March 2005, Issue 176

Zeroing in on ZigBee (Part 2)
Chipsets and Source Code

WHERE TO NEXT?

After ZigBee is established, it will take over some applications where Bluetooth and other PANs or cable replacement technologies are currently in use. For the consumer, however, the biggest change will have to do with new applications that suddenly become economical after ZigBee reaches critical mass.

A wireless smoke detector serves as an example of a cost-reducing application. A smoke detector that hooks up to a home security system costs approximately $40. Add the wiring and installation charges, and your looking at a $90 price tag. Smoke alarms without built-in communications cost as little as $6. After manufacturers implement industry-wide ZigBee profiles for smoke alarms and security systems, you’ll be able to add in a wireless smoke detector for roughly $10. Upgrading an existing security system base station with a ZigBee adaptor would be a cost-effective solution, even if you’re adding only a single ZigBee smoke sensor or intruder detector.

The key hanger is an example of a new application. Right now, you probably use an RF key fob to remotely lock and unlock your car. Well, Freescale recently demonstrated a ZigBee solution that adds functionality to these devices. The ZigBee network will alert you when you misplace your keys. Say your cell phone and key fob are ZigBee nodes. The cell phone is the coordinator. The cell phone knows when it’s home because it recognizes the other ZigBee nodes (e.g., a garage door lock, a universal remote, etc.). When these nodes are out of range, the cell phone assumes you’re out and about. If your keys leave the cell phone’s communication range, the phone will alert you before you get too far away. The scheme could be applied to other devices too. No more lost sunglasses!

ZigBee is poised to become ubiquitous in home offices and factories, but I think the market may be overestimated. I’m a little skeptical about wireless technologies like ZigBee being relied on to control at least half of the functions inside commercial premises and certainly in most applications inside a wafer fabrication plant or auto factory. Sure, security is built in, but what if some malevolent person turns up with a high-power, wideband jamming device? DSSS is good at ignoring wideband noise up to a certain threshold, whereupon performance rapidly decays to nothing. Ditto for most other wireless systems.

Such an approach doesn’t target the security layer. It simply makes transmission impossible because the clear channel assessment would always indicate a busy channel. Short-range versions used for prohibiting cell phone usage have been around for years. You can bet that these will be sold illegally on the ’Net to target the appropriate bands, if they aren’t already.

Obviously, ZigBee nodes in a building will have sensible built-in behavior just in case communication fails. At worst, a sophisticated arsonist might want to defeat all smoke detectors in a certain area. At best, customers at a petrol station will be unable to pump petrol or pay for their purchases while some larrikin smirks knowingly from across the road. What if the business next door to the scene of the prank is operating a crane with a ZigBee-enabled emergency stop switch? Put simply, you need a secure physical medium such as copper or fiber (not air) connecting serious business assets. Wireless is for convenience rather than critical functions.