WIRELESS INFRASTRUCTURE

  • 08/25/2015
    7:00 AM
  • Rating: 
    0 votes
    +
    Vote up!
    -
    Vote down!

Wireless Site Surveys Done Right

Here's what you need to know to conduct successful passive and active wireless site surveys.

I see many wireless site surveys with the typical Received Strength Signal Indictor (RSSI), signal-to- noise ratio (SNR), and Signal Quality (SQ) measurements. Some have professional overlays of a floor plan or a 3D building diagram. Others have a heat map indicating different signal levels represented by color. And some are able to illustrate the coverage difference between 2.4 GHz and 5 GHz, antenna types or channel interference.

In most cases, I see wireless architects opting for what is called a "passive survey." With this methodology, you walk around and record signal strength and noise from whatever access point is within range. What I don’t see done enough is information to better document or predict the clients’ wireless experience. One example would be documenting the access point’s radio measurements of how well it hears the wireless client. Another would be noting the client radio measurements and throughput, jitter and packet loss statistics. This kind of information is typically gathered in an active wireless site survey.  

In this blog post, I'll describe what's involved in passive and active wireless site surveys. Depending on your situation, it may be necessary to conduct both.

Passive wireless site surveys

Walking around, collecting competing SSID and general radio stats is very helpful in getting a general idea of what is near your network. I call it a quick glance of your "wireless landscape." Be careful as to what equipment you use when recording your radio stats and note if this is for a new install or troubleshooting an existing network since both have different approaches.

For existing networks, I would strongly suggest you determine what wireless devices will be connecting to the access point and work from the lowest common denominator. Working on an outdoor wireless network for a residential area is totally different than working on a warehouse with forklifts that require wireless access. Some criteria to consider: Is the network inside/outside? Long range/short range? Are the clients fixed or moving?

If you are not sure, you should always err on the side of caution. For example, I wouldn’t use my laptop and a 5 dBi 2.4 GHz directional antenna to perform a passive survey on a BYOD WiFi network that will service smartphone/laptop/tablet users. That's because most of your clients will probably have an omnidirectional antenna, some may have MIMO capabilities, and other antenna configurations. I doubt the average user will have a high-gain directional antenna.

Consider your users' radio/antenna specs when possible. Walking around with a sensitive directional antenna might mislead you into believing interference or other networks are closer than they really are. You need a good understanding of your tools and wireless environment so you don’t get pulled into the wireless rabbit hole. Trust me, trying to determine the gain of your tablet's or smartphone's antenna is not easy, but I have seen reports of antenna gains ranging from -2 to .5 dBi.

Some survey tools allow you to change the antenna calibration or sensitivity, which is a good feature to look for when evaluating or purchasing a tool. An alternative would be to simply use a smartphone, tablet and laptop that you know has a weak antenna when you perform your passive site survey. Sure, it will take a bit longer, but at least there will less doubt with your results.

If you decide to use your laptop and other devices, make sure you use a WiFi utility that reports the signal strength. Avoid recording the number of signal strength bars or operating terms such as good, excellent, poor, etc.

I like Homedale’s free portable WiFi Monitor for my Windows computer's default WiFi card and even whipped up a review of sorts on my website.

Active wireless site survey

If you have never performed an active wireless site survey, it is not as overwhelming or intimidating as you would think. I've listed the steps below; I bet you already probably perform two of the four steps daily. I don’t want to mislead you into thinking that you need to perform an active survey for all networks, but consider it for critical or problem networks.

1.       Record your RSSI or signal strength

  • If you do not have WiFi survey software, just print off a floor plan and put a letter or reference point where you are located and keep your notes on another sheet of paper.
  • Try to avoid the trap of providing WiFi everywhere. Start with areas that you or the client has identified first or trouble areas. Troubleshooting will have a different flow compared to a pre-installation survey, which is why you should always note the purpose for performing the current survey.

2.       Authenticate and successfully join the network

  • This is a quick way to prove that you have a two-way usable signal between yourself and the access point

3.       Perform TCP and UDP tests

  • This simply involves generating some data to and from a server using TCP like an FTP, or copying data back and forth to your network drive. You should have an idea of how long this should take to determine a pass or fail. A better approach would be to record the average throughput.
  • Here is where good old iperf comes in. I use it in TCP mode to record throughput (simulates data applications) and UDP mode (simulates VOIP, or latency-sensitive applications) to record packet loss and jitter.

4.       Record the access point’s radio statistics for your device

  • This can be tricky depending on  the access point model.
  • In some cases where I do not have access to the clients’ wireless devices, I use a hand-held WiFi analyzer like Fluke Networks AirCheck. I simply stand near the access point and record my test equipment’s signal strength and noise levels.
  • If possible, login into the controller or access point and record the radio stats for your test equipment.

If you are working on a pre-installation survey, try to get the access point location, model and antenna. Ideally, run the access point in standalone or autonomous mode and try to reproduce as much of the real installation as possible. Factors such as height, building materials and temperature should be replicated when possible.

I hope my advice on this topic will help you with your wireless site surveys or WLAN troubleshooting. Please share your wireless site survey experiences and thoughts in the comments section below.


Comments

1 AP Per classroom, or per X Square Feet, etc

Nice article, Tony. I'm curious how you feel about the general notions of "1 AP per Classroom" and "simply place an AP every x,000 square feet" as a shortcut to the site survey process?

 

-Lee Badman

Re: 1 AP Per classroom, or per X Square Feet, etc

Hi Lee,

Not to sound too vague, but there's nothing wrong with that rule of thumb as long as you are familar with that specific AP radio and antenna specs as well as the general classroom RF environment.

I was involved with a project to deploy WiFi at a school and we used that approach which was about 75% successful.  Then we went back to figure out the remaining 25%.

Its worth mentioning that if the rooms are fairly small, you want to make sure the AP's TX power is low to ensure a small footprint.

Regards,

Tony Fortunato

welcome!

Hi Tony -- Welcome to Network Computing! Thanks for this helpful guidance on wireless site surveys. I look forward to your future posts!

Re: welcome!

Thanks for the warm welcome

Can't wait to share more info with your readers.

site survey device

Hi Tony,

Enjoyed reading your excellent post - thank you!

Loved the thoughts on the site survey device. One thing most systems integrators seem to require, is site survey device standardization: To keep readings consistent among all the surveyors in your team, as well as the end customer who might be doing surveys / spot checks to maintain the network, perhaps one should - in some cases at least - consider using the same site survey device? Today, there's a few good quality USB adapter alternatives for this.

I would personally recommend this for passive surveys, to keep the signal strength and noise readings consistent, comparable, and accurate.

Of course, ideally, you would simultaneously measure with a standardized adapter, as well as the lowest common denominator (say, an Android phone), for example. Possible on some survey tools.

Look forward to your future posts!

Regards,
Jussi Kiviniemi

Re: site survey device

Great points and i appreciate the feedback.

Re: Wireless Sit Surveys Done Right

"What I don't see done enough is information to better document or predict the clients' wireless experience."

Indeed. WLAN is not my area of expertise, so I'm often mystified how my wireless device can report a strong connection and my experience can be so terrible.  A little extra qualitative info could double the value of an existing hardware investment. There are many complex factors involved in setting up a wireless network, but, as you say, Tony, it's the job of those doing the survey to be on top of all of this. It's the same as any job, really - there's a difference between doing something and doing it well. I'd consider this required reading for someone doing this as part of their job.

Re: Wireless Site Surveys Done Right

Very nicely put.

Thanks for the feedback..  always appreciated