[ntp:questions] Re: WLAN synchronization @ 10-100us

Jonathan Buzzard jonathan at uk.me.buzzard
Wed Dec 10 09:12:10 UTC 2003


On Tue, 09 Dec 2003 15:33:02 -0800, Magnus wrote:

> Dale Worley <worley at dragon.ariadne.com> wrote in message news:<87d6aypvqw.fsf at netnews.comcast.net>...
>> rovery78 at yahoo.com (Magnus) writes:
>> > In a wired network this is possible by adding customized hardware that
>> > inserts time-stamps in packets just before they reaches the physical
>> > level of the stack.
>> 
>> I have no idea if this would work, but:  Since you're talking about
>> customized hardware anyway, how about attaching a device to the server
>> that broadcasts a pulse of some sort once a second (on some unlicensed
>> frequency).  This pulse would be very simple, its only requirement is
>> that it provide an accurate (within 10 us) indication of exactly when
>> the "top of the second" is.  Then attach receivers of this pulse to
>> the clients, and use the receiver's output as a PPS source for NTP.
> 
> It could be a possible solution if synchronizing over standard
> communication protocol turnes out to be not an option. Does anyone
> know if this has been done?

Yes it is called MSF, DFC77, JJY, WWVB, TDF, or HGB, if you are use LW
frequencies. You are not however going to achieve 10us accuracy if
you ask me. There are a host of other transmissions on SW frequencies
as well. Obviously you are looking to do your own version of these,
but I don't see how you are going to beat the existing government
backed systems. 

JAB.

-- 
Jonathan A. Buzzard                 Email: jonathan (at) buzzard.me.uk
Northumberland, United Kingdom.       Tel: +44 1661-832195




More information about the questions mailing list