[ntp:questions] NTP architecture recommendation

Richard B. Gilbert rgilbert88 at comcast.net
Sun Aug 19 11:56:28 UTC 2007

David J Taylor wrote:
> Richard B. Gilbert wrote:
> []
>>Well, time is certainly one of the four variables that a four
>>satellite fix gives you, the others being latitude, longitude, and
>>elevation, but I don't see that time is in any way equal to position!
>>UTC is the same in Paris, New York, and Los Angeles and those three
>>positions are in no way equal!
> Errors in the time delay of the signal from one satellite, whether 
> deliberate or accidental, will result in errors of position.  Selective 
> Availability (SA) is a deliberate reduction in the accuracy of the time 
> measurement capability, resulting in a limitation on the accuracy of 
> position measurement.  It would also make the UTC in Paris, Berlin and 
> Stockholm differ slightly were UTC measured purely by GPS observations.
> Cheers,
> David 

True enough.  But there are generally seven or eight satellites above 
the horizon and Motorola had a feature called "TRAIM" or Time Receiver 
Autonomous Integrity Monitoring which made use of additional satellites 
for cross checking.  It makes sense to do something like that.

In fact, ISTR at least one occasion in which a satellite "lost it's 
mind" and had to be shut down, for a time, by ground control.  The rogue 
satellite was beaten into submission and was back on line in a day or so.

More information about the questions mailing list