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

jimp at specsol.spam.sux.com jimp at specsol.spam.sux.com
Tue Mar 8 18:59:59 UTC 2011


unruh <unruh at wormhole.physics.ubc.ca> wrote:
> On 2011-03-08, jimp at specsol.spam.sux.com <jimp at specsol.spam.sux.com> wrote:
>> JohnAllen <johnbenallen at gmail.com> wrote:
>>> Maybe I read this too quickly, but the report published today by the
>>> UK Royal Academy of Engineering (see
>>> http://www.raeng.org.uk/news/publications/list/reports/RAoE_Global_Navigation_Systems_Report.pdf
>>> and also the BBC coverage at http://www.bbc.co.uk/news/science-environment-12668230)
>>> seems to be saying that many organisations are vulnerable to GPS
>>> failures because their IT systems rely on GPS for precise time.
>>> 
>>> Can this be true? I would have thought that most systems are using
>>> NTP, and synchronising with diverse enough time sources that
>>> unavailable or incorrect GPS time would not cause short-term problems.
>>> 
>>> The relevant part of the report is on pages 13-14, where it says:
>>> 
>>> "GNSS timing is important for telecommunications applications.
>>> Synchronous
>>> technologies are much more efficient than asynchronous technologies
>>> but require a
>>> time source with appropriate accuracy, stability and reliability to
>>> operate effectively
>>> or at all, and GNSS can provide this. While ground-based clocks are
>>> accurate enough
>>> for this purpose (especially with the availability of chip scale
>>> atomic clocks (CSAC)),
>>> the synchronisation of many such clocks is problematic. GPS allows the
>>> derivation of
>>> synchronised UTC through resolving the signals from a number of
>>> satellites at a
>>> known position. Only a ???good guess??? of the current time is required
>>> and quartz clocks
>>> have therefore been adequate for this process making synchronous time
>>> keeping
>>> significantly more cost effective.
>>> 
>>> The use of time can be split into three clear and separate aspects:
>>> frequency
>>> control, time of day and common epoch (usually UTC) time slot
>>> alignment (also
>>> known as ???Phase???).
>>> Stability of radio communications transmission, constant digital traic
>>> low, time
>>> slot alignment and traditional services over next generation Ethernet
>>> based
>>> infrastructure are some of the features that good time and timing
>>> bring to
>>> communications networks.
>>> Financial systems increasingly need precise time stamping to
>>> prioritise trades and
>>> to provide an audit trail."
>>> 
>>> NTP is not mentioned anywhere in the report.
>>
>> Nor would I expect it to be.
>>
>> There is a big difference between keeping a computer's time of day clock
>> set to the current time (NTP) and maintaining timing or frequency control
>> in a telecom system.
> 
> And exactly what is that difference? While ntp is perhaps too slow to
> respond to local frequency changes, how do you see the difference
> between keeping a computer's idea of local time accurate from keeping a
> telecom's idea of local time accurate?

Most telecom systems care very little that it is exactly 12:34:56 Tuesday
and a lot that the leading edge of the XYZ sync pulse occurs every ABC
milliseconds and is DEF milliseconds wide, for example.

The difference is the difference between "time" and "timing".

Some systems don't care what the time of day is at all but do care about
timing.


-- 
Jim Pennino

Remove .spam.sux to reply.




More information about the questions mailing list