? Find a way to send the signal through the
>train
>tracks. ;) Only way to really fix the tunnel problem!
>
>Seriously though, I'd talk to some of the wireless providers in Australia.
>Australia is quite advanced at digital wireless services, they might be all
>over
>an additional excuse to expand their east coast network where most Aussie
>cities
>are anyway.
>
>Charlie
>
>Also note that a 1Mbps downstream and a 56 upstream probably isn't a good
>matchup (an offhand guess); even if you're doing 100% downloading of files,
>your
>upstream bandwidth could quickly become the bottleneck just with ACKs
>assuming
>normal windows TCP/IP. Something closer to a 10 to 1 ratio might be more
>appropriate. Check some of the litterature regarding assymetry and TCP/IP.
>
>
>
>Andreas Timm-Giel wrote:
>Second problem for L- and
>
> > S-band systems will be the prices for capacity on satellites (just as a
> > reference: Inmarsat takes around US$ 7 / min for a 64 kbit/s link)
>
> > >
> > > Hi,
> > >
> > > I'm doing a research project about providing Internet access to moving
> > > targets on the eastern seabord of Australia. In this case, the moving
> > > targets are trains.
> > >
> > > We really need to provide around 1MB or so downstream, maybe 56k or so
> > > uplink. I've been doing a lot of searching but facts seem to be hard
> > > to come by.
> > >--
>
>Charlie Younghusband
>Network Systems Engineering
>Xiphos Technologies http://www.xiphos.ca/
>514-848-9640 (f) 514-848-9644
This archive was generated by hypermail 2b29 : Wed Oct 04 2000 - 20:18:07 EDT