A picture, they say, is worth a thousand words -- and if youre a network administrator or an RF engineer charged with managing a wireless network in a hostile environment, the right images from a spectrum analyzer could speak volumes.
Ill admit it: I dont have the typical home environment. I have two 5.8 GHz cordless phones, each with three handsets, as well as a 2.4 GHz cordless phone with two handsets. When I survey the 2.4 GHz spectrum prior to testing wireless products, I see as many as 10 APs. Who knows what else is out there. This seemed like the perfect test environment to check out AirMagnets Spectrum Analyzer 2.0.
Priced at $3,995, this is clearly a tool aimed at enterprises that must maintain a highly reliable wireless network. It may seem hard to justify that price, but if your wireless network slows to a crawl because of interference you cant identify, the loss of productivity could easily offset the cost.
Installation of the AirMagnet Spectrum Analyzer is quite simple. You install the software, then insert the bundled dual band PC card into your notebook. The card can locate interference on all 14 channels (only 1-11 are legal in the US) in the 2.4 GHz radio band used by 802.11b/g, and the channels in the 5.150-5.350 and 5.725-5.850 GHz bands used by 802.11a.
In order to detect Wi-Fi signals, you also need a Wi-Fi card preferably one on which you can disable the transmitter. For my tests, I used an IBM Thinkpad T40, and enabled the embedded wireless. I set the preferences to associate to a non-existent AP so that its transmissions would be limited.
Since my editor doesnt allow me an infinite number of words, I thought I would include a number of screen shots along with explanations of the test scenarios for each. And if a picture is truly worth a thousand words, well, lets hope that my editor pays me by the word... [Editor's Note: I didn't.]
If youre a user of AirMagnets Spectrum Analyzer 1.0, you might find some of the features in 2.0, such as session record and playback, compelling enough for an upgrade. You can record, play, open, save and pause session files. Version 2.0 also includes new device classifiers, including 5 GHz microwave ovens, European Bluetooth devices, continuous FSK, OFDM, QAM, FM and CW devices, as well as burst FSK, FSK-FHSS, OFDM, QAM and QAM-FHSS. It can also differentiate between analog video, CW and FM. To help you analyze data, new charts include devices vs. channel, devices vs. time, duty cycle vs. channel and duty cycle vs time.
When you launch AirMagnets Spectrum Analyzer, the screen says powered by Cognio. Spectrum Analyzer 2.0 is based on Cognios Spectrum Expert 3.0. The difference in the AirMagnet version is that it integrates with AirMagnet Survey 3.0. This provides real-time FFT (Fast Fourier Transform) visibility, a spectrum snapshot for each data point, and spectrum reports when surveying with Survey 3.0.
The opening screen presents you with a default grouping of windows. While this view is quite useful, you can configure the program to display exactly the windows and charts that you want, and you can save multiple configurations for future use.
On the left side of the screen is a box that shows active devices, and, if you have a Wi-Fi card in receive mode, Wi-Fi APs. Here, you can see three cordless phones and seven APs. Below that box is the control panel that controls frequencies displayed and amplitude scaling.
The top left corner graph is the Real Time FFT plot. This shows power, expressed in dBm as a function of frequency measured during the period of one frequency sweep. Typically, sweep time is one second. Note that the channel edges for the non-overlapping 2.4 GHz Wi-Fi channels -- 1, 6 and 11 -- are shown in dotted blue lines for reference.
The graph to the right of the Real Time FFT plot is the FFT duty cycle. This shows the percentage of time that the power, from both 802.11 and interfering devices, is 20dB or more above the noise floor. Again, the horizontal axis is the bandwidth being monitored. Like the real time FFT plot, channel edge markers for channels 1, 6 and 11 are shown. Youll see in other screen shots how the FFT duty cycle changes due to either 802.11 traffic or interference.
Below the Real Time FFT plot is the Swept Spectrogram plot. This plot actually combines the data shown in the first two graphs, and allows you to see power in the RF spectrum as a function of frequency, measured over time. Typically, each sweep line is one second, so this view shows you about three minutes of history. Signal strength is indicated by color. Blue is -112 dBm (noise) and red is -42 dBm (very strong signal).
The last chart is a stacked bar chart showing devices per channel.
AirMagnet Spectrum Analyzer 2.0 has five tabs that run horizontally across the top of the interface. The first tab shows your configured screen for the 2.4 GHz spectrum. The next tab shows your configured screens for the 5 GHz spectrum.
Here, I chose only to view the FFT Duty Cycle, the Swept Spectogram and the Real Time FFT plots. Note the fairly clean environment, with most power levels down in the noise.
Its easy to add or remove charts or plots to the main screen. Just right-click on the screen, and the dialogue box below pops up with individual options.
Add Plot Dialogue
Add Chart Dialogue
The third tab gives you a list of devices. In this view, devices are identified and grouped together. In the screen shot below, note how the cordless phones are grouped together. To generate this screen, I made a phone call from a 2.4 GHz phone to a 5 GHz phone. As you can see from the fairly strong signals, both of the base stations and handsets were in the same room as the Spectrum Analyzer.
Ive always heard that microwave ovens cause interference in the 2.4 GHz band, but never really had a way to test it. Now, with AirMagnets Spectrum Analyzer 2.0, I could. For the microwave test, I took the Spectrum Analyzer notebook and placed it in the center of the kitchen, about eight feet away from the microwave oven. I put a plastic bowl with about four cups of water into the oven (to provide a load for the microwave), and turned it on high. As you can see from the screenshot below, the microwave oven was clearly identified by the Spectrum Analyzer.
A right-click on the device brought up detailed information about the source of the interference.
Note the relatively strong signal (-51.2 dBm) and the affected channels (8-11).
I decided to switch back to the main view to see the real time plots, and sure enough, all four panels showed the interference caused by the microwave.
The Real Time FFT shows significant energy centered around channel 11, but spanning several channels. The FFT duty cycle plot (energy 20 dB above the noise floor) shows a 30% duty cycle. The Swept Spectogram also shows a strong signal (solid yellow and red at the bottom of the display), and the Device vs. Channel chart shows microwave interference on channels 8-11.
About 60 seconds after I shut off the microwave, I took another screen shot. (below) Actually, this shows that the microwave had been on for about a minute, shut off (while I changed the water) and started up again. Note the gaps in the Swept Spectogram plot. Also note that the FFT Duty Cycle plot returned to a low level, and the microwave disappeared from the Devices vs. Channel chart.
I thought it might be interesting to see what a draft N product looked like when doing a file transfer. For this test, I set up a notebook with a draft N (Broadcom-based) card, and associated it to a draft N (Netgear WNR854T, Marvell-based) router. Both the router and the client were in the same room, operating on channel 3, and the client card indicated a data rate of 270 Mbps.
From the Swept Spectogram, its pretty easy to see when the file transfer started, and that it was using 40 MHz of bandwidth at least part of the time.
The last tab across the top is the device finder. In this window, you can actually locate the source of interference.
To use the device finder, you click on an active source of interference. In this case, I selected one of the cordless base stations. This screen works somewhat like a Geiger counter. The closer you get to the source of interference, the stronger the signal gets, and it plots the signal over time. In this example, I started with the notebook in the kitchen. I moved the notebook up the stairs into my office and watched the signal get stronger and stronger. Finally, when I was about a foot away from the base station, the signal peaked. Unfortunately, you cant use this feature to find APs. Other products in AirMagnets line will help you track down APs.
AirMagnet Spectrum Analyzer 2.0 is an incredibly powerful diagnostic tool for troubleshooting layer 1 wireless problems, and its surprising easy to learn and use. If you manage a corporate wireless network and have the budget, this is a tool that should be in your toolkit. The first time you isolate interference thats slowing down your wireless network, youll be a hero.
Article courtesy of Wi-Fi Planet