dslreports logo
site
 
    All Forums Hot Topics Gallery
spc

spacer




how-to block ads


Search Topic:
uniqs
2073
share rss forum feed

watts3000

join:2002-01-21

how to properly do a wireless survey

Guys I've always wondered for the professional wireless gurus how do you properly perform a wireless survey and what tools do you use be it free or ones that cost do you use to carry out the survey? Also what questions do you attemtp to answer during this survey for example I assume how many access points will be needed, what type of traffic will be going across these access points, how will the access points connect core of the network etc anyway thanks guys.



phantasm11b
Premium
join:2007-11-02

A true wireless survey is a complicated beast. I usually contract mine out to companies who specialize in it. However, some things to look for and keep in mind are the building materials, presence and location of any ductwork and or voids, ceiling height, amount of windows, and mirrors if any.

As for the amount of AP's to place, Cisco's recommendation is 1 AP per 5,000sqft of floor space. That number is for data only. If you plan to run voice over wifi, subtract a 1/3 from 5,000sqft. If you plan to run location tracking, that's another 1/3 off of the 5,000sqft recommendation.

I've used both autonomous (standalone AP's with their own IOS image) and controller based. The controller based AP's are nice as the controller provides a single point of management for all the AP's. Autonomous can have their use but I find them to be cumbersome in large deployments.

This is just off the top of my head, wireless design practices vary depending on what you're trying to achieve and what you plan to use to achieve it.



tubbynet
reminds me of the danse russe
Premium,MVM
join:2008-01-16
Chandler, AZ
kudos:1
reply to watts3000

we use 'airmagnet' for our wireless surveys.
expensive, for sure, but it is nifty.
in greenfield, you set up a few aps (using battery or long ethernet with a switch) in locations that you think will work. then use the software and walk until you hit an rx-threshold. then move the aps and repeat.

brownfield is easier. walk all over ans spit a report out.

q.
--
"...if I in my north room dance naked, grotesquely before my mirror waving my shirt round my head and singing softly to myself..."



Camaro
Question everything
Premium
join:2008-04-05
Westfield, MA
kudos:1
Reviews:
·Comcast

You can go even cheaper if you nix the survey software and just use the AP with the battery backup and windows built in WIFI manager. You are going to be overlapping anyways so being to precise isn't a issue, unless the software is on your company's dollar.


HELLFIRE
Premium
join:2009-11-25
kudos:18
reply to watts3000

The elcheapo / DIY route would be a laptop, an AP and something like inssider.

The professional way would be as phantasm11b See Profile mentions, consider about floor / building layout, building
material considerations, etc. It's not an easy beast, and if you want it done right the first time, stick to the
professionals.

As to the rest of your questions, go in there with at least some idea of what you expect from the wireless from
your / business perspective and needs -- # of users, day to day use patters, reliability, traffic load / patterns,
expected growth / function in the next 3 years or so, etc. One of the WORST ways to work I find is get someone
who wants it because it's "gee-whiz, gotta have it!," but has no clue of the underlying technology or what they
want it to do for them, then when it's built is blowing blood vessels left and right screaming "you didn't do it right!"

My 00000010bits.

Regards



tubbynet
reminds me of the danse russe
Premium,MVM
join:2008-01-16
Chandler, AZ
kudos:1
reply to Camaro

if the customer is expecting heatmaps and test results -- professional software is the only way to go.

q.


cramer
Premium
join:2007-04-10
Raleigh, NC
kudos:9
reply to watts3000

We used to do it by wandering around with an iPAQ (compaq's version of the palm pilot.) I don't remember what app was used.

Today, we have numerous bits of software to simulate AP coverage before even ordering the APs. And that same software can query the installed APs to get a coverage "heatmap", to make sure there's no gaps or RF overlaps. An on-site survey would only be needed where there are outside RF signals already present. (and that data is fed into the modeling software.)

[Note: I don't deal with any of this anymore.]


aryoba
Premium,MVM
join:2002-08-22
kudos:4
reply to phantasm11b

said by phantasm11b:

A true wireless survey is a complicated beast. I usually contract mine out to companies who specialize in it. However, some things to look for and keep in mind are the building materials, presence and location of any ductwork and or voids, ceiling height, amount of windows, and mirrors if any.

Make sure that wireless survey is done when all of those walls, ceilings, doors, windows, beams, and any interior items are in place. I recalled that we did wireless survey once on customer site prior final interior arrangement. Once the customer decided to move things around to come up with the final interior arrangement, the wireless connection started to get spotty


phantasm11b
Premium
join:2007-11-02

said by aryoba:

said by phantasm11b:

A true wireless survey is a complicated beast. I usually contract mine out to companies who specialize in it. However, some things to look for and keep in mind are the building materials, presence and location of any ductwork and or voids, ceiling height, amount of windows, and mirrors if any.

Make sure that wireless survey is done when all of those walls, ceilings, doors, windows, beams, and any interior items are in place. I recalled that we did wireless survey once on customer site prior final interior arrangement. Once the customer decided to move things around to come up with the final interior arrangement, the wireless connection started to get spotty

Yup. Been through that. The powers that be decided to install some metal accent pieces on the walls because they thought they would look good. It turns out the mesh they installed drops the signal strength by about 5db. Made my life hell for a bit.