802.11N Design

Hi Rick,

Disclaimer:  I am in marketing, and work for a WLAN vendor.  Your shields should be up, and BS detector at full strength! 

That said, what I believe is that the 802.11n specification includes hundreds of pages of changes from 802.11a/b/g, including the use of constuctive interference in multipathing to extend range and throughput.  However, this leads to highly unpredictable coverage when swapping out access points in a rip & replace approach, like you described.  There's a good diagram of this in our white paper here: http://www.merunetworks.com/collateral/white-papers/2012-wp-the-state-of-802-11n-standards-markets-network-design.pdf.pdf which reminds me of a either a phoenix in flight, or a blast pattern on the sidewalk - but it's not 1/6/11 circles any more!

We always recommend an initial site survey to figure out access point quantity and placement, antenna choice, etc.  But with a single channel architecture, such as ours, deployments are radically easier, often with no site surveys required for extending existing coverage or adding capacity in a given location.

We have a bunch of videos online that explain further.  If you like, shoot me a note and I would be happy to provide them to you.

 

Best regards,

Rich Nedwich
Marketing Director, Meru Networks
email: Rnedwich@merunetworks.com

 

 

Message from jason.cook@adelaide.edu.au

Message from reb@ncsu.edu

We purchased the Ekahau Site Survey software and have being doing new designs using this, mostly in planning mode at this stage. However for this years project we were able to fund a low level person do audits of almost all of our current buildings (to be continued next year for the remainder). Everything is getting collated to allow us to identify the worst buildings for prioritising, then use the existing surveys to design new 5ghz networks. Whilst we haven’t yet gotten to the point of playing with any stage after taking the audits yet, you are able to export from Ekahau into Cisco NCS/Prime which we have. How useful this really is, is yet to be seen.

 

It’s a plan that’s slowly come around over the last 2 years, we started by bringing in an external company to audit 10 buildings for us and design new networks. Part of that was for us to learn, so I followed them around and gathered plenty of useful information about RF design from experienced people. Every 6-12 months we’ll be involving them in a new design which is just a few hours of work. Enables us to get verification on how we do it, and updated information. This is the main reason we chose Ekahau over other vendors like Airmagnet, so we can easily work with this company.

 

So far I’ve been very happy with the software, quite easy to use and their supports been good when we have issues. There’s a few features lacking right now, that I’ve been told are on the road map. Most specifically a feature to take an existing survey, and auto-create a new design using that data. E.g no need to draw all the walls and their db values if you have actual RF values and current AP details. It also understands multi-story buildings and takes that into account, however can-not take into account a room that spans multiple floors like lecture theatres. The lacking features aren’t show stoppers, just requires a little more work, consideration and understanding of RF.

 

Unfortunately I can’t give much “this is how well it works” as we aren’t to that stage quite yet. We are comfortable with the plan, next year I can probably report on how well it’s worked for upgrades. As for the new sites, everything has been going very well.

 

They do have a free very cut down version

http://www.ekahau.com/products/heatmapper/overview.html

And also some “interesting” videos about their products, simple to follow at least. I think it’s aimed for management to understand.

http://www.wifithatworks.com

and some stuff on youtube

http://www.youtube.com/results?search_query=ekahau+site+survey

 

 

--

Jason Cook

Technology Services

The University of Adelaide, AUSTRALIA 5005

Ph    : +61 8 8313 4800

 

I'm using AirMagnet SurveyPro with a Proxim USB dongle. I've found that the USB dongle has lower receive signal strengths than our laptops, so its closer to a device with less power/lower antenna gain. Additionally there are options to simulate other chipsets, though I've never used it.

I do a manual survey using a spare AP (or demo unit, in the case of our recent RFP) using the methods described in the CWNA manual. In some cases I already have an idea of were APs should go, so I'll just start out at given location and spot check it before doing the full survey. There are software features that allow you set requirements and see how your environment meets them. I also use Airmagnet's Spectrum Analyzer to check the environment for interference.

--
Heath Barnhart, CCNA
Network Administrator
Information Technology Services
Washburn University
Topeka, KS

On 10/23/2012 12:30 PM, Rick Brown wrote:

Up to this point we had been doing manual site survey's in the 2.4 GHz space.  We recently upgraded all the existing 802.11a/b/g access points to 802.11n access points doing simply a one for one replacement.  The plan is to go back through and design for 5GHz. 

I'm curious as to what folks are doing.  Are you using manual site surveys, or wireless design software?  For those that have used both, your thoughts especially!  I'm also interested in the type tools that you are using for signal level measurement and analysis.

We have done some testing between various brands of tablets and software apps in addition to Fluke's aircheck meter. The levels read were all over the place.  The Fluke seems to always read better signal level that all else which means we think we may need to use a fudge factor just to be safe. 

Please feel free to comment directly to me if you'd rather not comment on list.

Thanks!

Rick

--

********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.

********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.

Message from ssmith@siu.edu

Over the years I've been doing surveys my basic methodology has changed at least once a year.  Many cases it's due to the technological changes in the 802.xx hardware or configurations, but more recently the BYOD type devices, as you are describing.

I would recommend a good survey should start with 20mW fixed power on the AP, and then utilize multiple client type devices and locking down your survey pro NIC to 20 or 25mW as well.  This does give you some head room in both directions and depending on what your AP manufacturer is this gives you a better understanding on what "could" happen once it's in production.

Certainly don't get me wrong here.  There's still plenty of variables that come into play, but in general this is a good starting point.

Lastly, I would recommend getting your hands on as many over the counter type client devices and understanding how they behave within the wireless spectrums.

Ok, this one is really the last one, iDevices don't play properly within the RF standards.  There are known issues about these type of devices and how you should include them into your survey or at least survey plans/design.

I am starting to find hear about unexpected results when factoring in the performance of devices that have lower transmit power, despite a well performed physical site survey. (iDevices at 20mW, for example)   Conducting a site survey with a laptop using 100mW will produce different results than a ‘droid using 10 or 20 mW.  The receive signal will be relatively close, but depending on your device population, you may want to pay more attention to the receive signal at the AP in addition.  Most of the more advanced survey tools will be able to do this, but you won’t find that feature on the iPhone/Android shareware tools.  Just one more thing to consider.

 

Full disclosure: I don’t personally survey any more, but I am very interested in understanding best practices on how to compensate for tablets/phones when conducting surveys.

 

Caston Thomas InterWorks Mobile 586.530.4981 Office   248.608.0000

 

 

 

 

 

 

Ekahau has a built in selection of client adapters that simulate end user results - you can select various laptop, IPphones, etc.  Still waiting for an ipad/droid selection.  You can manually reduce client power to customize. ********** Participation and subscription information for this EDUCAUSE Constituent Group discussion list can be found at http://www.educause.edu/groups/.