SmartWireless

I see that BT is now purveying a Home Hub 3 with "SmartWireless" (TM)
technology. Apparently this changes the AP's wireless channel to the
one with the least interference.

So if this shiny new AP cannot hear a lower powered AP then it's likely
to trample all over its signal. It's seems to me that this is a design
rule of the selfish, and unless everyone plays by these new rules people's
network channels will get arbitrarily stomped over.

Thoughts, please?
Chris

"Chris Davies" <chris-(E-Mail Removed)> wrote in message news:(E-Mail Removed)...
>I see that BT is now purveying a Home Hub 3 with "SmartWireless" (TM)
> technology. Apparently this changes the AP's wireless channel to the
> one with the least interference.
>
> So if this shiny new AP cannot hear a lower powered AP then it's likely
> to trample all over its signal. It's seems to me that this is a design
> rule of the selfish, and unless everyone plays by these new rules people's
> network channels will get arbitrarily stomped over.
>
> Thoughts, please?
> Chris

Agree 100%

I first saw this "feature" on a D-Link DIR-615 supplied by Virgin Media.
Once I realised what it did my natural instinct was to turn it off and
choose a channel manually after a site survey.

But once these things become the norm they will be constantly fighting
amongst themselves for 2.GHz spectrum space and things will be worse.

I got a massive (A0?) poster through the post from BT, extolling the virtues of
Smart Wireless (TM).

Do the public really chose an ISP on the claims made about the supplied access point performance??? (Rhetorical)



--
Graham.

%Profound_observation%

"Chris Davies" <chris-(E-Mail Removed)> wrote in message
news:(E-Mail Removed)...
>I see that BT is now purveying a Home Hub 3 with "SmartWireless" (TM)
> technology. Apparently this changes the AP's wireless channel to the
> one with the least interference.
>
> So if this shiny new AP cannot hear a lower powered AP then it's likely
> to trample all over its signal. It's seems to me that this is a design
> rule of the selfish, and unless everyone plays by these new rules people's
> network channels will get arbitrarily stomped over.


If you were choosing which channel to use, could you do any better?

Of the 11 (perhaps 13) available, each channel overlaps the adjacent
channels to some extent, so simultaneous use of channels 1, 5 and 11 in the
same location represents the maximum spectrum load. Any other channels used
will reduce the effective range and throughput some or all the channels.
Your choice is obviously to transmit over the weakest other channel ...

The whole concept is misguided. We moved from repeaters to switches to
avoid large collision domains, and wireless brings back a large collision
domain with a much less efficient transport protocol. Much better to use
wires!

--
Graham J

"Graham J" <graham@invalid> wrote in message
news:4dcd5e5f$0$2505$(E-Mail Removed)...
>
> "Chris Davies" <chris-(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...
>>I see that BT is now purveying a Home Hub 3 with "SmartWireless" (TM)
>> technology. Apparently this changes the AP's wireless channel to the
>> one with the least interference.
>>
>> So if this shiny new AP cannot hear a lower powered AP then it's likely
>> to trample all over its signal. It's seems to me that this is a design
>> rule of the selfish, and unless everyone plays by these new rules
>> people's
>> network channels will get arbitrarily stomped over.
>
>
> If you were choosing which channel to use, could you do any better?
>
> Of the 11 (perhaps 13) available, each channel overlaps the adjacent
> channels to some extent, so simultaneous use of channels 1, 5 and 11 in
> the same location represents the maximum spectrum load. Any other
> channels used will reduce the effective range and throughput some or all
> the channels. Your choice is obviously to transmit over the weakest other
> channel ...
>
> The whole concept is misguided. We moved from repeaters to switches to
> avoid large collision domains, and wireless brings back a large collision
> domain with a much less efficient transport protocol. Much better to use
> wires!

2.4Ghz networking is a complete waste of time here, there are 34 networks
competing for use of the 3 overlap free channels which has slowed everyone
to a crawl. I gave i a while back and bought myself a dual band router. I
have the only 5Ghz network within range.....

Paul

In article <iqjr0p$dfs$(E-Mail Removed)>, (E-Mail Removed)ap (Paulg0)
wrote:

> I gave i a while back and bought myself a dual band router. I
> have the only 5Ghz network within range.....

Seriously considering that option myself.

I have BT Openwound and Fon on 1, 3, 5, 6, 8, 9, 11 and 12 here!
I'm running on channel 2, but my SSID is my callsign, and my power is a smidgin
higher than the standard 100mW

--
Paul Cummins - Always a NetHead
Wasting Bandwidth since 1981

Meanwhile, at the uk.telecom.broadband Job Justification Hearings, Graham.
chose the tried and tested strategy of:

> "Chris Davies" <chris-(E-Mail Removed)> wrote in message
> news:(E-Mail Removed)...

>> So if this shiny new AP cannot hear a lower powered AP then it's likely
>> to trample all over its signal. It's seems to me that this is a design
>> rule of the selfish, and unless everyone plays by these new rules
>> people's network channels will get arbitrarily stomped over.

In what sense do people's network channels /not/ get arbitrarily stomped
over without such a system? Very few AP owners are aware there may be
channels that their AP uses; fewer still will know that adjacent channels
overlap each other to a large degree. To avoid interfering with someone
else's network, you need to know the position of all their nodes in relation
to all of your nodes. This would be hard enough with two networks near each
other, but quickly becomes mind boggling as you add more networks and nodes.

http://en.wikipedia.org/wiki/Hidden_node_problem

> But once these things become the norm they will be constantly fighting
> amongst themselves for 2.GHz spectrum space and things will be worse.

I don't think it's all that new a feature anyway. I've seen automatic
channel selection on other APs.

Given the need for independently operated wireless networks with little
regulatory involvement, a "tragedy of the commons" type situation is
inevitable.

Sure, you could come up with some system where adjacent APs try to
coordinate to avoid stomping on each other, but the success or otherwise of
such a system would be subject to being implemented correctly and in a
compatible fashion by every AP maker. And that's without considering those
trying to game such a system to get more throughput for themselves.

The commons at 5GHz, however, haven't been trashed yet [possibly because the
range is as far as at 2.4GHz].

> Do the public really chose an ISP on the claims made about the supplied
> access point performance??? (Rhetorical)

As Graham J alluded to, if performance was a factor, wireless wouldn't be
considered.

--
<http://ale.cx/> (AIM:troffasky) ((E-Mail Removed))
10:34:30 up 7 days, 14:33, 6 users, load average: 0.57, 1.10, 0.61
"People believe any quote they read on the internet
if it fits their preconceived notions." - Martin Luther King

alexd wrote:
> Meanwhile, at the uk.telecom.broadband Job Justification Hearings, Graham.
> chose the tried and tested strategy of:
>
>> "Chris Davies" <chris-(E-Mail Removed)> wrote in message
>> news:(E-Mail Removed)...
>
>>> So if this shiny new AP cannot hear a lower powered AP then it's likely
>>> to trample all over its signal. It's seems to me that this is a design
>>> rule of the selfish, and unless everyone plays by these new rules
>>> people's network channels will get arbitrarily stomped over.
>
> In what sense do people's network channels /not/ get arbitrarily stomped
> over without such a system? Very few AP owners are aware there may be
> channels that their AP uses; fewer still will know that adjacent channels
> overlap each other to a large degree. To avoid interfering with someone
> else's network, you need to know the position of all their nodes in relation
> to all of your nodes. This would be hard enough with two networks near each
> other, but quickly becomes mind boggling as you add more networks and nodes.
>
> http://en.wikipedia.org/wiki/Hidden_node_problem
>
>> But once these things become the norm they will be constantly fighting
>> amongst themselves for 2.GHz spectrum space and things will be worse.
>
> I don't think it's all that new a feature anyway. I've seen automatic
> channel selection on other APs.
>
> Given the need for independently operated wireless networks with little
> regulatory involvement, a "tragedy of the commons" type situation is
> inevitable.
>
> Sure, you could come up with some system where adjacent APs try to
> coordinate to avoid stomping on each other, but the success or otherwise of
> such a system would be subject to being implemented correctly and in a
> compatible fashion by every AP maker. And that's without considering those
> trying to game such a system to get more throughput for themselves.
>
> The commons at 5GHz, however, haven't been trashed yet [possibly because the
> range is as far as at 2.4GHz].
>
>> Do the public really chose an ISP on the claims made about the supplied
>> access point performance??? (Rhetorical)
>
> As Graham J alluded to, if performance was a factor, wireless wouldn't be
> considered.
>
Very good post, if I may say so.

Which just supports my growing belief that WiFi access is fundamentally
pants.

alexd <(E-Mail Removed)> wrote:
> In what sense do people's network channels /not/ get arbitrarily stomped
> over without such a system?

The part of the system that frustrates me is the automatic switching of
channels, and not so much the automatic (one-off) selection.

When I got stomped on the other week I was able to find another channel,
and as I'm reasonably sure that my neighbours won't be fiddling with
their APs, I feel I'm "safe" for now. If their APs started switching
around seemingly arbitrarily, I'd be forever playing catch-up.


> Sure, you could come up with some system where adjacent APs try to
> coordinate to avoid stomping on each other

This system is called SmartWireless. It claims to try to avoid channels
with interference. Consider three routers A, B, C, where A and B are
adjacent, B and C are adjacent, but A and C cannot hear each other. If
all three have SmartWireless then the system works. If one does not have
this new feature, then it loses out.


> such a system would be subject to being implemented correctly and in a
> compatible fashion by every AP maker. And that's without considering those
> trying to game such a system to get more throughput for themselves.

Absolutely.


> As Graham J alluded to, if performance was a factor, wireless wouldn't be
> considered.

There are different levels of performance. For the single user PC
(with maybe a smartphone or 'pad), data transfer is bounded by the
broadband speed. For many users with multiple PCs I would suspect that
actually there is not a lot of interaction between those PCs, so again
the performance is bounded by the broadband speed. Those of us dicussing
this topic, with or without multiple PCs, are very much in the minority.

Chris

Chris Davies wrote:
> alexd <(E-Mail Removed)> wrote:
>> In what sense do people's network channels /not/ get arbitrarily stomped
>> over without such a system?
>
> The part of the system that frustrates me is the automatic switching of
> channels, and not so much the automatic (one-off) selection.
>
> When I got stomped on the other week I was able to find another channel,
> and as I'm reasonably sure that my neighbours won't be fiddling with
> their APs, I feel I'm "safe" for now. If their APs started switching
> around seemingly arbitrarily, I'd be forever playing catch-up.
>
Possibly..or note. see below
>
>> Sure, you could come up with some system where adjacent APs try to
>> coordinate to avoid stomping on each other
>
> This system is called SmartWireless. It claims to try to avoid channels
> with interference. Consider three routers A, B, C, where A and B are
> adjacent, B and C are adjacent, but A and C cannot hear each other. If
> all three have SmartWireless then the system works. If one does not have
> this new feature, then it loses out.
>
>
>> such a system would be subject to being implemented correctly and in a
>> compatible fashion by every AP maker. And that's without considering those
>> trying to game such a system to get more throughput for themselves.
>
> Absolutely.
>

frequency agile systems can sort out the interference problems .
Essentially both ends agree to hop to a new channel. This doesn't
eliminate issues, but it means that the full spectrum is better utilised.

I am not sure if any wifi systems use this, but its common in e.g. radio
control on the 2.4Ghz band.



>
>> As Graham J alluded to, if performance was a factor, wireless wouldn't be
>> considered.
>
> There are different levels of performance. For the single user PC
> (with maybe a smartphone or 'pad), data transfer is bounded by the
> broadband speed. For many users with multiple PCs I would suspect that
> actually there is not a lot of interaction between those PCs, so again
> the performance is bounded by the broadband speed. Those of us dicussing
> this topic, with or without multiple PCs, are very much in the minority.
>
I could not get the clean 100Mbps I have over cat5 between my server and
the desktops, using wifi.


> Chris