Last December the Bluetooth Special Interest Group (SIG) officially adopted version 4.2 of the Bluetooth core specification. Besides increasing the speed and improving privacy, the good news is that the new specification includes the Internet Protocol Support Profile (IPSP). The profile will allow Bluetooth Smart sensors to access the Internet directly via 6LoWPAN, the IETF standard which enables the transmission of IPv6 packets in networks of resource constrained devices. After the announcement, nordic soon released a protocol stack and an SDK for its nRF51 Series SoCs. The stack and the SDK support the ISPS for IPv6 based communication.
There is a clear trend in IoT networking, especially in the smart home domain: the use of 6LoWPAN as networking technology. Bluetooth 4.2, in fact, is not the only protocol to have chosen this IETF standard as a mesh network enabler. Thread, the recently announced protocol led by Nest, will also be developed on top of 6LoWPAN. The trend is also demonstrated by Zigbee which has completely been redesigned in order to support 6LoWPAN in it’s 2.0 and now 3.0 specifications.
While 6LoWPAN might not be optimal in terms of device energy consumption, it enables a more seamless integration between the IoT and the Internet. I have heard several people and read several blog posts wandering why every device needs an IP address and arguing that IP and Web technology should only be handled at application gateway level. The reason is that application gateways (or hubs, how they are commonly and in my opinion wrongly called nowadays in the IoT space), are rather complicated machines which are required to do most of the times tough (and nasty) protocol translations which can lead to operational problems. An application gateways reduces the transparency because it is aware of the application logic and is tightly coupled to the smart objects. When the IoT devices speak 6LoWPAN the communication between the devices and the Internet happens via an IP router enabled with a 6LoWPAN adaptation layer. The difference is that the application logic can be pushed into the cloud (or even to another device in another network). While the difference between an application gateway and a router might seem minimal to a non telecom expert, from the point of view of the overall architecture complexity, flexibility, interoperability and scalability it can be huge. And it is the main reason why the Internet and the Web architecture has succeed. A more detailed discussion on the advantages of embedded IP (and Web) technologies can be found on my paper Embedded Web Technologies for the Internet of Things.
Bluetooth technology is nowadays included in every smartphone on the market and as such has become the uncontested leader in those use cases requiring short range communication between a device (headphone, smartwatch, fitness bands, shoes, etc.) and the smartphone. The point-to-point non mesh nature of Bluetooth has always prevented the protocol to be generalized and extended to different use cases and markets. There have only been some (proprietary) attempts to extend Bluetooth to mesh use cases, like the CSR Mesh solution. The adoption of 6LoWPAN in the 4.2 specifications will surely change this and it is a clear sign that the Bluetooth SIG intends to penetrate other market segments, in particular the smart home. Firstly, 6LoWPAN elimintes the problem of the limited communication range, as a consequence of its mesh nture. In addition, Bluetooth 4.2 devices can bypass the smartphone and talk to the Internet directly via a router (equipped with 6LoWPAN communication). Up to now, Bluetooth devices have been able to send data to the Internet only via the smartphone. The new specification allows devices to communicate with the Internet even when the smartphone is not around, which can be often the case for smart home systems. Another reason why Bluetooth wants to penetrate the smart home is the integration between smart home and wearable devices. My wristband, shoes, socks, etc. will be able to control home devices without having to go through API’s or through web services like IFTTT. The integration of smart home and wearable computing is going to be a hot topic during the coming months and is going to unleash exciting use cases. Misfit, one of the most famous wearable device companies, has recently announced a connected light bulb and has demonstrated its interest in investing on the integration of wearable and smart home devices.
There are obvoius reasons to believe that Bluetooth can become the leader technology in the smart home and that will finally eliminate ZigBee (and other more traditional home automation protocols like Z-Wave). Bluetooth is a widely used standard, with strong interoperability capability. This is not the case for ZigBee which has made several mistakes in terms of IoT interoperability. The success of Bluetooth 4.2 in the smart home space will also depend on the battery performance of the devices. Part of the Bluetooth success is due to the significant improvement that the 4.0 specification (Bluetooth Low Energy) brought in terms of battery consumption compared to its predecessor specifications. But with 6LoWPAN onboard it might be complicated to keep very long battery lifetimes.
The penetration of Bluetooth in the smart home, as well as the little time left to ZigBee to survive in the IoT protocol space, will also depend on Thread and its success in the smart home domain. With many chances Thread will also gain a large slice of the smart home market, considered that it is promoted by companies Google, Nest, ARM, etc. However, it is not yet clear what the next moves of Thread wil be. For the moment they missed the end of 2014 as deadline to release the first protocol specifications. This, however, doesn’t mean much considered how complicated it is to define a protocol. Therefore I believe that Thread will soon come up with interesting moves to “conquer” the smart home space and the battle with Bluetooth 4.2 will be exciting. Less exciting will be, for the ZigBee fans, to see the unavoidable end of ZigBee as a smart home protocol.