Martin Gercke wrote:
If a coordinator is sending out beacon frames and children (FFD,
possible routers) associate to him, how and when are they sending
THEIR beacons for THEIR children.
There can be even more "layers" so that the children are parents for
their own children and these children are again partents ... and so on
until the restrictions of the PAN are reached (max. depth)
Yeah, we were confused by that question too. It looks like there's no
explicit answer in the 802.15.4 specification. We found some additional
information in the ZigBee specification (pages 245-247: Scheduling
beacon transmissions).
We figured out that a routing device deals with two superframes. The
first is the parent's one, and the second is its own one. So there can
be several superframes around a device (its two + neighbors'), but they
should not overlap.
If the children send their beacons at the same time as their parents
there will be interference.
So either they
- send their beacon in the CAP of their parent (small amont of time
and if this is done for their children as well the beacon periode will
shrink every time)
- send their beacon in the time after CFP and the start of the next
beacon (again, the beacon period will decrease ...)
There is a possibility to place several superframes in one beacon
interval if the latter is long enough. Superframe duration and the
beacon interval are controlled by MacSuperframeOrder and MacBeaconOrder
attributes.
Am I missing something or is beacon mode only working for 1 parent and
the associated children?
Martin
Timofei.
_______________________________________________
Tinyos-help mailing list
[email protected]
https://mail.millennium.berkeley.edu/cgi-bin/mailman/listinfo/tinyos-help
