rfc957.txt

RFC 相关的技术文档

Mills                                                          [Page 15]



RFC 957                                                   September 1985
Experiments in Network Clock Synchronization


   Given the considerations above, it would seem feasable for hosts to
   synchronize logical clocks to a particular power grid, but only if
   corrections were transmitted often enough to maintain the required
   accuracy and these corrections were delivered to the hosts
   essentially at the same time.  Assuming a worst-case 400-ppm slewing
   rate and one minute between correction broadcasts, for example, it
   would in principle be possible to achieve accuracies in the 20-ms
   range.  There are a number of prediction and smoothing techniques
   that could be used to inhance accuracy and reduce the overhead of the
   broadcasts.

   Host DCN3, which uses a line-frequency clock interface, was unlocked
   during the experiment period so that the offset between the PEPCO
   clock, which is locked to the eastern power grid, could be measured
   with respect to the reference host DCN1.  Host DCN7, which uses the
   same interface, remained locked to DCN1.  In spite of the previously
   noted instability of the power grid, DCN7 remained typically within
   30 ms of DCN1 and only infrequently exceeded 100 ms in the vicinity
   of large changes in system load that occured near 0800 and 1700 local
   time. Over the seven-day period from 2 July through 8 July the mean
   offset was less than a millisecond with standard deviation about 24
   ms, while the maximum was 79 ms and minimum -116 ms.

   Experiments were also carried out using ICMP Timestamp messages with
   hosts known to use line-frequency clock interfaces in California,
   Norway and Germany.  The results indicated that the western power
   grid is rather more stable than the eastern grid and that the
   overseas grids are rather less stable.  In the Oslo, Munich and
   Stuttgart areas, for example, the diurnal variation was observed to
   exceed ten seconds.

4.3.2.  On Clocks Synchronized via Network Links

   As mentioned previously, all network links used to synchronize the
   clocks were carrying normal data traffic throughout the experiment
   period.  It would therefore be of interest to investigate how this
   affects the accuracy of the individual clocks.

   Table 2 summarizes the mean and standard deviation of the measured
   offsets between the WWVB radio clock and various hosts as shown in
   Figure 2.  Measurements were made over the 24-hour period for each of
   several days during the experimental period.  Each entry shown in
   Table 2 includes the mean of the statistic over these days, together
   with the maximum variation.





Mills                                                          [Page 16]



RFC 957                                                   September 1985
Experiments in Network Clock Synchronization


      Host  Mean          Deviation    Link Type and Speed        
      ----------------------------------------------------------- 
      DCN1  .08/.02       0.53/.02     WWVB radio clock (1200 bps)
      DCN5  -13.61/.04    1.1/0.4      Ethernet (10 Mbps)         
      DCN6  0.27/.18      5.8/1.0      DDCMP (4800 bps)           
      FORD1 38.5/1.6      2.5/0.5      DDCMP (9600 bps)           

                       Table 2. Link Measurements

   The departure of the mean shown in Table 2 from zero is related to
   the drift of the crystal oscillator used in the hardware interface
   (see Table 1).  As described previously, FORD1 was synchonized to the
   GOES radio clock with neglible offset, so that the mean and standard
   deviation shown can be accurately interpreted to apply to the GOES
   radio clock as well.

   The results show that the uncertaincies inherent in the
   synchronization algorithm and protocols is in the same order as that
   of the reference clocks and is related to the speed of the links
   connected the reference hosts to the other hosts in the network.
   Further discussion on the FORD1/GOES statistics can be found in the
   next section.

   Further insight into the error process can be seen in Table 3, which
   shows the first derivative of delay.

                 Host    Dev     Max     Min     Error 
                 ------------------------------------- 
                 DCN3    2.3     12      -17     10    
                 DCN5    1.5     45      -45     5     
                 DCN6    9       94      -54     40    
                 DCN7    1.4     6       -7      5     
                 FORD1   3.4     68      -51     15    

                   Table 3. First Derivative of Delay

   The mean and standard deviation of delay were computed for all hosts
   on a typical day during the experimental period.  In all cases the
   magnitude of the mean was less than one.  The standard deviation,
   maximum and minimum for each link is summarized by host in Table 3.
   A common characteristic of the distribution in most cases was that
   only a handful of samples approached the maximum or minimum extrema,
   while the vast majority of samples were much less than this.  The
   "Error" colum in Table 3 indicates the magnitude of the estimated
   error when these extrema are discarded.




Mills                                                          [Page 17]



RFC 957                                                   September 1985
Experiments in Network Clock Synchronization


   A very interesting feature of the observations was the unexpectedly
   low standard deviation of DCN3, which was locked to the power grid
   and thus would be expected to show wide variations.  Upon analysis,
   this turned out to be a natural consequence of the fact that the
   Hello messages are generated as the result of interrupts based on the
   line frequency when the local clock had just been incremented by
   1/60th of a second.

   The synchronizing protocol and implementation were carefully
   constructed to minimize the loss of accuracy due to sharing of the
   network links between data and control traffic, as long as sufficient
   resources (in particular, packet buffers) are available.  Since the
   various network links shown in Figure 2 operate over a wide range of
   rates, it is possible that undisciplined bursts of traffic can swamp
   a host or gateway and precipitate a condition of buffer starvation.

   While most hosts using paths through the experimental configuration
   were relatively well-disciplined in their packetization and
   retransmission policies, some Unix 4.2bsd systems were notorious
   exceptions.  On occasion these hosts were observed sending floods of
   packets, with only a small amount of data per packet, together with
   excessive retransmissions.  As expected, this caused massive
   congestion, unpredictable link delays and occasional clock
   synchronizing errors.

   The synchronizing algorithms described above successfully cope with
   almost all instances of congestion as described, since delay-induced
   errors tend to be isolated, while inherent anti-spike and smoothing
   properties of the synchronizing algorithm help to preserve accuracies
   in any case.  Only one case was found during the ten-day experiment
   period where a host was mistakenly synchronized outside the
   linear-tracking window due to congestion.  Even in this case the host
   was quickly resynchronized to the correct time when the congestion
   was cleared.

4.3.3.  On the Accuracy of Radio Clocks

   One of the more potent motivations for the experiments was to assess
   the accuracy of the various radio clocks and to determine whether the
   WWV radio clock was an appropriate replacement for the expensive WWVB
   or GOES clocks.  A secondary consideration, discussed further in the
   next section, was how the various clocks handled disruptions due to
   power interruptions, leap seconds and so forth.






Mills                                                          [Page 18]



RFC 957                                                   September 1985
Experiments in Network Clock Synchronization


4.3.3.1.  The Spectracom 8170 WWVB Radio Clock

   As the result of several years of experience with the WWVB radio
   clock, which is manufactured by Spectracom Corporation as Model 8170,
   it was chosen as the reference for comparison for the GOES and WWV
   radio clocks.  Washington, DC, is near the 100-microvolt/meter
   countour of the WWVB transmitter at Boulder, CO, well in excess of
   the 25-microvolt/meter sensitivity of the receiver.  The antenna is
   located in a favorable location on the roof of a four-storey building
   in an urban area.

   Using the data from the instruction manual, the propagation delay for
   the path from Boulder to Washington is about 8 ms, while the
   intrinsic receiver delay is about 17 ms.  The clock is read via a
   1200-bps asynchronous line, which introduces an additional delay of
   about 7 ms between the on-time transition of the first character and
   the interrupt at the middle of the first stop bit.  Thus, the WWVB
   radio clock indications should be late by 8 + 17 + 7 = 32 ms relative
   to NBS standard time.  While it is possible to include this delay
   directly in the clock indication, this was not done in the
   experiments.  In order to account for this, 32 ms should be
   subtracted from all indications derived from this clock.  The
   uncertaincy in the indication due to all causes is estimated to be a
   couple of milliseconds.

4.3.3.2.  The True Time 468-DC GOES Radio Clock

   The GOES radio clock is manufactured by True Time Division of
   Kinemetrics, Incorporated, as Model 468-DC.  It uses the
   Geosynchronous Orbiting Environmental Satellite (GOES), which
   includes an NBS-derived clock channel.  Early in the experiment
   period there was some ambiguity as to the exact longitude of the
   satellite and also whether the antenna was correctly positioned.
   This was reflected in the rather low quality-of-signal indications
   and occasional signal loss reported by the clock and also its
   apparent offset compared with the other radio clocks.

   Table 4 shows a summary of offset statistics for the GOES radio clock
   by day (all day numbers refer to July, 1985).










Mills                                                          [Page 19]



RFC 957                                                   September 1985
Experiments in Network Clock Synchronization


                 Day     Mean    Dev     Max     Min   
                 ------------------------------------ 
                 2       31.6    9.4     53      -76  
                 3       19.8    22.1    53      -64  
                 4       42.8    17.1    >150    19   
                 5       39.3    2.2     54      -45  
                 6       37.8    2.7     53      19   
                 7       62.2    13.0    89      22   
                 8       38.2    2.8     90      -7   

                   Table 4. GOES Radio Clock Offsets

    On all days except days 5, 6 and 8 long periods of poor-quality
   signal reception were evident.  Since the antenna and satellite
   configuration are known to be marginal, these conditions are not
   considered representative of the capabilities of the clock.  When the
   data from these days are discarded, the mean offset is 38.4 ms with
   standard deviation in the range 2.2 to 2.8.  The maximum offset is 90
   ms and the minimum is -45 ms;  however, only a very small number of
   samples are this large - most excursions are limited to 10 ms of the
   mean.

   In order to compute the discrepancy between the GOES and WWVB clocks,
   it is necessary to subtract the WWVB clock delay from the mean
   offsets computed above.  Thus, the GOES clock indications are 38.4 -
   32 = 6.4 ms late with respect to the WWVB clock indications.  which
   is probably within the bounds of experiment error.

4.3.3.3.  The Heath GC-1000 WWV Radio Clock

   The WWV radio clock is manufactured by Heath Company as Model
   GC-1000.  It uses a three-channel scanning WWV/WWVH receiver on 5, 10
   and 15 MHz together with a microprocessor-based controller.  The
   receiver is connected to an 80-meter dipole up about 15 meters and
   located in a quiet suburban location.  Signal reception from the Fort
   Collins transmitters was average to poor during the experiment period
   due to low sunspot activity together with a moderate level of
   geomagnetic disturbances, but was best during periods of darkness
   over the path.  The clock locked at one of the frequencies for
   varying periods up to an hour from two to several times a day.

   The propagation delay on the path between Fort Collins and Washington
   is estimated at about 10 ms and can vary up to a couple of
   milliseconds over the day and night.  While it is possible to include
   this delay in the clock indications, which are already corrected for