A Pleasant WiSpy Surprise

The WiSpy spectrum analyzer has long made wireless folks feel ambivalent. We love the cheap price and the USB form factor, but we hate the fact that it lacks the device identification capability that you get with the Cisco Spectrum Expert and Fluke AirMagnet Spectrum XT. I've always been one of those folks who tends to think it's worth the money to have a more professional-grade product, but while working at a hotel last week my WiSpy really helped me out.


Before I get to my story, I'd like to give a little background on WiSpy. WiSpy is a USB spectrum analyzer from Metageek. I was first told about it about four years ago by Devin Akin, who at the time was the top technical guru for the CWNP Program. When I clicked on the link he sent to my email, I was amazed. Metageek had created a 2.4 GHz spectrum analyzer for $99.

Like any good compulsive gadgeteer I ordered my WiSpy shortly thereafter and started playing with it. Unfortunately, it didn't take long for me to realize that my initial enthusiasm was misplaced. You see, the original WiSpy was a lot like the current WiSpy dBi. It only analyzed the 2.4 GHz band (which admittedly at the time was more annoying than deal-breaking) and it had no external antenna interface. I still kept the WiSpy in my bag and used it for demonstrations, but for professional work I stuck with the Cognio Spectrum Expert (since acquired by Cisco).

Today things have changed quite a bit, as Metageek now offers a wide range of spectrum analyzers. And when you go with the high end model (WiSpy dBx), you get full 2.4 GHz/5 GHz analysis and a RP-SMA detachable antenna interface.

In large part due to these enhancements WiSpy has been lauded by all sorts of folks at the best way to do WiFi spectrum analysis at a low cost. The WiSpy dBx is just $599 while AirMagnet Spectrum XT, for example, retails for over four times that amount. Still, there has always been an elephant in the room with WiSpy that some sharp-as-a-baby's-ass wireless writers tend to overlook: the compatible software doesn't do device identification.

Device identification is a major help in solving a lot of typical interference problems. Looking at a basic FFT or spectrum density graph is nice, but if I'm a field engineer who needs to solve problems fast, I'd rather pay more and get software that tells me exactly what type of device I'm supposed to be looking for.

What does this have to do with me using WiSpy at last week, you ask? Well, I was working at a hotel and we wanted to setup some APs using the 5 GHz band. I asked my contact if there are any existing wireless systems I should be aware of, and I was assured that the area was clear. Being my usual skeptical self, I decided to fire up WiSpy anyway. Lo and behold, a trickle of wireless was appearing throughout the UNII-3 channels (149-161). It turns out that their phone system for paging the facilities and food & beverage folks uses the 5.8 GHz ISM band, which overlaps the whole UNII-3 band.

Just to make sure I fired up WildPackets OmniPeek (w/ WUSB600N adapter, as usual) and scanned the 5 GHz band. Sure enough, the hotel had no APs on the channel. Well done, WiSpy. Disaster (being overdramatic here) averted.

Footnote: Some people may note that the premise of this post was flawed in some ways. I shouldn't need a spectrum analyzer to tell me to avoid 5 GHz channels 149-161 when I'm setting up APs indoors. The UNII-3 band is best suited for outdoor WiFi so I should have been focusing on the UNII-1 band (channels 36-48). That is true, but I find it's best to still scan 5 GHz channels outside of the UNII-1 band so that you can tell the people who are going to manage the network which specific channels should be avoided.

Comments

  1. hehe. Nice article. Does that make me the guy behind the guy behind the guy with the sniffer? :) I couldn't resist.

    Devin

    ReplyDelete
  2. Ben,

    Though I teach folks how to use the Cognio Spectrum Expert (AirMagnet's version) as well as AirMagnet's own Spectrum XT. I try very hard to NOT use the built-in device identification.

    Instead I teach how as a 'human' you can also beat the computer in identification, the ability to 'see' things/patterns in the FFT plots and Swept Spectograms.

    It's kind of like learning to read sonograms or x-ray results. It takes lots of practice. But in the end, a 'human' will always be able to beat a computer in pattern identification. (Think of hearing your child's cry at a noisy playground)

    Keith@inpnet.org
    http://wirelesslanprofessionals.com

    ReplyDelete

Post a Comment

Popular posts from this blog

Spectrum Deception

Free Sniffing in Windows! (Kind Of)

What's New (and Missing) in the WiFi for iPhone 6