Wireless Access

Hello all,

 

Whilst this isn't Aruba centric, I know a lot of people in the forums use Airmagnet and I just thought I'd share this with the group.

 

Maybe this will help avoid some challenges and time wastage for some of us?

 

In version 8.6 of Survey Pro (still the latest as I write this), they claim 802.11ac compatibility with a number of adapters. Due to our teams/business receiving a lot of 802.11ac survey requests, we pushed on with testing and execution of 8.6 with a couple of 802.11ac adpaters.

 

Our biggest finding was that although Fluke claim 802.11ac survey capability, this is limited to PASSIVE surveying ONLY. You CANNOT currently DO ACTIVE surveys with ANY 802.11ac adapter. There are also some other minor gaps and challenges, but for me this is by far the biggest.

 

Thanks.

Hi,

 

I am doing active surveys quite well with the internal Intel 7260-ac.  There are some limitations as you say, but it seems to work really well for me......apart from the usual Intel device reporting the signal too high.

 

The limitations are that you can't do dual active surveys when using the 7260-ac.  I can't change the roaming paramters within Airmagnet either.

 

Apart from that I am having a pretty good experience with it compared the the massive Airmagnet headaches in the past with previous adapters.

 

Have you tried that Intel adapter?

The big question is "Under what circumstances would you do an active survey"?

Generally always, if possible.  Better to have the info and not need it than the other way round.

 

Currently I do dual surveys with the 7260-ac doing active and the Proxim USB doing the passive.

 

Racking.Monkey, initially I had probs with 8.6 and active survey on the ac adapter.  I installed the patch release 8.6.1 which addressed the issues, though the Release Notes don't mention anything in there.  Airmagnet support advised me to upgrade to that to fix the issues.

 

Compared to previous versions and adapters I am having very few hassles now.  Previously my survey latop very nearly got a flying lesson from a great height on many occasions.....but now is all good.

 

:smileyhappy:

Michael_Clarke,

 

I have never done an active survey so I was wondering what the value of one is.

 

Would you do a survey with two different cards, even though they have different properties?

 

The value is that in addition to the signal strength, you get info on what the actual connected rate is, retries, lost packets and roaming behaviour.

 

No problem with different adapters.  Different cards report signal strength differently, but you can calibrate the adapter which will give a more accurate figure......apparently.

 

And using two adapters at the same time saves a lot of time....

Michael_Clarke,

 

Thank you.

 

Yeah, agree with everything MC said about actives. Broadly speaking, some customers like to do actives as it gives them more "comfort". My view is it's a bit of a glamour show.

MC, you wouldn't believe how annoying that is that you have an adapter capable of actives! For the simple reason I asked Fluke outright, "which ac adapters wiil do actives today". They said in writing "none"! Fluke can be really frustrating!

I'll google your adapter and see what's what. Cheers.

Forgot to say...

Yeah, we classically run dual Proxims. Quite good historically.

And yeah, I'm not sold on "Calibration". Looks like BS to me!

Yeah, the jury is still out on the calibration.  I did test it a bit at some stage, by surveying a channel without calibration  and then looking at the raw signal with a spectrum analyser.  The difference was about 15 dB.  When I looked at the data calibrated, it pretty much matched what the spectrum analyser said.  That was only a quick 15 min test on one channel though.

 

:smileyhappy:

It's kind of like saying, "this adapter doesn't interpret itself properly" isn't it? Weird.

 

Thanks again for the views.

FWIW, the most we use calibration for is to re-normalize the results for drastically different devices. We often have partners going in to do surveys for environments where handhelds will be used, but the handhelds obviously can't be the survey devices. So they will do a manual normalization and take a few samples from different areas within the area and compare survey laptop SNR to handheld SNR (and gotta do it both at the handheld looking at the radio output, as well as the controller with the 'Rx seen by the controller/AP from the handheld). Some handhelds we have seen suffer as much as a 9-12 Rx/Tx differential from an average laptop. They then will write a custom calibration to account for it post-survey and give them the results 'coverage with a laptop' and 'coverage with your handheld' kind of report. Same for voice devices since that is so critical. While vendors do a good job of stating Rx sensitivities allowing you to really map SNR, the Tx side of their devices are usually not given.

 

http://myaccount.flukenetworks.com/fnet/en-us/supportAndDownloads/KB/IT+Networking/AirMagnet+Survery/Bad+signal+levels+-+AirMagnet+Survey+Pro

Hi Jerrod,

 

In your scenario, yes I can see why that might well help.

 

Thanks for the input.

I have just installed the AC7260 in my laptop and it will only work in active mode whihc is no good as my Proxim dongle holds the licence/mac address that airmagnet is tied too, any thoughts ?

Why not just use the Proxim?

I do use the Proxim but i also use the internal device so i can both Passive and Active surveys at the same time as this gives me throughput data when i companre trhe active and passive on the same SSID.

Unfortunately i am unable to do this now as i can't get the AC7260 to recognise it isn't the primary device.