Pooling protocol

[torjon]

Did you plan to make pooling protocol in v3 like Proxim WORP or Mikrotik ?
It's very important with many 5GHz clients in noisy environment (hidden node problem).

[lonnie]

Polling is not something we have near the top of the list. We do not feel it is as big a problem as people say. In the old Karlnet days people used to tell me they had better results with larger systems when they turned off polling.

[ninedd]

Hi. What about RTS/CTS in V3?

[lonnie]

RTS/CTS is on the list for near term addition.

[budi]

Polling is not something we have near the top of the list. We do not feel it is as big a problem as people say. In the old Karlnet days people used to tell me they had better results with larger systems when they turned off polling.

Dear Lonnie,

Do you ever try to install V3 back-to-back with Several
Motorola Canopy, and you couldn't use 5 Ghz because legal issue ?

Even OSBridge had polling & disable csma/ca in their
firmware.

http://www.osbridge.com/download/CHANGELOG_5Gi.txt

Thx
//Budi

[luka]

Even OSBridge had polling & disable csma/ca in their
firmware.
I would like to hear concrete answer regarging this topic too...

[lonnie]

What "correct" answer do you want? I am not going to dispute that osbridge has polling or any other feature, so the "statement" goes unchallenged. It was not really a question.

I would like to hear correct answer regarging this topic too...

[lonnie]

I've never faced a Canopy system and I have no knowlegde about it. I do think that cloaking will be effective, but that is only from several customers who have tried it.

Dear Lonnie,

Do you ever try to install V3 back-to-back with Several
Motorola Canopy, and you couldn't use 5 Ghz because legal issue ?

Even OSBridge had polling & disable csma/ca in their
firmware.

http://www.osbridge.com/download/CHANGELOG_5Gi.txt

Thx
//Budi

[luka]

You were so fast that I havent time to edit post and change the word.

I meant concrete, is there a concrete reason why do you think polling/csma thing is not much important. Do you have implemented in Staros some other strategies for interference mitigation and avoidance? I dont want to know your internals, just if is there other thing which may eventually work as or better than proposed solutions?

In the last time many other sw vendors are working on some polling and similar mechanisms. For instance, there is some test at http://aqua.comptek.ru/test/HiddenNode/hidden_node_en.html which emulates clients and hidden node problem in 802.11 network. The worse situation can be in area with mixed standards, and interference problems and 802.11 weakness from other "more aggressive" products is the problem probably many of users are facing.

[lonnie]

We have not really been a polling fan over the years. I have seen it mainly as a "feature" that is extolled to correct all of the problems associated with 802.11 systems. In reality I have talked with many operators who remove polling and obtain better performance.

I know from first hand troubleshooting that many people create their own problems with poor LAN design and they further compound that with bad cabling (both RF and cat5).

I have helped many people to build a network using my methods and techniques and they simply do not have the problems that the rest of the world is saying that polling will fix.

You were so fast that I havent time to edit post and change the word.

I meant concrete, is there a concrete reason why do you think polling/csma thing is not much important. Do you have implemented in Staros some other strategies for interference mitigation and avoidance? I dont want to know your internals, just if is there other thing which may eventually work as or better than proposed solutions?

In the last time many other sw vendors are working on some polling and similar mechanisms. For instance, there is some test at http://aqua.comptek.ru/test/HiddenNode/hidden_node_en.html which emulates clients and hidden node problem in 802.11 network. The worse situation can be in area with mixed standards, and interference problems and 802.11 weakness from other "more aggressive" products is the problem probably many of users are facing.

[budi]

We have not really been a polling fan over the years. I have seen it mainly as a "feature" that is extolled to correct all of the problems associated with 802.11 systems. In reality I have talked with many operators who remove polling and obtain better performance.


Dear lonnie, the problem we're saying isn't created by
user / human error / design error (like : connector,pigtail,ps,etc)
but inheritanced problem from CSMA/CA protocol aproach that
used in 802.11x wireless chipset (Atheros,prism,broadcom,realtec,philips,etc)

The CSMA/CA protocol aproach to medium sharing (frekuensi)
is designed for indoor application where we could assume :
- We could hear all node
- All the devices is operating in the same standard using CSMA/CA.

The approach is like an boy scout :
- polite & honest (they check first if the medium is idle for x
time (difs waiting time) and transmit if the medium is idle
for more than difs time

But in the real world (frequency / life) you would get same
result / conclusion :

- Not all the people / radio system is honest, or
polite. Many radio system behaviour is very rude
and unpolite. Example like : Motorola Canopy system.

If 802.11x based radio system is put head to head
to this rude radio system, you would be waiting and waiting
for chance to transmit, that maybe never come :-)

Because Moto Canopy system is always transmit all the time,
not give a crap, even if the medium is busy.

This is the problem that many your user would facing, especially
in urban areas, where many big WISP or goverment agency
deploy Moto Canopy system indiscriminately.

And many of user because legal issues, couldn't use 5 GHz
freq.

And they would be forced using the same system in order
to coexist with that rude wireless system.

If you don't believe my word, please borrow some Moto Canopy
System, and please deploy it at near your AP / Backhaul using
the same freq / or nearby frequency.

You would scratch your head, with that kind of interference / noise.

Thank you very much for your consideration.

PS :
The osbridge has been acknowledge this problem, and
disable csma/ca in their wireless system and replace it
with polling based system.

We need more interference robustness in the event to
enable us coexist with Canopy based system.

A few hundred kbps more b/w is meaningless,
if its drop to nothing when faced with fierce / rude
Wireless system.

High Regards
Budi