The Internet of Things Protocol Wars

Transformational technologies are inevitably the battleground for competing implementations and protocols, and the Internet of Things (IoT) is no exception. However, the evolutionary trajectory of two different markets makes this battle an interesting and unpredictable one. IoT is the nexus of Internet computing (web technologies, tablets, smartphones, etc.) and an older breed of consumer electronics such as air conditioners, light switches, and security systems. While these two technology markets are working increasingly well together, there is one big battle that has been taking shape for the last 18 months, and it may come to a head this year.

Droves of "smart" devices are hitting the market. They can be controlled from an app, from a remote server, or from other devices. Almost all of the communication for these devices is wireless, which sounds innocuous enough. We've been controlling things without wires for a very long time, but it turns out that wireless communication strategies are shaping up to be the most contentious aspect of IoT development.

The Sides

On one side of the field is the familiar face of 802.11, which we all affectionately call WiFi. Like a dashing knight, it is reliable, battle-tested, and continuously improving. We understand WiFi. Our coffee shops and libraries sport little WiFi stickers on their front doors. Its bandwidth facilitates rapidly moves large amounts of data. It’s clearly the most visible protocol in our everyday computing environment. Proponents of WiFi view its claim to the IoT throne as a truism. How could it possibly be otherwise?

On the other side of the field stands a small band of contenders. They don't have the stature, the mass appeal, or the hefty experience of WiFi. You might condescendingly call them "upstarts." But this rag-tag band has something WiFi doesn't: specialization.

Among the dozen or so contenders against WiFi, there are some notable names. Z-Wave and Zigbee protocols are most recognizable, and the new Bluetooth Low Energy (BTL or BTLE) protocol is also in this camp. Not to be overlooked, Lutron's RadioRA and the Insteon protocol are also in this camp, carrying the advantage of a different bloodline: the decades-old home automation industry. Devices with these protocols are showing up at the neighborhood big-box consumer electronics and home improvement stores (e.g. Home Depot, Lowes). With their promise of energy efficiency and mesh networking, the devices leveraging these technologies are well-suited to the market.

Are these non-WiFI devices in it for the long run, or will they simply vanish overnight? Will one protocol really rule them all? And if so, will it be WiFi?  While it is too early to predict, there are some potential weaknesses in WiFi’s armor. First, let's look at its strengths.

WiFi's Strengths

Those backing WiFi do so with good reason. WiFi has some clear strengths.

Ubiquity: The biggest thing WiFi has going for it is its ubiquity. The market that is going to buy into the IoT space overlaps strongly (if not entirely) with the market that already has existing WiFi technologies. The fact of the matter is that we've already got WiFi routers in our homes. The infrastructure is there.

Open and Standardized: A tremendous advantage for WiFi is the fact that it is an open, unencumbered standard that has been implemented by a variety of chip manufacturers. Healthy competition drives innovation among makers, but WiFi's ubiquity counterbalances this by enforcing de facto compatibility.

Developer Friendly: It’s clear from my experience that developers are more comfortable using IP-based protocol stacks than special-purpose low-level protocols. This is where the Internet ecosystem really influences the Internet of Things. Most developers won’t want to spend time on unfamiliar networking technologies if they can just use simple WiFi tools and focus more of their time on other features.

Chinks in the WiFi Armor

WiFi's competitors have their own strong points. Because they are specialized, these technologies have definite advantages that general purpose WiFi simply cannot compete with.

Mesh Networking: Many of the low-power protocols use some type of mesh networking. In a nutshell, mesh networking refers to a network's ability to relay messages through nodes on the network. For example, a light switch may not only receive messages intended for it, but also pass on messages intended for other devices. This boosts message reliability and also network distance. Some varieties, like Insteon, even support dual-mode mesh networking, where messages can be sent wirelessly and over the power line.

Efficiency: WiFi is not a low-power solution, nor is its protocol stack lightweight. We know this simply from using our laptops and phones with and without WiFi enabled. This is important to recognize, because many embedded devices have smaller power supplies, less memory, and fewer CPU cycles than a laptop. WiFi simply consumes too many resources for some embedded devices to handle. However, protocols like Z-Wave are designed to run with minimal power consumption and fewer processor and memory resources.

Security: Think about all the data you can access once you’re logged into a WiFi network. There are laptops, phones, servers, and media devices that send vast amounts of data over the local WiFi network. Now add some low-power, resource strapped devices that need logins and passwords to use WiFi, but don't have significant security measures. What's the new weakest point on your network? What happens if a WiFi light bulb is compromised? What data suddenly becomes exposed on the WiFi network? A well-publicized problem with Belkin Wemo earlier this year brought this issue to the IoT world’s attention.

Contrast that security scenario with a limited, special-purpose protocol like Zigbee. In the absolute worst case, an attacker who manages to get physically close enough to a device and somehow manages to compromise it may have the ability to send messages to other Zigbee devices. In most cases, this wouldn’t be as bad as an attacker having access to all the data on the network.

The Coming Showdown

When it comes to the networking and protocol layer, what can we expect in the next few years? The final outcome is far from certain, but based on the current market trajectories here are some relatively safe bets.

WiFi For Entertainment, Cameras, and Gateways: WiFi's biggest foothold is in the media-rich IoT space. Video cameras, streamed media (including voice-interactive technologies), and data intensive apps will likely always be WiFi. Many systems also have Internet-connected hubs or gateways that make it possible to control a device or group of devices from the cloud or from a mobile device. WiFi has no current competitor for these features.

Two Protocols Enter, One Protocol Leaves: Z-Wave and Zigbee are both intensely competing for the home automation space. Often times a market segment is wide enough for competing protocols, but home automation is not one of them. Consumers will quickly tire of trying to determine whether a device's network protocol is compatible with their other stuff. Right now, Zigbee is fragmented with incompatible implementations, but they can (and probably will) fix this situation right away.

Still Room for an Open Standard: The downside to Lutron, Z-Wave, Zigbee, and Insteon is that all of them are proprietary. Unlike WiFi, which relies upon open standards, these protocols are tightly bound to vendors or consortiums. If one of these protocols becomes the first to break away from this tight coupling, it may quickly emerge as the de facto protocol.

Finally, it is unlikely that any one protocol will ultimately "win out"and be the sole wireless platform for IoT. Already, Belkin's Wemo line is moving toward a combination of WiFi and Zigbee to control their new lighting solution. The reasoning is straightforward: WiFi is an easy bridge to existing home networks, but among the devices, Zigbee is more efficient. Philips Hue is also built this way. Hubs, those ugly little hockey pucks that seem to ship with every IoT solution, are the distasteful side-effect of multi-protocol solutions.

Thus, while multiple protocols will inevitably survive and even flourish, one challenge of the budding IoT space may be clearing out enough closet space for hubs and dongles.