[rescue] Poor Man's WiFi extender
rjw at alembic.com
Tue Oct 26 00:04:59 CDT 2004
> Message: 9
> Date: Mon, 25 Oct 2004 01:00:38 -0400
> From: Phil Stracchino <alaric at caerllewys.net>
> Subject: Re: [rescue] Poor Man's WiFi extender
> To: rescue at sunhelp.org
> Message-ID: <20041025050038.GA9081 at prydain.caerllewys.net>
> Content-Type: text/plain; charset=us-ascii
> On Sun, Oct 24, 2004 at 09:53:22PM -0700, Ron Wickersham wrote:
> > so far as i know the WiFi devices that can do diversity in the real world
> > are receive only diversity, and due to header timing, stick to only 802.11b
> > devices (802.11g just can't do diversity). i'd recommend the Senao
> > NL-2511CD Plus which has +23dBm output (200mW) and good receive sensitivity.
> > we use them extensively in a community network in Sonoma County, Calif, USA
> > (albeit not in diversity as ships floating in the hills are exceedingly rare).
> Have you considered using WaveRider gear instead? It's NLOS, so passing
> ships shouldn't disrupt it.
> (Of course, early next year you should be able to buy 802.16a hardware,
> which is much faster and is also NLOS...)
WaveRider sounds great but the cost of their "starter pack" for 5 stations
costs $1000/station and our cost is less than a tenth of that scrounging
old hardware and at times home-brew antennas. we are limited to line-of-
sight but get more than twice the bandwidth and some of our stations are
more than 3 times the range quoted in the WaveRider spec.pdf. if we
were a "commercial" operation and paid for labor, things would be different
but we are all unpaid volunteers and just having a good time bringing
wideband to our rural area where the phone company doesn't have an interest
in providing service and the cable company has only provided service in the
last couple of months that cover just a few of the locations that are close
my motive in posting was to share our experience and point out that even
though 802.11g cards are available with 2 antenna connectors for diversity,
the scheme doesn't work if the stations are separated more than a couple of
hundred feet. that is not the case with 802.11b cards and the longer
header (in microseconds) gives the receiver time to check both antennas and
choose the better one and settle down before the payload comes along. and
the higher power of the cards we use for longer connections might be enough
to make a difference for the over-water link in the shadow of a ship.
maybe not, but 8dB more signal certainly makes a difference when you have
a marginal signal.
looking forward to the introduction of 16a gear and the possibility that
some manufacturers will be able to sell cards that are not as expensive
as the WaveRider stuff.
More information about the rescue