[ntp:questions] UK report on GPS vulnerabilities seems to overlook NTP

Chris Albertson albertson.chris at gmail.com
Tue Mar 8 23:09:34 UTC 2011


On Tue, Mar 8, 2011 at 1:14 PM, Uwe Klein
<uwe_klein_habertwedt at t-online.de> wrote:

> IMHO the basic concept of your system is broken when you have
> sync to such high requirements and need external infrastructure
> to achieve this.
> this then is an extremely fickle system that lacks robustness.

I can't defend the design of CDMA cell technology.  But I'm sure a lot
of it was driven by trying to get as many calls as possible into a
limited bandwidth.      Another requirement for precision timing comes
from the need to measure the signal's speed of light delay.  They use
this to locate a phone by noting the differences in the delay to
several towers   A uS is about 1000 feet so they need to do this far
better than to a uS.

Does it lack robustness?  Some phone companies publish their system
availability statistics.  I guess we could look it up.  No need to
speculate if such a system would work or not.

Timing is actually simple and robust.   The central part is a very
stable local 10MHz oscillator.  All the timing is derived from that
local source.  Then they have A GPS receiver that outputs one pulse
per second.  Every second they measure the time from the leading edge
of the pulse to the next zero crossing of the oscillator  and
periodically adjust the oscillator frequency to keep that time a
constant.   It is robust in that if GPS goes away all that happens is
the oscillator is no longer measured.  But if it is well built, the
oscillator will run correctly for a long time without adjustment.
The system does not crash if GPS goes away.

=====
Chris Albertson
Redondo Beach, California



More information about the questions mailing list