[CUWiN-Dev] G and A and realworld throughput

[David Young]

On Tue, Jul 12, 2005 at 01:40:08AM -0500, Stelios Valavanis wrote:
> please copy me on replies as i'm not on the developer list.
> 
> here in chicago we're about to deploy a 20 node cuwin mesh backhaul network 
> with cisco APs shooting out to the end users. the ciscos are getting donated 
> fyi but i did get the metrix kits with dual radios for when cuwin can do dual 
> radio.
> 
> anyway i was warned that using A radios is not tested. i've got atheros a/b/g 
> cards for which i believe there are drivers. i'd like to use 802.11a because 
> the spectrum is less crowded and there are more channels to choose from. so 
> my first question is whether or not i will have problems with using A cards.

Stelios,

I cannot think of any reason the driver should not work for 802.11a.
I have some a/b/g cards on our indoor testbed.  I will give .11a a try
sometime next week.

The problem I anticipate is a lot more mundane than device driver
compatibility: we set the operating mode to 802.11b at "compile time."
You have to fiddle with some scripts to use any other mode.  To use
802.11a, you have a few choices:

        * compile-time configuration: you can build a CUWiN distribution
          that sets the mode to 802.11a, instead.
          (Change the text 'mode 11b' to 'mode 11a' in
          cuw/trunk/src/boot-image/extras/etc/ifconfig.wireless.tmpl.)
          CUWiN can help you with that.

        * you can change our web configuration widget so that it lets
	  you program the operating mode (.11a, .11b, or .11g).  CUWiN
	  can also help you with that.

When you add a second radio, things get a little bit more complicated,
because presumably you want to operate them on different channels, or
even on different bands.  The first problem is that the web configuration
widget only works on one radio; that will have to be changed.  It's
not hard, just nobody has done it, yet.)  You can do "compile-time"
configuration of multiple radios, but you have to make assumptions
about the order that radios are enumerated---that is, you have to take
care that radio #0 and radio #1 are not reversed in your Metrix boxes,
or else that they are not reversed in your configuration files.

(Someday I want to automate the channel & mode selection, based on the
channel characteristics and information gleaned from the routing protocol.
This is a long way off, of course!)

> my second question is about real-world throughput. obviously i'm not going to 
> get all 54Mbps because of network overhead but how does that end up after 3 
> or 4 or 10 hops to get to your uplink? my understanding is that the bw 
> doesn't halve with every hop like WDS does.

I remember reading about a 9-hop 802.11b trial in Colorado that was
usable, but slow.  WDS versus not-WDS does not matter, the problem is
that there is one radio, and it's half-duplex, so you cannot "pipeline"
a packet stream.  You can be certain that after 3 hops, you will not get
more than 1/3rd the 1-hop bandwidth, but I have no reason to believe the
bandwidth decline is exponential.  I will not recommend going more than
3-4 hops until I have experimented with long paths, myself.

Dave

-- 
David Young             OJC Technologies
dyoung at ojctech.com      Urbana, IL * (217) 278-3933