Doug Ratcliffe wrote:
Actually, they have implemented a CSMA/CA bypass on their new 3.0 beta
versions, using their NStreme Polling protocol.
Cool.
GPS Sync has been very high up on their list, however, the issue at the
moment is that conventional serial GPS units lack the necessary timing
precision for anything other than raw timesync.
I can see where that would be the case 4800 baud is not going to be more
accurate enough for a TDM system which is running "fine" slices. "Fat"
time slices would work if you limited the number of systems you were
willing to negotiate sync with. You wouldn't get too much waste using a
less than optimal clock that way. What accuracy are they seeing on
serial GPS systems?
That said, I can't see why they couldn't sync based on ethernet broadcast
packets (lets say, the master radio sends a signal to the slave radios to
transmit). That would only work on a per-tower basis, but a master GPS sync
could sync up on a lesser precision basis, more often, allowing all the
towers to stay fairly in sync.
If you only had one clocking unit in the network the network would not
need GPS sync. The hardware clock on the clocking unit would be
accurate enough. The clocking unit could broadcast sync to all
associated units. This would probably require a star topology but it
would work just fine, especially if you used GPS to tell each unit how
far it was from the master clock so that it could account for RF
propagation delay.
-m-
--
WISPA Wireless List: [email protected]
Subscribe/Unsubscribe:
http://lists.wispa.org/mailman/listinfo/wireless
Archives: http://lists.wispa.org/pipermail/wireless/
