I attended the 75th Internet Engineering Task Force (IETF) meeting in Stockholm last week. Here is some of the latest news from IETF activities related to the Internet of Things:
6LoWPAN - The 6lowpan working group is currently moving four Internet-Drafts towards last call in the standards track (Improved Header Compression, 6LoWPAN Neighbor Discovery) and informational track (Use Cases, Routing Requirements). I presented the latest ND draft, which has good consensus and will be going through a round of technical improvements within 2 weeks. Next the WG is looking to re-charter to continue on subjects such as security, MIBs etc.
ROLL – The roll working group presented the RPL (pronounced “ripple”) routing protocol draft, which has been accepted as an official working group document today. This will be the basis for routing over low-power and lossy networks including 6LoWPAN, which still needs lots of contribution to reach a full solution.
New 6lowapp effort - We held a very successful meeting about applications in resource-constrained networks. About 60-70 people attended presentations from Carsten Bormann, Don Sturek (from Pacific Gas & Electric and ZigBee/IP) along with a set of 2-minute stand-ups. The presentation is available here. The feedback from IETF Area Directors was that there is obvious support, motivation and requirements (and that we are in a hurry) – so start working! A 6lowapp mailing list and wiki page will be coming soon, keep tuned.
There was a great presence at the IETF from IPSO members, who held several meetings during the same week. Thanks to the move of ZigBee and the energy industry towards all-IP smart energy we say many new participants in Stockholm. It is really positive to see the collaboration between the IETF, IPSO, ZigBee and Utilities in this area.
Last week the ZigBee Alliance made a ground-breaking announcement -
ZigBee Alliance Plans Further Integration of Internet Protocol Standards
What this means in practice is that future versions of ZigBee specifications will incorporate IETF 6LoWPAN and ROLL IPv6 standards. In practice the most straight-forward way to achieve this is by adapting the ZigBee Application Layer (ZAL) over UDP, as specified in http://www.ietf.org/internet-drafts/draft-tolle-cap-00.txt.
I am very excited about this – and believe this is a win-win for the whole industry. This lets the ZigBee Alliance, IPSO, the IETF and the IEEE (802.15.4) work in the same direction with much greater impact on achieving the Internet of Things. My own bet was that we would see ZigBee/IP in 2010 – great that it came a year early!
According to the press release and what I have seen happening in the industry – ZigBee has received a lot of pressure from the energy industry and chip makers to take advantage of native IP technology. This recent ZigBee alliance with Homeplug surely helped the decision as well. Texas Instruments was the first chip maker to make a press release with their strong support of ZigBee’s announcement, here at Sensinode we’re proud to be TI’s 3rd party partner for IP networking -
Texas Instruments supports ZigBee Alliance plan to integrate Internet Protocol standards for smart energy applications
Something is bothering me. I keep hearing that ZigBee and 6LoWPAN are competing technologies. The ZigBee Alliance has taken a stand to force such a confrontation. Its like comparing apples and, well, New York 
6LoWPAN = IPv6 = The Internet
Think about it. The Internet… the most successfull, innovative, massive network ever created. Now what was that Zig thing called again? Does anyone even remember the proprietary, link-specific networking protocols from the 90s?
Here is why ZigBee is not competetive, and shouldn’t be compared to 6LoWPAN and IPv6:
- ZigBee = small-scale isolated ad-hoc networking. 6LoWPAN = massively scalable networking as an end-to-end part of the Internet, it is IPv6!
- ZigBee = limited to a single radio standard. 6LoWPAN = applicable to any low-power, low-rate wireless radio (or even wired! See Watteco). IP protocols tie together heterogeneous networks.
- The only good part of ZigBee is application protocol profiles. And guess what, there is an IETF specification for using ZigBee profiles over UDP/IP. http://www.ietf.org/internet-drafts/draft-tolle-cap-00.txt
- ZigBee is not a standard, it is a special interest group. Will it be around in a few years? The IETF produces open, long-lived, standards. IPv6 will be around for 20+ years.
- Large-scale enterprise automation, M2M, metering systems etc. require end-to-end addressing, security, mobility, traffic multiplexing, reusability, maintainability, and web-services which are globally scalable… this is the kind of thing IPv6 was designed for.
I only see one option for ZigBee, and that is to get properly networked. I bet soon we’ll be seeing something called ZigBee/IP 2010.
ZigBee over UDP/IPv6
