Home wireless without the power trip

A new generation of low-power radio technologies is creeping into our homes, in the form of wireless light switches and remote-controlled plug sockets. But the next generation of home-automation kit is all going to communicate every which way, assuming a common language can be agreed upon.

Zigbee, Z-Wave, and Bluetooth Low Energy would all like to play in this space, and Intel is even trying to squeeze Wi-Fi into the role. The industry believes the time is right for watches, key fobs and door locks to be wirelessly enabled, and that it's finally time to replace the aging infrared remote controls that have been controlling our TVs for the last few decades.

Not that the IDrA (Infrared Data Association) is taking this lying down. The group is quick to point out that the most expensive part of building IR into a device is the plastic window covering the transmitter, and has demonstrated a 1Gb/sec version of the standard. But line of sight is always going to be a problem for IR, so for most applications radio works better.

Still seeing infrared?

Several radio standards are competing for the title, all offering limited bandwidth and very low power consumption, but differing both in the details of their protocols and in how they expect the architecture of our (home) wireless networks to evolve.

Groups such as the Bluetooth SIG and Ozmo Devices (who have been demonstrating Intel's Cliffside low-power Wi-Fi) see the future filled with client devices connected to more powerful servers. In the case of Ozmo the "server" could be a games console or home computer, while the Bluetooth SIG still reckons the mobile phone will be the most popular server for their "Bluetooth Low Energy" standard. This divergence is based on the technologies onto which the two approaches piggy-back: Wi-Fi and Bluetooth respectively.

The standard formally known as Wibree, now christened "Bluetooth Low Energy Technology", is designed to make use of much of the Bluetooth silicon - not just the antenna as Nokia's original Wibree proposed. The SIG reckons that should make the cost of bundling Low Energy an insignificant increment on that of putting Bluetooth into a device. Phones supporting Bluetooth should, from the end of 2009, automatically support the Low Energy technology, though it remains to be seen if there will be any Low Energy devices for them to talk to by then. The idea is that watches, light switches etc. will only have the Low Energy components, though the cost of those is still unknown.

Cliffside is a technology from Intel that does much the same thing with Wi-Fi circuits. The argument is that as every device is going to have Wi-Fi anyway you might as well make use of that circuit for your low-power connections too. Cliffside allows a device to connect to a traditional Wi-Fi network while simultaneously talking to low-power-Wi-Fi devices, such as those recently demonstrated by Ozmo Devices.

All this traffic at 2.4GHz might see the channel getting a bit crowded, if it wasn't already. But Ozmo reckons their technology will scale well to 5.8GHz - another chunk of unlicensed space to which Wi-Fi connections are increasingly shifting.

So the question becomes whether Bluetooth or Wi-Fi is going to be the wireless technology embedded in everything by default - on which technology the low-power variation can most usefully piggy-back.

Ozmo reckons its devices can out-perform Bluetooth, offering 9Mb/sec bandwidth and with 2.5 times the battery life of a Bluetooth device. But Bluetooth Low Energy should have 10 times the battery life of a normal Bluetooth device, even if it won't be able to offer the same bandwidth.

Point to the window, point to the door

A 10x improvement in battery life is all very well, but to reach a life measured in years rather than days you need to move away from existing technologies and adopt something a little more radical.

Zigbee is a low-power protocol that can also fit into 2.4GHz, and operating at such low bandwidth it can squeeze in around other applications - one of the 16 bands used by Zigbee at 2.4GHz is actually a Wi-Fi guard band, so will never be filled by networking kit. Zigbee can also hang around at 433Mhz, but that's a very chaotic band for interference, or 868Mhz, though there's not much room to breathe there.

Zigbee was designed from the ground up to be a very low-power networking technology, and one that allows nodes to act as relays, creating an auto-forming mesh network to extend coverage. Zigbee and its competitor Z-Wave both offer a battery life measured in years, and the first consumer products using the technologies are now on the shelves.

AlertMe.Com is typical. It uses Zigbee-equipped sensors operating at 2.4GHz and with a battery life of several years to monitor doors and windows, as well as temperatures and user interaction, with mesh-enabled nodes operating as status lights while routing connections around the network. The mesh that AlertMe.Com creates only connects to a single hub, but there's no reason for a Zigbee deployment to be limited in that way.

AlertMe.Com is working on more devices, including a remotely-controlled plug socket, which should provide a standard platform for home automation. This first deployment will be controlled by AlertMe.Com as all traffic must be routed through their servers, but it won't be long before we see more open Zigbee hardware on the shelves.

So if you believe the future will be devices connecting to peripherals then it's most likely going to be Bluetooth Low Energy or low-powered Wi-Fi, depending on which is the dominant embedded technology. But if you see a future in networks of nodes around the home then Z-Wave or Zigbee is the technology of choice.

If, on the other hand, you want to see high-definition video or uncompressed audio flying around the house then you'll need something else entirely, and we'll be looking at some of the options there soon.