Ethernet continues to gain traction as
a cost-effective way to achieve higher bandwidth. With the emergence of ‘all
Ethernet’-based networks, packet-based timing synchronization is now of fundamental
importance to ensure maximum network performance as more demanding technologies
and applications are deployed. The transition to Ethernet from traditional
Plesiochronous Digital Hierarchy (PDH) and Synchronous Optical Network
(SONET)-based networks requires efficient timing synchronization techniques in
backhaul networks to synchronize base stations and avoid dropped calls as the
call is handed off from one base station to the next. Similarly Data center
networks & Electrical Sub Stations also require tighter synchronization to
ensure the accuracy and performance.
To address the synchronization needs, 1588
V2 came into existence. IEEE 1588 V2 is a protocol designed to synchronize
real-time clocks in the nodes of a distributed system that communicate using a
network.
Overview of Historical Time
Synchronization Technologies
In
a data network, time synchronization allows all of the different devices on
that network to use a common clock to coordinate all of their activities.
Network integrators currently have a number of different time synchronization
options available. Each has its own advantages and disadvantages.
The
following are the time synchronization methods: IRIG-B (AM), IRIG-B (DC
Shifted), 1PPS, GPS, NTP, IEEE 1588 v1, IEEE 1588 v2. Typical accuracy in
substation improved from 1 mS to 1 uS. The methods have variation in parameters
like date & time of the day indication, requirement of dedicated cabling
requirement, cost effectiveness and scaling with large no. of devices.
Inter-range
Instrumentation Group (IRIG): The IRIG standard defines a serial time code
format for use with serial communications networks. First standardized in 1956,
IRIG signals are a legacy technology used with older serial systems. IRIGB
205-87 is the latest update of this standard.
Network
Time Protocol (NTP): NTP is a time protocol for data networks; it was
first established in 1985. NTP relies on a hierarchical, layered system to
promulgate the current time throughout the network. NTP imposes hierarchical
tree architecture on the network to avoid cyclical dependencies.
Global
Positioning System (GPS): GPS satellites orbiting the earth use highly
accurate atomic clocks. Satellite signals carrying timekeeping information can
travel at the speed of light to receivers on the ground. These light-speed
signals are also corrected according to the principles of general relativity,
which gives each receiver on the ground highly accurate time information.
IEEE
1588 V1: This standard defines a protocol enabling precise synchronization of
clocks in measurement and control systems implemented with technologies such as
network communication, local computing and distributed objects. The protocol is
applicable to systems communicating by local area networks supporting multicast
messaging including but not limited to Ethernet.
Evolution of 1588 V2
The
Precision Time Protocol (PTP) provides a standard method to synchronize devices
on a network with sub microsecond precision. The protocol synchronizes slave
clocks to a master clock ensuring that events and timestamps in all devices use
the same time base. PTP is optimized for user-administered, distributed
systems, minimal use of network bandwidth and low processing overhead.
PTP
was originally defined in the IEEE 1588-2002 standard, officially
entitled “Standard for a Precision Clock Synchronization Protocol for
Networked Measurement and Control Systems”. In 2008 a revised standard, IEEE
1588-2008 was released. This new version, also known as PTP Version 2,
improves accuracy, precision and robustness but is not backwards compatible with the
original 2002 version.
Difference between version 1 and
version 2 of IEEE 1588
V1
does not support transparent clocks or profiles whereas V2 allows for
transparent clocks (including End-to-End and Peer-to-Peer delay options) and
industry profiles. Store and forward Ethernet switches exhibit latency times
that can vary depending on the data that the switch is currently processing.
For instance if the device is transmitting a 1500-byte packet, the latency will
be much greater than if the transmit queue was empty or transmitting a 500-byte
packet. Transparent clock mode can account for the varying latency times; the
switch timestamps the time packet as it enters, measures the residence time,
and corrects the time packet either as it leaves (one-step mode), or with a
follow-up message with the correction field in it (two-step mode). Accumulation
of switch latency or jitter errors is eliminated with transparent clock mode.
V2
introduced the delay measurement mechanism. The propagation delay time is
measured only between the switch and its upstream peer. This is an alternate
method to measuring the total end-to-end path delay from the slave clock to the
master clock that eliminates two likely problems with the previous scheme: In a
large network the end-to-end method will traverse many switches, each with
varying and unpredictable latency time that leads to timing inaccuracy and
jitter, compounded by the possibility of asymmetric data paths, Secondly, all
the end-to-end path delay request messages must be answered by the master clock
that can cause a traffic and processing bottleneck at the master in a large network.
In the peer-to-peer delay mechanism, path symmetry is guaranteed and there will
never be processing or traffic overloading due to the one to one relationship.
V1
packets are larger, making more traffic whereas V2 packets are smaller. V1 is
now completely redundant and is obsolete.
V2
introduced announce messages which improved the operation of the BMC (Best
Master Clock) algorithm which made reconfiguration faster, so V2 is more fault
tolerant.
Maintaining Synchronization using
1588 V2
In
a packet transport system, clocks communicate with each other over the
communication network using PTP. All clocks, whether master or slave, lead back
to and ultimately derive their time from – the ‘Grandmaster’ clock.
There
are 5 types of PTP clock devices.
Ordinary
Clock – A single port device that can be a Master or Slave Clock.
Boundary
Clock – A multi port device that can be a Master or Slave Clock
End
to End Transparent Clock – A multi port device that is not a Master or Slave
Clock but a bridge between the two. Forwards and corrects all PTP messages.
Correction achieved by addition of the bridge residence time into a correction
field within the header of the message.
Peer
to Peer Transparent Clock – A multiport device that is not a Master or Slave
clock but a bridge between the two. Forwards and corrects Sync and Follow up
messages only. Correction achieved by addition of the bridge residence time +
the peer to peer link delay, into a correction field within the header of the
message
Management
Node – A device that configures and monitors clocks.
Master
and slave are kept in sync by exchange timestamps, which are sent within PTP
messages. There are two types of message in the PTP protocol.
Event
Messages – Timed messages whereby an accurate timestamp is generated both at
transmission and receipt of the message.
General
Messages – Messages which do not require timestamps but may contain timestamps
for their associated event message.
There
are two mechanisms used in PTP to measure the propagation delay between PTP
ports:
The
Delay Request-Response Mechanism
This
mechanism uses the messages Sync, Delay_Req, Delay_Resp and Follow_Up.
The
Peer Delay Mechanism
This
mechanism uses the messages Pdelay_Req,
Pdelay_Resp and Pdelay_Resp_Follow_Up. It is restricted to
topologies where each peer-to-peer port
communicates PTP messages with, at most, one other such port.
There
are two phases in the normal execution of the protocol:
Phase
1 – Master-Slave hierarchy establishment
In
each port of any Ordinary or Boundary clock there is a PTP state machine. These
state machines use the ‘Best Master Clock Algorithm’ (or BMCA) to establish the
Master for the path between two ports. The statistics of the remote end of a
path are provided to each state machine by the Announce message. Since the local
clocks statistics are already known by the state machine, a comparison can be
made as to which is the best Master.
Phase
2 – Synchronizing Ordinary and Boundary Clocks (using the delay
request-response mechanism or Peer delay mechanism)
Method-1
Clock
synchronization phase starts after the Master-Slave hierarchy has been
established. This phase consists of the exchange of PTP timing messages on the
communications path between the two clocks.
There
are two parts to this synchronization method:
1.
Measurement of the propagation delay between Master and Slave using the delay
request-response mechanism.
2.
Performing the clock offset correction. Once the propagation delay is known the
Master can send Sync and optional Follow_Up messages containing its master
timestamp.
Method-2
After
the Master-Slave hierarchy has been established the clock synchronization phase
can start.
There
are two parts to this synchronization method:
1.
Peer-to-peer ports maintain a measurement of the link propagation to each peer
by using the peer delay mechanism.
2.
Performing the clock offset correction. Once the link propagation is known, the
master sends Sync and optional Follow_Up messages containing its master
timestamp.
Application of 1588 V2
GSM and UMTS Base station
Synchronization
One
of the most common applications currently being cited for 1588 V2 is for the
synchronization of various wireless telephony and data services, e.g. GSM,
UMTS, CDMA, WiMAX etc. These are gradually transitioning from a TDM-based
backhaul network to a packet-based network. The problem with eliminating the
TDM interface is that this is often used as a source of synchronization for the
base station itself. In order to permit correct handover between adjacent base
stations in the presence of Doppler shift generated by a moving mobile handset,
the RF frequency at a GSM or UMTS base station must be accurate to within 50ppb
(parts per billion) of the nominal frequency at all times When the TDM backhaul
is replaced by a packet network, the synchronization requirement is fulfilled
by 1588 V2.
Smart and Synchronized Electrical
Substation Automation
The
latest buzzword in the power industry is Smart Grid, a revolution that
promises to make power distribution more efficient, sustainable, and
cost-effective by applying information technology. The basic concept is simple:
upgrade the power grid to accomplish the same amount of work with less
electricity by reacting intelligently to changes in power supply and demand
with a more responsive, adaptable, and decentralized power distribution
network.
Smart
Grid requires the electrical infrastructure to be smart & intelligent.
This is achieved by means of front-ending computers connected to remote devices
such as switch gears, sensors, power generators, and circuit breakers through
intelligent electronic devices (IEDs). Other priority front-end computing tasks
in substations include data acquisition, data computing, and protocol
conversion between the DNP, IEC, Modbus, and other proprietary protocols used
in substation communications.
To
create a Smart Grid, all the network nodes must work together seamlessly.
That’s where time synchronization comes into picture. Accurate timekeeping
allows the network to coordinate activity more effectively. For example, one
embedded computing task is to keep precise data logs of all the substation
computers, switches, and IEDs. It’s important to keep accurate timestamps of
all the events in these data logs, which are often only milliseconds apart.
Accurate timekeeping ensures that these logs can be used to correctly manage
and diagnose any problems on the network.
Electric utilities have recognized that 1588 V2
offers network-based precision time synchronization that is reliable and
accurate enough (i.e. sub-microsecond) for use in electric power applications.
IEEE 1588 V2 is currently being considered for inclusion within Edition 2 of
the IEC 61850 standard for the design of
electrical substation automation.
In
a network based on IEEE 1588 V2, the grandmaster clock determines the reference
time for the entire substation automation system. The Ethernet switch acts as
the boundary or transparent clock, and additional devices (such as merging
units, IEDs, and protection devices) are designated as ordinary clocks. All of
these devices are organized into a master-slave synchronization hierarchy with
the grandmaster clock at the top.
Capital Markets
1588
V2 provides a foundation for accurate performance measurement and transaction
logging that is required for next generation electronic trading platforms,
exchanges, and other trading venues. It provides accurate system clock and
synchronization across server clusters required to measure application
performance in an ultra-low-latency environment such as HFT (High Frequency
Trading). Accurate measurement is the first step toward gaining advantage in a
competitive market where fast trading speed matters.
Conclusion
Synchronization
is an important part of today’s IP networks. By using 1588 V2 protocol,
carriers can achieve synchronization with accuracy matching that of alternative
solutions without the cost or need to build overlay networks required by those
solutions. This standard provides an essential technology that allows carriers
to efficiently deploy IP networks with accurate synchronization requirements
being met. Field trial showcasing excellent interoperability and performance
results, 1588 V2 is a proven solution for IP synchronization.
Ankur Rawat & Sasindra M Prabhu, Tech Mahindra
[email protected]
