Worrying about cordless phones, Bluetooth headsets, and microwave ovens interfering with your 802.11 b/g wireless LAN? These pale in comparison to wireless security cameras. That’s right, security cameras, especially those little critters that operate on the unlicensed 2.4 GHz band, same as 802.11 b/g.
Microwave ovens are often considered to be the king of 802.11 b/g interference, generating RF that splatters across the 2.4 GHz spectrum (and who knows what else). Luckily, access points and clients that maintain a safe distance don’t get cooked.
Next on the list of offenders are cordless phones. With cordless phones at least the signals are controlled on a relatively narrow band within given frequency ranges (one for the base and another for the handset) compared to the ovens.
Then there’s Bluetooth, pretty much at the bottom of the list due to its low power and constant real-time frequency hopping which lowers the duty cycle on any given channel. This is not a major detractor for 802.11 unless Bluetooth kicks into high power mode,
Enter wireless security cameras. Forget about big brother watching you. These babies are capable of completely crippling your network. You’ve heard the stories: Hospital puts up security system, wireless goes down, etc. Words like decimate and obliterate come to mind.
I ordered a few of these cameras for a mere 25 bucks online. I kept telling myself that it can’t be this easy to bring down a WLAN. It is. The ones I got even have a removable antenna on the back. I went to the lab and grabbed one of our external 802.11 directional booster antennas and screwed it right in. Furthermore it has a dip switch for 4 channels which I quickly learned allows you to pick any part of the 802.11 2.4 GHz channel range you wish to bring down, er, operate in. Nice.
Turn one on within a dozen or so feet of an access point and nobody gets on. To the user, the wireless LAN disappears (those little bars on the laptops turn all white – no signal!). Within 20 to 30 feet you’ve got a dangerously low signal-to-noise ratio for packets to/from the AP. Get that close to a client and it doesn’t bother to send anything. Within about 40 to 50 feet of an access point (or client) the error count and wireless frame re-transmission is horrific to the point of disconnecting. Within a 100 feet or so you’ve got errors but stuff does manage to get through.
Why is this? A little (okay a lot) of detailed RF analysis with OmniSpectrum tells all. Without getting into all the technical nitty-gritty, the important metrics are: 1) the spectrum the camera is transmitting on relative to 802.11 channels; 2) duty cycle; and 3) signal strength.
Get a device that covers 4 or 5 adjacent channels with 100 percent duty cycle and a strong signal and you’ve got trouble. That’s exactly what these cameras are capable of. With 100% duty cycle overlapping, say, channel 1, placing such a device in the vicinity of access points and clients on that channel cause them to not even attempt to send a packet – the RF looks (and is) 100% occupied.
If the duty cycle is 100% but the signal strength is moderate to low, a client will typically send packets (as evidenced by capturing wireless packets with OmniPeek), but they are corrupted by the ensuing RF from the camera. I found it interesting that a device would even attempt to send a packet when sensing moderate RF activity. Since the activity is non-802.11, perhaps the driver figures it’s worth a shot.
How about the old adage of sending smaller packets (fewer bits, less time in the air) in such interfering or noisy environments? With the analyzer we can easily see that larger packets have a higher probably of being garbled. Thus, we could set our wireless clients to split those large packets using wireless fragmentation. But why bother? Get to the source of the problem!
Don’t let this happen to you. Use RF spectrum analysis to ensure that your environment is clean and stays that way. Your WiFi users will love you and you’ll keep big brother at bay to boot.