How will the Internet of Things communicate?

Jonathan Strickland


We've talked about the Internet of Things a lot on Fw:Thinking (see one of our earliest episodes on it at the end of this post). It's the future in which all sorts of devices and sensors communicate with one another -- and with distant computers and other systems -- to operate in a seamless fashion and transform our world.

All of that communicating has to happen over some sort of network. Some devices might be hardwired into an existing network but others will need to communicate wirelessly. An ideal implementation would be to create mesh networks. These are networks in which nodes relay information across the network. Your various devices and sensors could operate as those nodes. This makes communication across the network robust -- even if one node fails, other nodes can route the information to where it needs to go.

There's no shortage of wireless protocols already -- Wi-Fi, Bluetooth and ZigBee are all solutions. While each one communicates via radio waves, they do so in different ways. And now a proposed newcomer could join the ranks - Thread, a wireless communication network that's an evolution of ZigBee.

Here's a quick breakdown on each type of wireless solution and its pros and cons:

WiFi: This is a local area wireless technology that uses 2.4 GHz ultra-high-frequency or 5 GHz super-high-frequency radio waves. This tech is great for sending large amounts of data wirelessly between devices. But it also requires a lot of energy to operate and many devices in the Internet of the Things model only require a small level of data throughput. In other words, this is overkill for many applications and would require you to change batteries in all your devices on a regular basis.

Bluetooth: Introduced by Ericsson back in the 1990s, Bluetooth technology allows for personal data networks. It transmits data over the frequency band between 2.4 and 2.485 GHz. It operates over shorter distances than WiFi and requires less power to operate. You can pair devices like phones, smartwatches, headsets, speakers, computers and more together. Even with the limited range, early Bluetooth implementations were a big drain on battery life. With the development of Bluetooth v4.0 came the ability to implement low-energy features that conserve power more effectively. Typically, devices connect back to a centralized machine like a computer or smartphone -- they don't "talk" to each other so they aren't a candidate for mesh networks. One thing to note - this standard is still evolving, so perhaps in the future things could change.

ZigBee: This standard is well-suited for mesh networks. The ZigBee standard allows for low-powered devices to send data along a network, with each device capable of relaying the data toward its intended destination. This lets you set up a really effective network -- you don't have to worry as much about getting out of range as you would with a Bluetooth device. As long as your ZigBee gadget can chat with another gadget in the network, you're set. So what's the downside? The main problem is that there are lots of different flavors of ZigBee implementation -- in other words, the standard isn't as standard as it might need to be.

Thread: The brainchild of an alliance between Nest, Samsung, ARM and a few other companies, Thread aims to anticipate the needs of the Internet of Things. Based on the current specifications, Thread would be able to support a network of up to 250 devices. Every house could be its own network, meaning your home could have up to 250 integrated devices interacting with you on a daily basis. Like ZigBee, Thread would allow for mesh networks -- all those devices would be capable of relaying data. Thread hopes to avoid the ZigBee problem of fractured standards by requiring a certification program for anyone wishing to incorporate the technology into a product.

I love the idea of Thread but the plan could still -- forgive me -- unravel. The challenge of creating a standard is that you have to get people (and companies) to convert to it. If Thread works exactly as promised but no one designs products that incorporate it, it'll be useless. The Nest products (thermostat and smoke detector) already run on a version of Thread.

If Thread can prove its usefulness, I think it'd be a solid platform for the Internet of Things. But to do that, the brains behind Thread have to convince everyone from consumers to manufacturers that they're actually solving a problem rather than adding to the list of already available alternatives.