Thank you David for your thoughts on this. I too am a bit
skeptical. :) Which is why I decided to ask for comments from you
guys. There is the additional issue of "mesh" routing protocols but
I think (and I mean think) I may have that covered if the physical
stuff worked out ok. It's just the whole concept might fall flat for
a dozen reasons. I'm trying to figure out how many nodes I would
need to deploy in a real world test, but I also want to think this
out before getting that far and spend a bunch of money. The idea may
be so bad that testing is unwarranted.
But hey, there must be at least 8 people left in the country without
broadband, and I want to give it to them! ;)
Allen
At 01:05 PM 9/10/2007, David E. Smith wrote:
Allen Marsalis wrote:
I take it that nobody has ever built a 900MHz NLOS mesh network before.
Which is not a good sign to me. That's a sign that my idea
probably won't work.....
I'd be very skeptical just because of what I lovingly call the
"Tropos Effect."
Obviously, all these nodes eventually have to come back to...
somewhere that has a big bad Internet connection. Your office, a
central tower, whatever. If you're near that tower, you don't have
much of a problem, as your laptop is talking to a node that's
talking directly to that point of origin. If you're a few blocks
away, where your laptop talks to a node that's two or three hops
away, there's cumulative bandwidth loss and added latency, and just
more things that can go wrong generally.
Your proposal gets rid of the worst part of how Tropos does things.
They use the same radio both for inter-node communication and for
customers, same SSID, same everything; by using separate radios for
backhaul and customer access, you're already coming out ahead.
There will still be added overhead and latency, the more nodes you
have to go through, and the folks at the farthest reaches of the
network won't have as good an experience as the folks close to your
point of origin.
I'm a bit skeptical. The expense of 900MHz gear, and the sheer
number of units you'd need to for a wide coverage area, makes this
seem like a really difficult idea to pull off. Nevertheless, I wish
you luck, if you do choose to deploy something like that.
David Smith
MVN.net
--------------------------------------------------------------------------------
** Join us at the WISPA Reception at 6:30 PM on October the 16th 2007 at ISPCON
**
** ISPCON Fall 2007 - October 16-18 - San Jose, CA www.ispcon.com **
** THE INTERNET INDUSTRY EVENT **
** FREE Exhibits and Events Pass available until August 31 **
** Use Customer Code WSEMF7 when you register online at
http://www.ispcon.com/register.php **
--------------------------------------------------------------------------------
WISPA Wants You! Join today!
http://signup.wispa.org/
--------------------------------------------------------------------------------
WISPA Wireless List: wireless@wispa.org
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
