This thread has been locked.

If you have a related question, please click the "Ask a related question" button in the top right corner. The newly created question will be automatically linked to this question.

Problem with using beacons and Z-Stack 1.4.3

Other Parts Discussed in Thread: Z-STACK, SIMPLICITI, TIMAC

 Hi, I am currently trying to use beacons in Z-Stack 1.4.3 on a SmartRF04EB with a CC2431EM installed and several CC2431EM+battery boards.  However when I set the stack to use beacons the Coordinator (the SmartRF board) hangs where it displays the IEEE address on the LCD and doesn't get any further.  If I use ZTool I can see that Devicestate is DEV_COORD_STARTING.  I think that I have set the stack up correctly, I am using the NETWORK_SPECIFIC stack profile ID and set NWK_MODE to NWK_MODE_STAR in nwk_globals.h and have set DEFAULT_BEACON_ORDER and DEFAULT_SUPERFRAME_ORDER in ZDapp.c to 6 and 2 respectively.  From what I have picked up from the documentation this is all I should need to set to get beacons working however I can't seem to get them working.

 Any help on the matter would be much appreciated.

 Regards Arrian.

  • hi arrian,

    beacon mode networks are not supported by zigbee, hence our z-stack does not support beacon mode networks. For mesh, it is assumed that non-beacon mode operation of the MAC is used.

  • Hi, I understand that beacons cannot be used in mesh mode however I was under the impression from the Zigbee spec that you could use beacons in star or tree networks.  Have I miss-read this? or is it just that beacons are not supported at all in any mode in Z-Stack.

     Regards Arrian.

  • beacon mode networks are not supported by zigbee, hence our z-stack does not support beacon mode networks. For mesh, it is assumed that non-beacon mode operation of the MAC is used.

     

    This is all true, but....

    1) Beacon mode Star/Tree networks are specified in ZigBee-Spec (2004, 2006, 2007)

    2) Beaconing is not supported by ZigBee Alliance

    3) Beaconing is optional

    4) Beaconing is not supported by public download version of Z-Stack 1.4.3

  • Hi OZ, correct on all accounts. In addition, this is more of an issue of the "stack profile". Currently the only supported ones are home controls (Zigbee) and Zigbee Pro. Both assume that the network mode is mesh. Cluster tree mode (which involves beaconing) is not officially supported by such a stack profile, even though it is mentioned in the spec. I believe the WSN working group is lobbying to get future support in so that battery powered routers can be achieved. But currently it stands as status quo.

  • Hi, Thanks OZ and Z for your explanation.

     Cheers, Arrian.

  • Beaconing will only work well if accurate time synchronization is present and can propagate to Zigbee nodes. Actually, all that is supported now is only a call method here but the technical prerequisites have to be present and implemented as close to real-time as possible as process delays add up to other errors in time synchronization. For those not familiar with time synchronization there are good references around the CDMA technology which (2000 1.x or better) is synchronized by GPS (=stratum 1 time). I.e. every CDMA base station is in synch with GPS time - i.e. CDMA time is GPS time.

    Similarly, the beacon is used in Zigbee to accomplish this job. Beacon networks have the advantage that they can handle more time slots and as such are more spectrum efficient. That means that you can handle more sensors per second than without beacons. This was one of the original promises of Zigbee which up until today has been unfulfilled. Ergo, at the present level Zigbee is hardly ready to handle thousands of sensors per seond.

    Naturally, multi-node hopping requires a network based synchronization protocol if not all of the nodes are within range of the beacon itseelf (and you hardly want to put a GPS onto every sensor). As you can imagine, old fashioned protocols like NTP for PCs and servers are not originally designed to do that with such great accuracy so yu might just take a more RF based approach and write your own libary to support something that works. In most cases you will find that 2x your highest rate of inaccuracy (i.e. process delays, clock / oscillator drift, multipathiing etc.) is what you get as your timeslot basis. Still even with a not so accurate synchronization of down to 10 us you will see great improvement over conventional polling.

    We are currently testing beacon with Simpliciti and are carefully optimistic...

  • Hi

    Our TIMAC support beaconing. Maybe you can be interested in evaluating the TIMAC with the beaconing feature. 

    Check out the following papers and articles on your topic interest:

    A Non-beaconing ZigBee Network Implementation and Performance Study: http://ieeexplore.ieee.org/iel5/4288670/4288671/04289207.pdf

     

    A research paper: Network Signaling Channel for Improving ZigBee Performance in Dynamic Cluster-Tree Networks http://www.hindawi.com/GetArticle.aspx?doi=10.1155/2008/456535&e=ref

    Collision-Free Beacon Scheduling Mechanisms for IEEE 802.15.4/Zigbee Cluster-Tree Wireless Sensor Networks: http://www.open-zb.net/publications/ASWN2007.pdf

    LPRF Rocks the World