Home arrow Blog arrow Zigbee arrow The Differences Between 802.15.4-2003, 2006, a, c, d
The Differences Between 802.15.4-2003, 2006, a, c, d | Print |
Written by Akiba   
Sunday, 23 March 2008
There is still a lot of confusion between the different versions of the 802.15.4 specification.When you mention it, most people immediately think of the original 802.15.4 spec which is otherwise known as 802.15.4-2003. This was the spec that laid the foundation for low-rate wireless PANs (LR-WPANs). The specification described the radio, modulation, bitrate, headers, protocol, and services.

The original 2003 spec had three different radios to choose from when implementing it: 868 MHz, 915 MHz, and 2.4 GHz. By far, the 2.4 GHz radio dominated the product offerings, so much so that there are a many new papers discussing Zigbee co-existence with other 2.4 GHz radio technologies (802.11, Bluetooth).

The main reason that the 2.4 GHz spectrum came to dominate was that 2.4 GHz is a free spectrum in almost the entire world, and also the 2003 spec mandated that the 2.4 GHz would be allowed to run at 250 kbps while the 868/915 MHz frequencies had to run at lower frequencies due to their allowed modulation schemes. For reference, the 2003 spec limited the 868 MHz frequency to run at 20 kbps and the 915 MHz band to 40 kbps using BPSK modulation. Unfortunately, the 868/915 MHz radios never recovered from this and there is still a scarcity of radios at these frequencies.

In 2006, the LR-WPAN working group introduced revisions to the original 802.15.4 specification. The original name of the revisions was called 802.15.4b however since alphabet letters are less exciting than numbers, the industry ended up calling this the 802.15.4-2006 and relegated the old spec to the similarly innovative name of 802.15.4-2003.

The biggest contribution the 2006 spec did was to give more flexibility to the radios. See pix below. New modulation schemes were introduced for the radios in the 868/915 MHz spectrum, most likely because nobody was using them. One of the mistakes that they made in the 2003 spec is that the modulation schemes were different between the 2.4 GHz and 868/915 MHz frequencies which meant that not only did you need to change the local oscillator for the modulation frequency, you also needed to change the modulator/demodulator block in the chip when you processed the signals. This basically meant that you would need separate chips to support the 868/915 MHz and 2.4 GHz frequencies or you would have a huge chip with a large portion of the radio unused all the time.

The interesting thing about the 2006 spec is that one of the radio modulation combinations they offer is to have all of the radios (868/915/2400) using the same OQPSK scheme. This has the potentially interesting benefit of allowing a universal radio that can span all three frequencies. This radio would be able to work in North America at 915 MHz or 2.4 GHz @ 250 kbps, or Europe at 868 MHz @ 100 kbps and 2.4 GHz @250 kbps. Running at frequencies other than 2.4 GHz would have the nice benefit of not being subject to all the coexistence issues with 802.11 that people are worried about. 

802.15.4-2003 Radios

802.15.4-2006 Radios

Some of the other modification they made to the 2006 spec are as follows. These are taken from the Introduction section in the 802.15.4-2006 spec:

  • Frame timestamps
  • Added method for communication revision level on a frame by frame basis.
    • Note: They added a new field called frame version in the FCF for the MAC Header
  • Beacon scheduling
  • Broadcast message syncrhonization
  • Security improvements
  • GTS support made optional
    • Note: Thank God! (or whoever you believe in)
  • Allow manual control of the receiver
  • Simplify passive and active scan procedures
  • Allow for more flexibility in the CSMA-CA algorithm
  • Reduce association time in non-beacon networks

There is also an amendment to the 802.15.4-2006 spec called 802.15.4a which was approved in 2007. I suspect that the 802.15.4a Task Group was started before the 802.15.4b (802.15.4-2006) TG which is why they got the 'a' moniker. The 802.15.4a amendment defines alternative PHYs for certain applications that require:

  • Location Awareness
    • Asset tracking, Motion Detection and Tracking
  • Extended Range
    • Applications where multiple routers/repeaters cannot be implemented (don't we all need this?)
  • Enhanced Robustness 
    • Mission critical applications that require fewer retransmissions
  • Enhanced Mobility
    • Drive by meter reading, toll reading, sensors on moving objects
The alternative PHYs are a UWB and a 2.4 GHz Chirp Spread Spectrum PHY. For more information on the UWB PHY for 802.15.4, you can check out this link . Actually, is it me or does it sound like these requirements are pretty broad and generic? Anyways I haven't heard of any chips that support 802.15.4 UWB yet and I'm pretty sure they won't come out unless Zigbee takes off.

Along with the 2003, 2006, and a specs, there are also some other amendments in the pipeline. There are two PHY customization efforts going on which are called 802.15.4c and 802.15.4d. 802.15.4c is studying the possibility of creating a PHY targeted at the newly opened 314-316 MHz, 430-434 MHz, and 779-787 MHz bands in China. 802.15.4d is defining an amendment to the spec to support the new 950-956 MHz frequency band in Japan.

I'll probably discuss more about 802.15.4c and d in another post, especially 802.15.4d since I'm actually living in Japan. Until then, I need to get started on doing my taxes. Ugh.

Hits: 10521
Comments (2)Add Comment
802.15.4c and 802.15.4d approved
written by Dmitry Eremin-Solenikov, July 18, 2009
It seems, that according to IEEE RevCom meeting minutes, 802.15.4c and 4d are approved. They are available in the IEEE shop, however they still aren't available under get802 smilies/sad.gif
report abuse
vote down
vote up
Votes: +0
written by Akiba, July 18, 2009
Oh, sweet. I'm going to be all over 802.15.4d since its targeted at Japan. I don't think it's going to be a big issue if the specs aren't released yet since they're only PHY specs. You would pretty much only need them if you're designing a radio, in which case, the cost of the spec is nothing compared to the cost fabricating the IC smilies/smiley.gif From a software point of view, I'd just use standard 802.15.4 now and when chips come out that support China/Japan, then you can just adjust your driver.
report abuse
vote down
vote up
Votes: +0

Write comment

  No Comments.

< Prev   Next >