Saturday, November 5, 2011

AirMagnet Spectrum Analyzer

| Wireless Driver & Software

In Use

As I mentioned earlier, the Spectrum function features twelve different plots. Four of these – Real Time FFT, FFT Duty Cycle, Swept Spectrogram, and Power vs. Time – are fundamentally different views, one – Power vs. Frequency – is a “tweakable” form of the Real Time FFT, and seven are “Air Quality” plots for advanced RF analysis.The two plots that I found myself using most of the time were the Real Time FFT and Swept Spectrogram demonstrated in Figure 7. The nice, regular spectrum shown was generated by running IxChariot‘s throughput script continuously between a single wireless client and AP simultaneously in uplink and downlink directions.

Nothin' but Channel 6

Figure 7: Nothin’ but Channel 6
(click image to enlarge)

The Real Time FFT and Power vs. Frequency plots are ASA’s two forms of implementing basic spectrum analysis. The main differences between the two are:

  • the Real Time FFT has a fixed resolution bandwidth (RBW) of 156 KHz while the Power vs. Frequency’s RBW can be set from a minimum of 9.77 KHz to a maximum of 1.25 MHz
  • the Real Time FFT plot is generated every second from 5000 FFTs that are crunched by the SAgE engine in the Spectrum PC card while the Power vs. Frequency plot represents one FFT per second calculated in the Spectrum PC card’s MCU

In simple terms, this means that you choose the Real Time FFT if you’re looking for short-lived, but relatively broadband RF phenomena and the Power vs. Frequency plot if hunting for repetitive narrowband glitches. It took me awhile to figure out to use the Average mode for the Power vs. Frequency plot to get anything resembling what I was seeing with the Real Time FFT for my nice, steady WLAN activity.

At first, I thought the Swept Spectrogram plot was mostly eye-candy, but found that it can be a good way to detect changes in relatively quiet RF environments. Figure 8 shows what happens when I moved the laptop containing ASA near my microwave oven that I then turned on.

You can tell that something is going on by comparing the Real Time FFT portions of Figures 7 and 8. But by looking at the Swept Spectrogram – which is showing about 5 minutes worth of history – you can see that the microwave signal is both “bursty” and pretty much covers the entire 802.11b/g spectrum. You can also see that peak microwave signal strength falls between Channels 6 and 11 (the red areas in the Swept Spectrogram).

Channel 6 plus microwave oven
Figure 8: Channel 6 plus microwave oven
(click image to enlarge)

No comments:

Post a Comment