Ad-Hoc Mode Only

I just read in a post by Jeff Liebermann that ad-hoc mode only supports
9 nodes at time.

Why is this?

I would like to create an ad-hoc network that allows an essentially
unlimited number of nodes present.

What low-level exchanges are involved in an ad-hoc network. I would
also like to know what configuration of the transceivers would allow
the *absolute miminum* delay between the time N nodes present
themselves into a space and the time that they can start sending data
frames between each other.

Note that I am only interested in what happens at the link-layer and
below. Above the link-layer, I am not concerned.

Thanks,

-Le Chaud Lapin-

(E-Mail Removed) hath wroth:

>I just read in a post by Jeff Liebermann that ad-hoc mode only supports
>9 nodes at time.
>Why is this?

There's no real limitation in the 802.11 spec to the number of nodes.
There may be some limit to bridging table sizes, but nothing inherent
in the specification. Microsoft recommends 9 clients maximum, but
I've connected many more without much difficulty.
| http://pdos.csail.mit.edu/papers/gri...om01/paper.pdf

Ad Hoc is sometimes called peer to peer mode or Independent Basic
Service Set (IBSS). If you don't mind having your brain turned to
mush, read IEEE 802.11-1999.
| http://standards.ieee.org/getieee802...02.11-1999.pdf

>I would like to create an ad-hoc network that allows an essentially
>unlimited number of nodes present.

Unlimited is a very large number. Please note that wireless mesh
networks are essentially a collection of ad-hoc networks.

>What low-level exchanges are involved in an ad-hoc network.

The IBSS protocol is defined in IEEE 802.11-1999.

>I would
>also like to know what configuration of the transceivers would allow
>the *absolute miminum* delay between the time N nodes present
>themselves into a space and the time that they can start sending data
>frames between each other.

That's going to be a difficult. Optimimum timing for large systems is
usually provided by synchronous intelligent polling. However, the FCC
clearly specifies that there is to be no synchronization between Part
15.247 radios. That's to insure that one system does not hog all the
airtime in an area. That also means you're going to have collisions
and timing issues no matter what you do.

Hint: Search Google for "queueing model 802.11 ad hoc". There has
been quite a bit of research on optimizing wireless systems.

Also, you might want to look at:
http://monarch.cs.rice.edu
http://monarch.cs.rice.edu/papers.html
which has a large number of wireless mobile ad-hoc papers.

>Note that I am only interested in what happens at the link-layer and
>below. Above the link-layer, I am not concerned.
>
>Thanks,
>
>-Le Chaud Lapin-
--
Jeff Liebermann (E-Mail Removed)
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558

Jeff Liebermann wrote:
> Ad Hoc is sometimes called peer to peer mode or Independent Basic
> Service Set (IBSS). If you don't mind having your brain turned to
> mush, read IEEE 802.11-1999.
> | http://standards.ieee.org/getieee802...02.11-1999.pdf

Great info. I've visited the Monarch site a few times over the
years.

I started reading that each night before bed time. I searched the doc
for ad-hoc and came up with some links that did not say what was
involved. I tried to cheat by reading the overview only (Chapter 5)
and again, the information was light in nature. The only thing I am
concerned with is the necessary minimum exchange required before two
ad-hoc nodes can send frames to/from each other.

I would also like to know, aside from RSSI, what harware indicators
there are to support notification of a (perceived) change in PHY state.

I'm still going to read the 802.11 spec, but in truth, these are the
only two pieces of information that I need to determine whether what I
am trying to do will work.

-Le Chaud Lapin-