can I rule our interference?

I had a problem a while back, solved it. It wasn't interference. But
i'm wondering generally.. what the symptoms of interference are, and if
I can ever rule it out.

For example, suppose I change to each channel and sitll the same
problem, would that rule out interference or could it just mean there's
interference on all channels.

Interference is when you connect but had low responsiveness right?

What about signal? If I connect but have a good signal like -45db
amplitude reported by inSSIDer, and 5 bars reported by Windows. But
let's say for some reason it's unresponsive. Would that , the 5 bars
or the -45db, mean definitely not interference?

On 6 Sep 2011 23:11:14 GMT, "tommy alimo" <tommyalimo@yahoo.com>
wrote:

>i'm wondering generally.. what the symptoms of interference are,

If you ping the wireless router, instead of a nice constant and
consistent 1-3msec latency, it will hop around radically in value,
with the occasional added bonus of a time out. Something like this is
what interference looks like:

C:\> ping -t 192.168.111.1
Pinging 192.168.111.1 with 32 bytes of data:
Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
Reply from 192.168.111.1: bytes=32 time=20ms TTL=64
Reply from 192.168.111.1: bytes=32 time=105ms TTL=64
Reply from 192.168.111.1: bytes=32 time=15ms TTL=64
Reply from 192.168.111.1: bytes=32 time=75ms TTL=64
Reply from 192.168.111.1: bytes=32 time=5ms TTL=64
Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
The variations are the result of retransmissions and retries inspired
by lost packets.

>and if I can ever rule it out.

Yes, you can rule it out if your wireless router and client are both
located in an RF shielded room, basement, dungeon, or some location
where sources of RF are unlikely. Otherwise, interference is a
prevalent problem.

>For example, suppose I change to each channel and sitll the same
>problem, would that rule out interference or could it just mean there's
>interference on all channels.

There are only 3 channels (1, 6, and 11) so that's not going to be
effective. Many wireless routers set their operating channels to
"auto" by default, which makes channel selection a crap shoot. Many
MIMO (802.11n) routers use a 40Mhz wide channel instead of the usual
20Mhz resulting in the entire band being one channel.

>Interference is when you connect but had low responsiveness right?

Lack of responsiveness is one symptom. Erratic and slow downloads are
another.

>What about signal? If I connect but have a good signal like -45db
>amplitude reported by inSSIDer, and 5 bars reported by Windows. But
>let's say for some reason it's unresponsive. Would that , the 5 bars
>or the -45db, mean definitely not interference?

That doesn't work. You can have an absurdly strong signal and it will
still be susceptible to interference. There's no FM "capture effect'
in 802.11a/b/g. A fairly weak interference signal can drastically
reduce the thruput from a fairly strong signal.

So, what problem are you trying to solve?

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558
# http://802.11junk.com jeffl@cruzio.com
# http://www.LearnByDestroying.com AE6KS

"tommy alimo" <tommyalimo@yahoo.com> wrote:
> I had a problem a while back, solved it. It wasn't interference. But
> i'm wondering generally.. what the symptoms of interference are, and if
> I can ever rule it out.
>
> For example, suppose I change to each channel and sitll the same
> problem, would that rule out interference or could it just mean there's
> interference on all channels.
>
> Interference is when you connect but had low responsiveness right?
>
> What about signal? If I connect but have a good signal like -45db
> amplitude reported by inSSIDer, and 5 bars reported by Windows. But
> let's say for some reason it's unresponsive. Would that , the 5 bars
> or the -45db, mean definitely not interference?

On my program it says strength and data quality. There are three basic
channels.

Greg

Jeff Liebermann wrote:

> On 6 Sep 2011 23:11:14 GMT, "tommy alimo" <tommyalimo@yahoo.com>
> wrote:
>
> > i'm wondering generally.. what the symptoms of interference are,
>
> If you ping the wireless router, instead of a nice constant and
> consistent 1-3msec latency, it will hop around radically in value,
> with the occasional added bonus of a time out. Something like this is
> what interference looks like:
>
> C:\> ping -t 192.168.111.1
> Pinging 192.168.111.1 with 32 bytes of data:
> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=20ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=105ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=15ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=75ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=5ms TTL=64
> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
> The variations are the result of retransmissions and retries inspired
> by lost packets.
>
> > and if I can ever rule it out.
>
> Yes, you can rule it out if your wireless router and client are both
> located in an RF shielded room, basement, dungeon, or some location
> where sources of RF are unlikely. Otherwise, interference is a
> prevalent problem.
>

but if I don't get that symptom with the pings, then there can be no
(significant) interference problem?

I don't actually have any connection problem it's just a general
question so I know for the future and understand the situation better.


> > For example, suppose I change to each channel and sitll the same
> > problem, would that rule out interference or could it just mean
> > there's interference on all channels.
>
> There are only 3 channels (1, 6, and 11) so that's not going to be
> effective. Many wireless routers set their operating channels to
> "auto" by default, which makes channel selection a crap shoot. Many
> MIMO (802.11n) routers use a 40Mhz wide channel instead of the usual
> 20Mhz resulting in the entire band being one channel.

inSSIDer and xirrus wifi inspector show my router on channel 3, and I
can select any from 1 to 11, so what do you mean?

it's all very nice if of neighbourign routers, one is on 1, and one is
on 11, then i go on 6. But a bunch are on 1, and a bunch are on 6, and
a bunch are on 11.

So what the heck, why not just choose 3?

And also, let's suppose there is just one router and it's on 1,
wouldn't a good use of space for me to put mine on channel 5?
channels 1 and 5 don't overlap do they?


1 2 3 4 5

+- 20Mhz either side..
channel 1 = 2412Mhz
Channel 2 = 2417Mhz
channel 3 = 2422Mhz
channel 4 = 2427Mhz
Channel 5 = 2432Mhz

Each does a 20Mhz range, 10Mhz each direction

Channel 1 = 2402Mhz-2422Mhz

Channel 5 = 2422Mhz-2442Mhz

No overlap.

Why does it have to be one router on channel 1, another router 5Mhz
away from the edge of its spectrum? thus on channel 6. Why not
channels 1 and 5.. no gap, but no overlap either.

and would channel 1 be like up to 2421.9999 and Channel 5 be from
2422.000001 ? Even if not, 2422 is only one discrete number that'd
overlap on a spectrum with 20 of them, so is it so bad to have just one
frequency 2422Mhz overlapping?





>
> > Interference is when you connect but had low responsiveness right?
>
> Lack of responsiveness is one symptom. Erratic and slow downloads are
> another.
>
> > What about signal? If I connect but have a good signal like -45db
> > amplitude reported by inSSIDer, and 5 bars reported by Windows. But
> > let's say for some reason it's unresponsive. Would that , the 5
> > bars or the -45db, mean definitely not interference?
>
> That doesn't work. You can have an absurdly strong signal and it will
> still be susceptible to interference. There's no FM "capture effect'
> in 802.11a/b/g. A fairly weak interference signal can drastically
> reduce the thruput from a fairly strong signal.
>
> So, what problem are you trying to solve?

I'm just trying to understand wireless issues a bit better so I know
what's going on a bit more.

btw xirrus wifi inspector works nicely showing WPA and WPA2.

On 7 Sep 2011 21:52:58 GMT, "tommy alimo" <tommyalimo@yahoo.com> wrote:

>Jeff Liebermann wrote:
>
>> On 6 Sep 2011 23:11:14 GMT, "tommy alimo" <tommyalimo@yahoo.com>
>> wrote:
>>
>> > i'm wondering generally.. what the symptoms of interference are,
>>
>> If you ping the wireless router, instead of a nice constant and
>> consistent 1-3msec latency, it will hop around radically in value,
>> with the occasional added bonus of a time out. Something like this is
>> what interference looks like:
>>
>> C:\> ping -t 192.168.111.1
>> Pinging 192.168.111.1 with 32 bytes of data:
>> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=20ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=105ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=15ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=75ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=5ms TTL=64
>> Reply from 192.168.111.1: bytes=32 time=2ms TTL=64
>> The variations are the result of retransmissions and retries inspired
>> by lost packets.
>>
>> > and if I can ever rule it out.
>>
>> Yes, you can rule it out if your wireless router and client are both
>> located in an RF shielded room, basement, dungeon, or some location
>> where sources of RF are unlikely. Otherwise, interference is a
>> prevalent problem.
>>
>
>but if I don't get that symptom with the pings, then there can be no
>(significant) interference problem?

Theoretically, interference could still cause a throughput impact, even if you
aren't getting packet retransmissions hence the variable latency scenario that
Jeff shows above. Remember that 802.11 does CSMA/CA where CA = "Collision
AVOIDANCE" - the transmitter will check the channel before sending, and if there
is noise on it above a given threshold (the Clear Channel Assessment [CCA] -
typically -85dBm or so), it will defer.

>I don't actually have any connection problem it's just a general
>question so I know for the future and understand the situation better.
>
>
>> > For example, suppose I change to each channel and sitll the same
>> > problem, would that rule out interference or could it just mean
>> > there's interference on all channels.

One or the other.

>> There are only 3 channels (1, 6, and 11) so that's not going to be
>> effective. Many wireless routers set their operating channels to
>> "auto" by default, which makes channel selection a crap shoot. Many
>> MIMO (802.11n) routers use a 40Mhz wide channel instead of the usual
>> 20Mhz resulting in the entire band being one channel.
>
>inSSIDer and xirrus wifi inspector show my router on channel 3, and I
>can select any from 1 to 11, so what do you mean?

He meant that there are only 3 *nonoverlapping* channels: 1, 6, and 11.
If you have set your router on channel 3, then your signal will overlap with
other cells that are on *both* channels 1 and 6. So your performance should be
worse than if you had picked either 1 or 6.

>it's all very nice if of neighbourign routers, one is on 1, and one is
>on 11, then i go on 6. But a bunch are on 1, and a bunch are on 6, and
>a bunch are on 11.
>
>So what the heck, why not just choose 3?

Because your signal is now interfering with (/being interfered by) *two* bunches
of APs, the bunch on 1 and the bunch on 6.

>And also, let's suppose there is just one router and it's on 1,
>wouldn't a good use of space for me to put mine on channel 5?
>channels 1 and 5 don't overlap do they?
>
>
>1 2 3 4 5

Yes they do.

Here's a little white paper that you may enjoy reading:

Channel Deployment Issues for 2.4-GHz 802.11 WLANs
http://www.cisco.com/en/US/docs/wire...e/Channel.html

Cheers,

Aaron

----

>
>+- 20Mhz either side..
>channel 1 = 2412Mhz
>Channel 2 = 2417Mhz
>channel 3 = 2422Mhz
>channel 4 = 2427Mhz
>Channel 5 = 2432Mhz
>
>Each does a 20Mhz range, 10Mhz each direction
>
>Channel 1 = 2402Mhz-2422Mhz
>
>Channel 5 = 2422Mhz-2442Mhz
>
>No overlap.
>
>Why does it have to be one router on channel 1, another router 5Mhz
>away from the edge of its spectrum? thus on channel 6. Why not
>channels 1 and 5.. no gap, but no overlap either.
>
>and would channel 1 be like up to 2421.9999 and Channel 5 be from
>2422.000001 ? Even if not, 2422 is only one discrete number that'd
>overlap on a spectrum with 20 of them, so is it so bad to have just one
>frequency 2422Mhz overlapping?
>
>
>
>
>
>>
>> > Interference is when you connect but had low responsiveness right?
>>
>> Lack of responsiveness is one symptom. Erratic and slow downloads are
>> another.
>>
>> > What about signal? If I connect but have a good signal like -45db
>> > amplitude reported by inSSIDer, and 5 bars reported by Windows. But
>> > let's say for some reason it's unresponsive. Would that , the 5
>> > bars or the -45db, mean definitely not interference?
>>
>> That doesn't work. You can have an absurdly strong signal and it will
>> still be susceptible to interference. There's no FM "capture effect'
>> in 802.11a/b/g. A fairly weak interference signal can drastically
>> reduce the thruput from a fairly strong signal.
>>
>> So, what problem are you trying to solve?
>
>I'm just trying to understand wireless issues a bit better so I know
>what's going on a bit more.
>
>btw xirrus wifi inspector works nicely showing WPA and WPA2.

On 7 Sep 2011 21:52:58 GMT, "tommy alimo" <tommyalimo@yahoo.com>
wrote:

>it's all very nice if of neighbourign routers, one is on 1, and one is
>on 11, then i go on 6. But a bunch are on 1, and a bunch are on 6, and
>a bunch are on 11.

Sometimes it's not about how many AP's are on a channel but rather how
strong they appear to be, and programs like inSSIDer can help show you
that. I'd rather share a channel with a dozen very weak AP's versus a
channel with a couple of very strong AP's.

On 7 Sep 2011 21:52:58 GMT, "tommy alimo" <tommyalimo@yahoo.com>
wrote:

>but if I don't get that symptom with the pings, then there can be no
>(significant) interference problem?

Variable latency is just one of the symptoms of interference. It's
like a runny nose and a cold. If you have a cold, it's highly likely
that you have a runny nose. However, if you have a runny nose, it can
be caused by a cold, flu, hay fever, allergy, etc. It really depends
on the source of the interference. For example, a periodic
interference source will usually produce very consistent latency
figures. Instead of the usual 1-2 msec latency, you'll see a much
larger number, with little variation.

Because the ping packets are small, they don't test for all forms of
interference. A much better indication is a thruput test. I suggest
you download and use Jperf:
<http://code.google.com/p/xjperf/>
<http://openmaniak.com/iperf.php>
If you get nice stable upload and download speeds through the wireless
link, at a speed you would consider normal, then you do not have any
interference. If it varies radically, you have interference.

>I don't actually have any connection problem it's just a general
>question so I know for the future and understand the situation better.

Good question. One of the common questions I see is "My wireless was
working just fine until a few days ago. Now, it's erratic,
disconnects often, slow, and useless. I didn't change anything".
Usually, that means the neighbors have dragged home a new wireless
router or wireless toy, and are creating interference. Being able to
recognize the difference between interference and simple
misconfiguration is useful.

>inSSIDer and xirrus wifi inspector show my router on channel 3, and I
>can select any from 1 to 11, so what do you mean?

Aaron Leonard answered that question effectively. The occupied
bandwidth (how wide the signal appears) is roughly 22MHz for
802.11a/b/g. At 5MHz per channel, that's a bit over 4 channels wide.
With 11 channels available, that's about 3 usable channels that do NOT
overlap each over.
<http://en.wikipedia.org/wiki/802.11#Channels_and_international_compatibility>
Note that the degree of interference from adjacent channels depends
heavily on the distance between you and the source of interference. If
BOTH radios were in the same room, you might need more isolation than
if they were located in separate buildings.

>it's all very nice if of neighbourign routers, one is on 1, and one is
>on 11, then i go on 6. But a bunch are on 1, and a bunch are on 6, and
>a bunch are on 11.
>
>So what the heck, why not just choose 3?

Because you receive interference from others on both channels 1 and
channel 6. On a statistical basis, you'll be worse off than picking
either 1 or 6. In addition, you will be generating interference to
users of both 1 and 6, thus doubling the number of irate neighbors
banging on your door at odd hours.

>And also, let's suppose there is just one router and it's on 1,
>wouldn't a good use of space for me to put mine on channel 5?
>channels 1 and 5 don't overlap do they?

The channels are 5MHz wide. However, your signal is 22MHz wide.

>btw xirrus wifi inspector works nicely showing WPA and WPA2.

Cool. You might also try:
<http://www.performancewifi.net/performance-wifi/main/NetSurveyor.htm>
Unfortunately, it doesn't do a good job of identifying encryption
type, but does everything else well enough. To see interference
caused by non-802.11 devices, such as microwave ovens, you'll need a
spectrum analyzer.
<http://www.metageek.net>


--
Jeff Liebermann jeffl@cruzio.com
150 Felker St #D http://www.LearnByDestroying.com
Santa Cruz CA 95060 http://802.11junk.com
Skype: JeffLiebermann AE6KS 831-336-2558