Inter-operability of Draft-n equipment

Finalisation of 802.11n is still reportedly over a year away. There is
of course no guarantee (actually, little likelihood) that any of the
draft-n gear available now will be upgradable to the standard, when it
comes.

The conventional wisdom for using draft-n equipment at the moment is to
get all the pieces from the same manufacturer, to make sure they work
together as intended. But that's not always possible.

Have any of the computer or networking magazines, or any of the
consumer-protection testing labs, run any studies comparing how well (or
how poorly) 'mixed' environment networks run? Things like 'The A pcmcia
card works with B's router but not C's access point, while D's pci
adapter works with routers from A and C but not B', etc., etc.

Given that there are only, what?, four or five or six major
manufacturers of consumer- and SOHO-level equipment, it shouldn't be too
hard to put the more popular models through their paces.

cheers,

Henry

On Sat, 9 Feb 2008 16:07:24 +0200, (E-Mail Removed) (Henry) wrote in
<1ic1w1x.7afdfc1i6elhpN%(E-Mail Removed)>:

>Finalisation of 802.11n is still reportedly over a year away. There is
>of course no guarantee (actually, little likelihood) that any of the
>draft-n gear available now will be upgradable to the standard, when it
>comes.
>
>The conventional wisdom for using draft-n equipment at the moment is to
>get all the pieces from the same manufacturer, to make sure they work
>together as intended. But that's not always possible.

Why not?

My own conventional wisdom is to stick to 802.11g gear. Why pay a
premium for uncertain value? When more signal is needed, better
antenna(s) are often a better bet.

>Have any of the computer or networking magazines, or any of the
>consumer-protection testing labs, run any studies comparing how well (or
>how poorly) 'mixed' environment networks run? Things like 'The A pcmcia
>card works with B's router but not C's access point, while D's pci
>adapter works with routers from A and C but not B', etc., etc.

I've seen some small tests, but don't know anyone that's undertaken the
gargantuan task of wide interoperability testing.

>Given that there are only, what?, four or five or six major
>manufacturers of consumer- and SOHO-level equipment, it shouldn't be too
>hard to put the more popular models through their paces.

Actually very hard to get all the gear and test it properly, and hard to
see how it would pay off.

--
Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>

Henry wrote:
> Finalisation of 802.11n is still reportedly over a year away. There is
> of course no guarantee (actually, little likelihood) that any of the
> draft-n gear available now will be upgradable to the standard, when it
> comes.
>
> The conventional wisdom for using draft-n equipment at the moment is to
> get all the pieces from the same manufacturer, to make sure they work
> together as intended. But that's not always possible.
>
> Have any of the computer or networking magazines, or any of the
> consumer-protection testing labs, run any studies comparing how well (or
> how poorly) 'mixed' environment networks run? Things like 'The A pcmcia
> card works with B's router but not C's access point, while D's pci
> adapter works with routers from A and C but not B', etc., etc.
>
> Given that there are only, what?, four or five or six major
> manufacturers of consumer- and SOHO-level equipment, it shouldn't be too
> hard to put the more popular models through their paces.
>
> cheers,
>
> Henry
I gather you have no faith in the "Wi-Fi Alliance".
<http://certifications.wi-fi.org/wbcs_certified_products.php?search=1&advanced=1&la ng=en&filter_company_id=&filter_category_id=&filte r_subcategory=&filter_cid=&date_from=&date_to=&x=3 0&y=18&selected_certifications%5B%5D=33>

On Sat, 09 Feb 2008 16:20:46 +0000, LR <(E-Mail Removed)> wrote in
<(E-Mail Removed)>:

>Henry wrote:
>> Finalisation of 802.11n is still reportedly over a year away. There is
>> of course no guarantee (actually, little likelihood) that any of the
>> draft-n gear available now will be upgradable to the standard, when it
>> comes.
>>
>> The conventional wisdom for using draft-n equipment at the moment is to
>> get all the pieces from the same manufacturer, to make sure they work
>> together as intended. But that's not always possible.
>>
>> Have any of the computer or networking magazines, or any of the
>> consumer-protection testing labs, run any studies comparing how well (or
>> how poorly) 'mixed' environment networks run? Things like 'The A pcmcia
>> card works with B's router but not C's access point, while D's pci
>> adapter works with routers from A and C but not B', etc., etc.
>>
>> Given that there are only, what?, four or five or six major
>> manufacturers of consumer- and SOHO-level equipment, it shouldn't be too
>> hard to put the more popular models through their paces.
>>
>> cheers,
>>
>> Henry
>I gather you have no faith in the "Wi-Fi Alliance".
><http://certifications.wi-fi.org/wbcs_certified_products.php?search=1&advanced=1&la ng=en&filter_company_id=&filter_category_id=&filte r_subcategory=&filter_cid=&date_from=&date_to=&x=3 0&y=18&selected_certifications%5B%5D=33>

The point is that n can't be certified until it's approved.

--
Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>

John Navas wrote:

> The point is that n can't be certified until it's approved.
>
While that is stating the obvious, apart from the OP's first sentence
the post was about draft n. "Inter-operability of Draft-n equipment"

(E-Mail Removed) (Henry) hath wroth:

>Finalisation of 802.11n is still reportedly over a year away.

Actually, the slipping schedule seems to be a continuously moving
target that always seems to remain a year or two away. The current
guess is Dec 2009 for the grand finale, and about July 2009 (my guess)
for something that can be updated to whatever is approved.
http://grouper.ieee.org/groups/802/1..._Timelines.htm

>There is
>of course no guarantee (actually, little likelihood) that any of the
>draft-n gear available now will be upgradable to the standard, when it
>comes.

Well, Dec 2009 is a bit less than 2 years away. Anything you buy
today will probably be obsolete by then. It's not just the MIMO part
that will change in the next two years. Router and wireless features
will also improve. For example, dual SSID's for "guest" WLAN's. Built
in intrusion detection.

>The conventional wisdom for using draft-n equipment at the moment is to
>get all the pieces from the same manufacturer, to make sure they work
>together as intended. But that's not always possible.

Nope. All the same chipsets from the same manufacturer. There's no
guarantee that even a single manufacturers stuff will talk to other
products made with another vendors chips.

Incidentally, the spec includes provisions for two radically different
forms of MIMO. Spacial multiplexing and beam forming are totally
incompatible.

>Have any of the computer or networking magazines, or any of the
>consumer-protection testing labs, run any studies comparing how well (or
>how poorly) 'mixed' environment networks run?

Sorta. This one is several years old:
<http://www.smallnetbuilder.com/content/view/24214/96/>
and doesn't really do any interoperability testing. Some of the
reviews on this site test performance with compatible MIMO and
ordinary wireless clients. I haven't seen any Interop style bake off
testing yet.

>Things like 'The A pcmcia
>card works with B's router but not C's access point, while D's pci
>adapter works with routers from A and C but not B', etc., etc.
>
>Given that there are only, what?, four or five or six major
>manufacturers of consumer- and SOHO-level equipment, it shouldn't be too
>hard to put the more popular models through their paces.

Yep. However, new MIMO models come and go at an alarming rate.
Product lifetimes seem to be about 6 months. By the time the tests
are organized, run, reported, and published, all the products tested
will be obsolete. However, that's not why nobody is doing such tests.
It's because many of the claims presented by MIMO proponents are
unsupportable and all too easily proven to be baloney. Nobody wants
to sponsor a comparative test that demonstrates that MIMO is about
speed, not range.

--
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

The Wi-Fi Alliance, as usual, keep their interoperability test
procedures for Draft N under wrap.
As an aside I did find this web site which does have a pdf. of how they
did it for 11a,11b and 11g.

<http://www.iol.unh.edu/services/testing/wireless/testsuites/temp.html#802.11n%20Draft%202.0%20MAC%20STA%20Conf ormance%20Test%20Suite>

LR <(E-Mail Removed)> hath wroth:
>The Wi-Fi Alliance, as usual, keep their interoperability test
>procedures for Draft N under wrap.

Well yes. The test procedures are their primary asset. If anyone can
obtain the procedures and perform their tests, then someone might
discount their rather high cost of certification (About $25K). Of
course, that won't stop someone from trying:
<http://www.azimuthsystems.com/solutions-wi-fi-certified.htm>
<http://www.azimuthsystems.com/97710dd2-94ae-44f8-9b57-4f97befcfaef/news-press-detail.htm>
There are other vendors, but as usual, I can't find them in my
bookmark mess.

Meanwhile, I'm still dealing with WEP ASCII/BINARY conversion
incompatibilities and regularly find wireless clients that can't seem
to get a DHCP assigned IP from certified wireless routers. There are
a few other big holes in Wi-Fi certification.

So, when all the products are certified and largely indistinguishable,
vendors are competing on the basis of buzzwords and unsubstantiated
claims. That attract class action suits:
<http://www.belkin.com/class_notice/classnotice.asp>
<http://www.techworld.com/mobility/news/index.cfm?NewsID=4873&inkc=0>
<http://www.buffalotech.com/files/Official%20Notice.pdf>
Everybody lies, but that's ok because nobody listens.

>As an aside I did find this web site which does have a pdf. of how they
>did it for 11a,11b and 11g.
><http://www.iol.unh.edu/services/testing/wireless/testsuites/temp.html#802.11n%20Draft%202.0%20MAC%20STA%20Conf ormance%20Test%20Suite>

Never mind interoperability. Just getting MIMO to work in the first
place is somewhat of a challenge:
<http://rfdesign.com/next_generation_wireless/broadband_technologies/radio_challenges_testing_mimo/>

Most of the vendors are assuming that MIMO will go the way of DOCSIS
certification, where there is no interoperability testing between the
DSLAM and the DOCSIS modem. Instead, there are certification tests,
performed with a standardized (CableLabs) test suite, on each device.
In theory, if they communicate reliably with the test equipment, they
should talk reliability to each other. In general, that has worked
well.
--
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

On Sat, 09 Feb 2008 17:10:37 +0000, LR <(E-Mail Removed)> wrote in
<(E-Mail Removed)>:

>John Navas wrote:
>
>> The point is that n can't be certified until it's approved.
>>
>While that is stating the obvious, apart from the OP's first sentence
>the post was about draft n. "Inter-operability of Draft-n equipment"

Sorry, but I can't parse that.

--
Best regards, FAQ for Wireless Internet: <http://Wireless.wikia.com>
John Navas FAQ for Wi-Fi: <http://wireless.wikia.com/wiki/Wi-Fi>
Wi-Fi How To: <http://wireless.wikia.com/wiki/Wi-Fi_HowTo>
Fixes to Wi-Fi Problems: <http://wireless.wikia.com/wiki/Wi-Fi_Fixes>

John Navas <(E-Mail Removed)> wrote:

> (E-Mail Removed) (Henry) wrote

> >get all the pieces from the same manufacturer, to make sure they work
> >together as intended. But that's not always possible.
>
> Why not?

Why not? Erm ... what about the situation where a single manufacturer
just doesn't make all of the parts that you need?

Here's one scenario, but of course there will be others. Apple's latest
Airport Extreme Base Station is draft-n, as are all of their current
computers. But they don't have an internal upgrade card for older Macs
that came with built-in 'b' or 'g', nor do they sell any usb, pci,
pcmcia or expresscard add-ons. Older machines in the building have to
connect on 'g', which can slow down the whole network. Unless, of
course, _someone else's_ usb, pci, pcmcia or expresscard draft-n add-ons
will inter-operate functionally with Apple's gear. Get my point?

cheers,

Henry