[Cu-wireless] general update:

[David Young]

One problem with using omni antennas and 802.11b on the Race Street
corridor is that throughput will suffer. I don't think that this should
stop us from experimenting with omni's, not even on the corridor, but
it is something to keep in mind.

Say that there are stations at Zach's, Maiko's, and Sascha's (henceforth
stations Z, M, and S). Say that Z is out of range for S, but M can talk
to both Z and S.

When Z broadcasts to M, M is silenced, and M silences S. When M
broadcasts, S and Z are both silenced. Imagine S wants to send a packet
to Z through M. S sends an RTS, silencing M, who sends a CTS for S,
silencing Z. S sends to M. M wants to forward the packet to Z, now.
M sends an RTS, silencing S and Z. Z sends a CTS. M sends the packet.
To send a new packet, this process must be repeated.

Compare with the situation where we are using two transceivers with
directional antennas on different channels. You have a "pipeline,"
now, where M can receive one or more packets from S simultaneously with
transmissions to Z.

Again, I don't think that this sort of concern should stop us, now.
Also, those omnis will be super-useful in oodles of applications.
But they are not interchangeable with our directional antennas.

BTW, using Kumar's SEEDEX algorithm to arbitrate media access, omnis
might serve us better. I think we are a long, long way from using SEEDEX,
though.

Dave

On Wed, Mar 06, 2002 at 07:40:43AM +0100, by way of Illustrious niteshad <niteshad at whopper.de> wrote:
> One way to eliminate much of the complexity and expense incurred in our
> currrent plan for Race Street would be to use high gain omni-directional
> antennas.  Guerrilla.net has plans for a 10 dBi vertical collinear array,
> unfortunately, their mathematical theory section is so full of holes that a fishnet
> looks solid by comparision (e.g. they list the speed of light as 300000 m/s,
> it is actually 300000000 m/s) With such basic errors, I do not trust their
> numbers, or their expertise in physical/radio theory.  I am currently
> scouring the ARRL Antenna Handbook for the proper theoretical underpinnings of the
> vertical collinear array.  Once the theory is firm in hand, I could
> probably build two of these antennae for between $20 and $40.  The primary
> radiating elements are thin brass pipe and LMR-200 or LMR-400 co-ax cable. 
> (ecommwireless sells LMR-200 for _$0.28/ft._!)  
> 	If we replace the 12 dBi Pringles Can yagis with 12 dBi omni-directional
> (omni) antennas, we save cost on our network.  At each node, we need only
> one wireless card, rather than 2 at Maiko's, 2 at Sascha's and 2 at Race and
> Main.  According to my back of the envelope calculations, it will cost
> approx. $944 to run wireless from Zach's to the IMC using weatherproof Pringles
> cans, pigtails, free routers and $80 wireless PCI cards. (I assume that a
> weatherproof Pringles Can costs about $12, the pigtails are known to cost $26).
> 	Using slightly more expensive omnidirectional antennas, the cost actually
> drops to approx. $655.  This is due to the fact that only one antenna, and
> thus one costly wireless card and pigtail, are needed at each node, rather
> than two.  In other words, the more expensive antennas save us 30% of the
> cost in the long run.  I should hasten to add that the $655 figure was arrived
> at by estimating that the omni antenna will cost around $25 to build and
> weatherproof, wireless cards cost $80 and pigtails cost $26.  
> 	While I favor the omni in theory, I realize that the goal is to get the
> network up and running as quickly as possible.  Therefore, I suggest that we
> attempt to move forward with the proven Pringles Can design (actually, we
> might want to test the Coffee Can design, as it's simpler, cheaper and
> inherently waterproof.)  If the collinear array proves itself to be superior in
> performance, we can relatively easily replace the directional antennas with
> omnis.  Then we can redeploy the directional antennas and surplus wireless
> cards elsewhere in the community.  Possibly to people within line-of-sight of
> our network who lack the income for internet access.  
> 
> best regards,
> 
> Mark 
> 
> -- 
> Sent through Global Message Exchange - http://www.gmx.net
> 
> 
> _______________________________________________
> Cu-wireless mailing list
> Cu-wireless at lists.groogroo.com
> http://lists.cu.groogroo.com/cgi-bin/listinfo/cu-wireless

-- 
David Young                   On the Job Consulting
dyoung at onthejob.net     Urbana, IL * (217) 278-3933