Hopefully, by this point in this series, the reader will have a good
understanding of IEEE 1588 technology and the ways it can be applied.
It should be clear from the discussions of applications in test and
measurement, industrial automation, and telecommunications that this is
a very active field of investigation, involving many companies and
researchers around the world.
However, it is far too early to be sure that IEEE 1588 will achieve widespread
adoption, or that the additional capabilities it brings to time-based
operations will prove sufficiently useful to displace or augment
existing technologies.
It is clear that IEEE 1588 has the potential to make a major
contribution to applications with hard real-time constraints. It is
probably not too much of a stretch to claim that IEEE 1588 will be
successful in the field of industrial automation and motion control.
The area of test and measurement is less certain, and depends as
much on how quickly the distributed architecture and peer-to-peer
communication model exemplified by the LXI specification is accepted,
as it does on the IEEE 1588 standard itself.
Telecommunications applications at this point must be considered
speculative. There are clearly areas within telecommunications where
IEEE 1588 has the potential to make significant contributions, but it
is not at all clear whether it will ultimately be adopted.
Proposed Techniques to Enable IEEE
1588 in Telecommunications
The telecommunication network impairments that degrade the performance
of an IEEE 1588 system are latency fluctuations and asymmetry. If only
frequency alignment is required, then asymmetry and the absolute value
of the latency are of no concern. However, if time transfer—epoch
alignment—is also required, then all impairments must be considered.
Asymmetry. Asymmetry impairments can arise from a variety of causes.
Optical fiber systems supporting two-way communication will have
different path lengths in the two directions as a result of chromatic
dispersion, even if the nominal line lengths are identical. In
practice, companies already correct for this effect by means of
additional fiber on the short path. This is essentially a calibration
process.
Far larger contributions to asymmetry can be expected from queuing
effects in switches. In addition, some network operators use ring
topologies in which traffic flows in only one direction, which leads to
vastly different path lengths for the forward and reverse
communications between two arbitrary points. There are also protocols
that are asymmetric by design, e.g., asymmetric digital subscriber
loops (ADSL).
If these factors can be rendered time-invariant by network
engineering techniques, then the needed corrections can be made by
calibrating the resulting network. This is not very appealing, but may
be the best that can be done.
It remains to be determined how well asymmetry can be controlled in
actual operating environments. Algie estimates that the latency across
a typical metropolitan area network is less than 20 ms. If this is the
case, then asymmetry-induced epoch offsets would be on the order of at
most 10 ms, which is probably not good enough for many of the proposed
telecommunications applications.
As discussed later, early data are discussed indicating that in some
circumstances the latency is substantially less. If it were possible to
transfer time to the edges of the metropolitan networks using existing
equipment, then it may be possible to use current IEEE 1588 techniques,
such as boundary clocks, to distribute time within an enterprise.
The most difficult environment for controlling asymmetry for the
telecommunications examples discussed earlier in this series is the
metropolitan area network, and the backhauls to wireless cell sites. In
building distribution or distribution within equipment racks, the
networks are much more likely to have controlled and stable asymmetry
properties.
Latency Fluctuations
Latency fluctuation impairments arise from a variety of sources. The
most common is queuing fluctuations in switches. These are particularly
troublesome because they are likely to be very traffic-dependent. A
second source of timing fluctuation is the equipment needed to
translate between communication protocols present in the network, e.g.,
the transition between a TDMA and a FDMA protocol, or from a T1 line to
SONET.
On a longer time scale, changes in path routing will also introduce
differences in latency. Furthermore, these fluctuations may not be the
same on the forward and reverse paths, which will further aggravate the
asymmetry problem.
The only practical solution to fluctuation is filtering the offset
and delay values computed by the slave clocks. Simple filtering is not
likely to be effective due to the time and traffic load dependence of
the magnitude and distribution of the delay values.
More sophisticated algorithms could monitor the distributions, and
make use of the holdover properties of the local oscillator to maintain
a stable time scale while the parameters for a new distribution are
computed.
It will probably be necessary to separately monitor the forward and
backward paths, since there is no reason to assume that the effects of
traffic will be uniform in the two directions.
The observation times for these filtering algorithms will be long.
If we assume that the fluctuations are uniformly distributed and have a
magnitude of about 2 ms, then the NTP curve suggests that to achieve an
Allan deviation of 10-8, observation times of about 1 day
will be required.
The early data discussed later indicate that observation times on
the order of a few hours may suffice. Observations times on this order
can be supported by quartz oscillators. If this is not the case, then
more stable oscillators such as rubidium will be required, which will
increase the cost and make the use of IEEE 1588 in these applications
much less attractive.
IEEE 1588 Timing Redundancy
Many applications in both industrial automation and telecommunications
have redundancy requirements. Redundancy requirements are found in
circumstances where loss of life, unacceptable physical damage, or loss
of revenue can result due to system failure. The measures taken by
General Electric in the design of their turbine controllers are a
classic example of the use of redundancy to mitigate these risks.
In current telecommunications systems, carriers provide alternate
paths to ensure connectivity in the face of network outages resulting
from broken links or equipment. Similar measures are implemented to
protect current telecommunications timing and synchronization services.
If IEEE 1588 is to be used to provide telecommunications timing,
then means must be devised to satisfy these redundancy requirements, in
principle by using the same techniques currently used in
telecommunications.
For example, redundant timing in the wireless applications discussed
earlier in this series would be provided by redundant IEEE 1588 time
servers, as shown in Figure 9.6 below.
Figure
9.6. IEEE 1588 Redundancy in Telecommunications (courtesy of Zarlink
Semiconductor)
Here, primary and secondary time servers would provide timing to
separate IEEE 1588 grandmaster clocks. Each IEEE 1588 slave clock would
select the appropriate IEEE 1588 grandmaster based on a set of
algorithms, which have yet to be defined.
The complete absence of a grandmaster, for example, due to a network
failure between the grandmaster and the slave, would no doubt invoke
the same holdover techniques currently used in telecommunications.
The difficulty is that in the current version of the IEEE 1588
standard at the time of this writing, there is no provision for
implementing this form of redundancy.
The current standard allows selection of the grandmaster based only
on the best master clock algorithm. Since the primary and secondary
references in a telecommunications system will be categorized by IEEE
1588 as stratum 1 clocks,2 the best master clock algorithm will segment
the IEEE 1588 topology into two disjointed regions, which is not what
is needed in this application.
The IEEE 1588 standards committee has an agenda item to provide
mechanisms for implementing the redundancy requirements for both the
industrial and telecommunications applications.
Measurements of IEEE 1588 on
Metropolitan Networks
Since the spring of 2005, four companies have collaborated on a field
trial to test the viability of IEEE 1588 in distributing time over a
metropolitan network. The initial results of these measurements were
reported to the ITU by DaveTonks of Sematech. The topology of the trial
is illustrated in Figure 9.7 below.
The metropolitan network was selected by the carrier as being typical
in terms of equipment, operation, and network traffic.
Figure
9.7. Topology of the IEEE 1588 telecommunications field trial
An IEEE 1588 master clock, using an Agilent 5071A cesium clock as a
reference, injected IEEE 1588 packets into the network. These packets
were carried on a virtual private network (VPN) to a remote office on
the network where they were returned on a different VPN.
The path consists of six switch buffers in the Cisco 6509 equipment
forming the network, plus two additional buffers in switches local to
the experiment. Other than the VPNs, no special handling of the IEEE
1588 packets was used.
The returned packets were received by a slave clock that recovered
the IEEE 1588 timing information and generated a T1 signal as an input
to the Agilent OmniBER 718 network analyzer. Both the master and slave
clocks are Semtech prototype devices. Note that the 718 is measuring
the noise of the entire IEEE 1588 system, including the master and
slave clocks, as well as the noise introduced by the metropolitan
network.
Figure
9.8. MTIE field trial data for 5 June 2005 (courtesy of Semtech)
The 718 collected TIE data at 50 samples per second. The data are
transferred via a local control computer to a data store in Agilent's
laboratory in Palo Alto, California. TIE data are collected for
approximately 24 h before a new measurement sequence is initiated.
Additional data collected from the slave clock allow related
information, such as the actual round-trip delay as a function of time,
to be archived. All collected data could be retrieved from the store by
the trial participants.
The test began on 21 April 2005, and has been collecting TIE data
more or less continuously ever since. The data were still being
collected and analyzed as this book was being prepared. The MTIE and
TDEV measurements for 5 June 2005 are shown in Figures 9.8 above and Figure 9.9 below.
Figure
9.9. TDEV field trial data for 5 June 2005 (courtesy of Semtech)
In each figure, the mask shown is the plesiochronous digital
hierarchy (PDH) mask based on ITU recommendations. The observed values
of both MTIE and TDEV data fall well below the masks. A comparison of
the TDEV and MTIE data collected as of 5 June 2005 does not show
significant day-to-day variation.
Figure 9.10 below, which is
based on data provided by Dave Roe of Semtech, shows the maximum MTIE
values for each 24-h measurement period between 21 April and 5 June
2005.
Figure
9.10. MTIE field trial summary data: 21 April to 5 June 2005 (courtesy
of Semtech)
These MTIE and TDEV data suggest that, under certain conditions,
IEEE 1588 is able to transfer frequency with the necessary phase
stability over metropolitan networks. Clearly, additional investigation
is needed before this technique can be adopted.
Although one would expect that the 6-week trial period evaluated
here encompassed typical network loading conditions, additional testing
for performance with loads designed to stress the use of IEEE 1588 in
this application must be done. Likewise, testing on additional
metropolitan networks must be done to see if either carrier practices
or differences in network equipment are important.
Since some protocols require actual time transfer, rather than only
frequency, further experiments will be required to determine the
effects of asymmetry in telecommunication networks.Dave Tonks of
Sematech presented the results of preliminary time transfer experiments
using the trial network. These results indicate that time transfer is
possible in metropolitan networks to better than 100 ns.
Specific Concerns and Likely
Outcomes
In the author's opinion, IEEE 1588 is at a critical point in its
development. As noted, the IEEE has authorized a standards effort to
extend the existing standard. The responsible IEEE working group,
P1588, was in the early phase of discussions as this book was being
prepared.
The group that produced the original standard numbered no more than
a dozen members. By contrast, the current P1588 committee regularly
draws between 30 and 45 members to each meeting.
Furthermore, the current committee represents a far more diverse set
of interests and potential application requirements than did the
original group. This is an indication that IEEE 1588 is perceived as
useful technology, but carries with it the potential for fragmentation
of the standard.
IEEE 1588 is an infrastructure standard. End users must be able to
obtain a variety of affordable IEEE 1588-enabled devices for their
applications. All IEEE 1588 based systems will require not only end
devices particular to the specific application, but also common
components such as boundary clocks, clocks linked to a GPS system,
bridges between network protocols, software tools for monitoring
performance, and, in the future, transparent clocks and other devices
yet to be conceived.
The P1588 committee must ensure that the revised standard not only
provides the needed enhancements, but does this in a way that keeps the
standard simple enough to enable manufacturers to make cost-effective
common components useful for all.
None of the application areas is large enough by itself to create
the usage numbers needed to drive down the cost of these common
components to the point where IEEE 1588 is not only technically
attractive, but also financially compelling.
This will require compromise within the P1588 group, which is always
tricky to achieve with any group. Standards that are the union of
everyone's design preferences are rarely successful, and certainly do
not serve the end customers very well.
The outcome of the P1588 work is likely to be successful. The major
enhancements under consideration offer considerable improvements for
all application areas. This should provide sufficient incentive to find
common solutions.
While predicting the outcome of any standards committee effort, much
less the response of balloters, is an art, not a science, the early
discussions indicate that the P1588 committee recognizes the need for
compromise and is willing to devote the necessary time to achieve it.
There are also technology and economic concerns. Fundamentally, IEEE
1588 is a standard that accurately synchronizes clocks. The accuracy
currently required by most applications can easily be met by IEEE 1588.
However, accuracy requirements are likely to become more, rather than
less stringent.
It is therefore imperative that efforts to achieve low or even
sub-nanosecond accuracy succeed. At this level, there are all sorts of
technology problems, such as the PHYs, asymmetry in network links, and
oscillator stability and noise, that must be addressed.
The solutions will probably require silicon, rather than FPGA or
software techniques. Again, cheap silicon requires volume, which is
further reason to make sure the standard is no more complex than
needed. Data published at the 2004 IEEE 1588 conference and more recent
data provide reason for optimism that the needed accuracies can be
achieved.
Concern has been voiced that the MII or GMII interfaces between the
PHY and MAC layers in Ethernet technology will be integrated into
silicon, and not be available for use by IEEE 1588. This is definitely
an issue that will be resolved based on the demand that can be
generated by the IEEE 1588 user community.
Improvements to IEEE 1588
In the short term, there are a number of improvements and extensions to
IEEE 1588 that will emerge from the P1588 group. These include:
1) The transparent clock.
The standard will definitely have some form of transparent clock that
will make it feasible to construct reasonably long linear IEEE 1588
topologies. Transparent clocks may also prove useful in tree topologies
involving large numbers of end devices as a way to reduce the number of
cascaded servos at the top layers of the hierarchy.
2) Short frames and short
sync interval. The combination of these will prove useful in achieving
higher accuracy, in providing advantageous cost optimization of
oscillator quality, and in enabling several proposed telecommunications
applications.
3) Layer 2 mapping. A layer
2 implementation of IEEE 1588 has been requested by almost all
application areas. It holds the potential of enabling easier
silicon-based solutions, and more efficient switch technology.
4) Gigabit implementations.
Several of these have been reported to date, and more are sure to
emerge. Implementations on fiber optic media will also appear,
particularly if IEEE 1588 finds adoption in the power industry.
5) Wireless implementations.
Although not on the current agenda for the P1588 group, it is quite
likely that serious efforts will be made over the next few years to
implement IEEE 1588 on one or more of the wireless protocols.
6) The emergence of silicon
incorporating IEEE 1588. This will be driven by increasing adoption of
IEEE 1588, and broadening of the experience base to guide the design of
chips that incorporate appropriate support for applications using IEEE
1588. Intel was one of the first to announce such a chip, and more are
likely to follow.
Regarding applications, the next few years should see significant
numbers of products and installations in the industrial automation
area. These will occur initially in motion control, but will
subsequently spread to monitoring and general control situations as
well. In test and measurement, the adoption in the data acquisition
market is very likely. Adoption in high-end test depends on the success
of the LXI effort.
In telecommunications, there will be continued field trials. If
these prove successful, then the adoption depends on the willingness of
the major equipment suppliers to provide the necessary devices, and the
identification of service areas where there is a clear reason to switch
technologies. The early results reviewed earlier are encouraging, but
not yet definitive.
Final Thoughts
IEEE 1588 provides another tool to the designer of hard real-time
systems. It will be used in conjunction with the existing practices, as
appropriate.
What IEEE 1588 has introduced is the ability to actually specify and
execute operations based on time to an accuracy that is appropriate for
the kinds of hard real-time problems discussed here.
It has been commented that hard real-time programming is difficult
in part because computer science has abstracted away the notion of
time. Perhaps the introduction of IEEE 1588 will prove to be the needed
incentive to address this issue. If so, the next few years should prove
exciting.
Used with permission of its publisher, Springer Science and
Business Media, this series of articles is based on material from "Measurement,
Control and Communication Using IEEE 1588," by John C. Eidson and
can be purchased on line.
John C. Eidson, Ph.D., received a B.S. and an M.S. from Michigan
State University and a Ph.D. from Stanford University, all in
electrical engineering. He held a postdoctoral position at Stanford for
two years, spent six years with the Central Research Laboratory of
Varian Associates, and joined the Central Research Laboratories of
Hewlett-Packard in 1972. When HP split in 1999, he transferred to the
Central Research Laboratory of Agilent
Technologies. Dr. Eidson was heavily involved in IEEE 1451.2 and
IEEE 1451.1 and is the chairperson of the IEEE 1588 standards committee
and a life fellow of the IEEE.
