Cisco BR-350 Bridge setup -- Any Help with Error Messages - Associations?? ,, thanks

Hi all,

I'm using 2 cisco 350 radios on a wireless link about 2mi in length.

I've got the radio, 40' LMR400, 24dbi grid dish on each end of the
link. The power on each radio is set at 100 mw. The radios are set
as root and non-root w/o clients.

I can see each antenna from the other antenna with field glasses so
I've got no obstructions between the links. One antenna is mounted at
60' and the other antenna is mounted at 40'. The antennas are
horizontally polarized.

I've checked the channels and am using channel 9, which shows NO
radios on that channel. I've done this with netstumbler. I've
checked this by driving between the two points and watching the laptop
for radios. Also to be sure I connected a PCMCIA card in the laptop
to each antenna and checked for radios showing up this way also. So
i'm pretty sure that I've got no other radios on channel 9.


The link tests show about a 98% signal test at all times i've checked
it.

However, the log shows that the non-root radio without clients
associates and dis-associates itself from the root radio about every
3-5 minutes.

I can supply more information about this link if needed.

Thanks for the read and possible help,

Bob Smith

Robert Smith Consulting
Fort Bragg, Califonria


Bob Smith

Bob Smith <(E-Mail Removed)> hath wroth:

>I'm using 2 cisco 350 radios on a wireless link about 2mi in length.
>
>I've got the radio, 40' LMR400, 24dbi grid dish on each end of the
>link. The power on each radio is set at 100 mw. The radios are set
>as root and non-root w/o clients.

Sounds good.
TX power = +20 dBm
TX coax loss = 4dB
TX ant gain = 24dBi
Distance = 2 miles
RX ant gain = 24dBi
RX coax loss = 4dB
RX sens = -84 dBm (at 12 Mbits/sec)
Fade margin = unknown
| http://www.terabeam.com/support/calculations/som.php
Calculated fade margin is 33.8dB. That's more than enough.

Depending on what speed you are selling, you might consider dropping
the xmit power to a bit less than maximum.

>I can see each antenna from the other antenna with field glasses so
>I've got no obstructions between the links. One antenna is mounted at
>60' and the other antenna is mounted at 40'. The antennas are
>horizontally polarized.

Fresnel zone incursion? You need 27ft clearance at midpoint. Are
your antennas both more than 27ft off the ground? Any debris within
27ft of the line of sight at midpoint?

>I've checked the channels and am using channel 9, which shows NO
>radios on that channel. I've done this with netstumbler.

Yech. Ch 9 will get interference from both 6 and 11 as well as the
other channels in between. Channels are 5 MHz wide. 802.11 is about
25MHz wide. Try to stay on 1, 6, and 11, even if occupied.

Netstumbler only picks up access points that respond to probe
requests. They don't see client radios, AP's with SSID off,
proprietary protocols (frequency hoppers), etc. At least use Kismet,
which will see client radios and bridges. Borrow a spectrum analyzer
and do some sniffing for non-802.11 junk. A friend recently found a
local company doing injection molding in a small industrial park. They
had a very leaky microwave oven setup as a a plastic pre-heater to
remove the moisture before molding/casting. I've now found 3 over
powered 802.11 amplifiers that were generating more pollution than
communication.

Laundry list of potential sources of crud:
http://wireless.wikia.com/wiki/Wi-Fi#Interference

>I've
>checked this by driving between the two points and watching the laptop
>for radios. Also to be sure I connected a PCMCIA card in the laptop
>to each antenna and checked for radios showing up this way also. So
>i'm pretty sure that I've got no other radios on channel 9.
>
>The link tests show about a 98% signal test at all times i've checked
>it.

OK, that's more than sufficiently strong. What about S/N ratio? If
it's interference, it should be going up and down.

>However, the log shows that the non-root radio without clients
>associates and dis-associates itself from the root radio about every
>3-5 minutes.

WPA re-keying interval? Try it temporarily without encryption and see
if the problem goes away.

>I can supply more information about this link if needed.

Contour profile using Radio-Mobile? (Never mind).

--
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 Wed, 13 Sep 2006 23:59:17 -0700, Jeff Liebermann
<(E-Mail Removed)> wrote:

>Bob Smith <(E-Mail Removed)> hath wroth:
>
>>I'm using 2 cisco 350 radios on a wireless link about 2mi in length.
>>
>>I've got the radio, 40' LMR400, 24dbi grid dish on each end of the
>>link. The power on each radio is set at 100 mw. The radios are set
>>as root and non-root w/o clients.
>
>Sounds good.
> TX power = +20 dBm
> TX coax loss = 4dB
> TX ant gain = 24dBi
> Distance = 2 miles
> RX ant gain = 24dBi
> RX coax loss = 4dB
> RX sens = -84 dBm (at 12 Mbits/sec)
> Fade margin = unknown
>| http://www.terabeam.com/support/calculations/som.php
>Calculated fade margin is 33.8dB. That's more than enough.

This is what I figured also....

I haven't had that much experience with Cisco radios but the
associated/dis-associate problem could be a radio. I was watching
the root radio and I think I was seeing a boot/reboot happen
intermittantly. All the lites would turn RED then all out, then the
radio would start to work. This is a city system and the bad thing
about it is that they don't have any spare radios that I could
substitute in to see if it's the radio.



I've tried setting the speeds from 1mb to 11mb & auto. The speed
doesn't seem to help the situation

>Depending on what speed you are selling, you might consider dropping
>the xmit power to a bit less than maximum.
>
>>I can see each antenna from the other antenna with field glasses so
>>I've got no obstructions between the links. One antenna is mounted at
>>60' and the other antenna is mounted at 40'. The antennas are
>>horizontally polarized.
>
>Fresnel zone incursion? You need 27ft clearance at midpoint. Are
>your antennas both more than 27ft off the ground? Any debris within
>27ft of the line of sight at midpoint?


The Fresnel zone looks clean with field glasses. No trees or
buildings are in the way.

>>I've checked the channels and am using channel 9, which shows NO
>>radios on that channel. I've done this with netstumbler.
>
>Yech. Ch 9 will get interference from both 6 and 11 as well as the
>other channels in between. Channels are 5 MHz wide. 802.11 is about
>25MHz wide. Try to stay on 1, 6, and 11, even if occupied.

The local computer store has started putting up hot spots. There is
one about 2000' from the Root Radio. The radio they are using is a
tranzeo panel pointed South. My antenna is pointed NW
using netstumbler I see the tranzeo radio signal strength varying from
10 to 20. it's about 60 degrees angle from the center of the 24dbi
grid feed.

On the same tower I have another cisco radio on channel one pointed
180 degrees direction( see below)


<-------------- tranzeo
channel 11
nothing /
/
/
2000' /

Cisco w/24di grid
channel 9
no wpa/no wep SSID on

cisco w/24dbi grid
channel 1
wpa on SSID off
/
/
/
/



>Netstumbler only picks up access points that respond to probe
>requests. They don't see client radios, AP's with SSID off,
>proprietary protocols (frequency hoppers), etc. At least use Kismet,
>which will see client radios and bridges. Borrow a spectrum analyzer
>and do some sniffing for non-802.11 junk. A friend recently found a
>local company doing injection molding in a small industrial park. They
>had a very leaky microwave oven setup as a a plastic pre-heater to
>remove the moisture before molding/casting. I've now found 3 over
>powered 802.11 amplifiers that were generating more pollution than
>communication.
>
>Laundry list of potential sources of crud:
> http://wireless.wikia.com/wiki/Wi-Fi#Interference

I've also used Kismet and seen nothing in the area.. The non-root
radio has no buildings arround it for at least a mile. The root radio
does have the hotspot radio shown above.


>>I've
>>checked this by driving between the two points and watching the laptop
>>for radios. Also to be sure I connected a PCMCIA card in the laptop
>>to each antenna and checked for radios showing up this way also. So
>>i'm pretty sure that I've got no other radios on channel 9.
>>
>>The link tests show about a 98% signal test at all times i've checked
>>it.
>
>OK, that's more than sufficiently strong. What about S/N ratio? If
>it's interference, it should be going up and down.

I wish there was a way to see SNR.. I don't see how to do this with
the 350 radio. They just have antenna alignment and signal strength


However, the log shows that the non-root radio without clients
>>associates and dis-associates itself from the root radio about every
>>3-5 minutes.


>WPA re-keying interval? Try it temporarily without encryption and see
>if the problem goes away.
I have turned the encryption off, and am broadcasting the SSID

>>I can supply more information about this link if needed.
>
>Contour profile using Radio-Mobile? (Never mind).

thanks jeff, maybe the above will help explain it a little be more..


Bob


Robert Smith Consulting
Fort Bragg, Ca
NA6T

On Thu, 14 Sep 2006 15:14:44 -0700, Bob Smith <(E-Mail Removed)> wrote:

>I haven't had that much experience with Cisco radios but the
>associated/dis-associate problem could be a radio. I was watching
>the root radio and I think I was seeing a boot/reboot happen
>intermittantly. All the lites would turn RED then all out, then the
>radio would start to work.

That sure sounds like an unscheduled reboot. Anything in the logs?
| http://www.cisco.com/en/US/products/...080451858.html
If you think it's crashing, try the usual reset to defaults and reload
all the settings. If you're really adventureous, reflash with the
latest firmware.

>This is a city system and the bad thing
>about it is that they don't have any spare radios that I could
>substitute in to see if it's the radio.

There are several AIR-BR350 radios on eBay for $250 to $400 ea.

>I've tried setting the speeds from 1mb to 11mb & auto. The speed
>doesn't seem to help the situation

Did you try different channels? It's probably not the signal strength
with means that speed (and corresponding rx sensitivity) probably are
not part of the problem.

>>Fresnel zone incursion? You need 27ft clearance at midpoint. Are
>>your antennas both more than 27ft off the ground? Any debris within
>>27ft of the line of sight at midpoint?

>The Fresnel zone looks clean with field glasses. No trees or
>buildings are in the way.

You didn't answer my question. Are your antennas more than about 27ft
off the ground? If not, the ground is in the Fresnel zone.

>The local computer store has started putting up hot spots. There is
>one about 2000' from the Root Radio. The radio they are using is a
>tranzeo panel pointed South. My antenna is pointed NW
>using netstumbler I see the tranzeo radio signal strength varying from
>10 to 20. it's about 60 degrees angle from the center of the 24dbi
>grid feed.

Don't worry about it. The 24dBi dish antennas have a -3dB beamwidth
of about 7 degrees. You'll never hear anything coming from the side
or back. What you will hear is anything that's inline with the line
of sight (in both directions). Are there any radio towers along the
line of sight?

>On the same tower I have another cisco radio on channel one pointed
>180 degrees direction( see below)

That might cause a problem when they're that close. Just stay away
from channel 1 and you'll probably be fine.

>I've also used Kismet and seen nothing in the area.. The non-root
>radio has no buildings arround it for at least a mile. The root radio
>does have the hotspot radio shown above.

I'm fairly sure it's not an RF issue, but a spectrum analyzer might
show some non-802.11 interference.

>I wish there was a way to see SNR.. I don't see how to do this with
>the 350 radio. They just have antenna alignment and signal strength

I think it's in there somewhere. Maybe on the IOS command line. Most
probably can be extracted using an SNMP MIB browser. MIB list:
| ftp://ftp-sj.cisco.com/pub/mibs/supp...pportlist.html
I don't have a BR350 handy to figure out the appropriate OID or
incantation.

BR350 pages:
| http://www.cisco.com/en/US/products/hw/wireless/ps458/



--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558 (E-Mail Removed)
# http://802.11junk.com (E-Mail Removed)
# http://www.LearnByDestroying.com AE6KS

On Fri, 15 Sep 2006 00:52:48 GMT, Jeff Liebermann
<(E-Mail Removed)> wrote:

>>I wish there was a way to see SNR.. I don't see how to do this with
>>the 350 radio. They just have antenna alignment and signal strength
>
>I think it's in there somewhere. Maybe on the IOS command line. Most
>probably can be extracted using an SNMP MIB browser. MIB list:
>| ftp://ftp-sj.cisco.com/pub/mibs/supp...pportlist.html

Wrong URL. For the BR350 MIB's, see:
| ftp://ftp-sj.cisco.com/pub/mibs/supp...pportlist.html

The SNR is in the 802.11 Interface MIB at:
| ftp://ftp.cisco.com/pub/mibs/v2/CISCO-DOT11-IF-MIB.my
Search the document for "SNR".

Use a MIB browser to extract the numbers. I suggest GetIF 2.3.1 from:
| http://www.wtcs.org/snmp4tpc/getif.htm
Install the program.
Drop whatever MIB files you need into:
c:\program files\getif 2.3.1\mibs\
directory.
Erase the file named ".index". Getif will recreate it every time it
starts. When you add (or remove) any of the MIB files, erase it
again. Yeah, I know it's dumb but I didn't write this thing.
Operation of the program is "obvious" (which means read the help and
don't bug me). Good luck.

--
# Jeff Liebermann 150 Felker St #D Santa Cruz CA 95060
# 831-336-2558 (E-Mail Removed)
# http://802.11junk.com (E-Mail Removed)
# http://www.LearnByDestroying.com AE6KS