FICORA GFI 9501

2nd edition
6.6.2003

GUIDELINES FOR IMPLEMENTATION

SYNCHRONIZATION OF THE DIGITAL

TELECOMMUNICATION NETWORK

This guideline document is related to FICORA Technical regulation

26 C/2003 M.
GFI 9501 1 (10)

Contents

CONTENTS ..................................................................................................................... 1
FOREWORD .................................................................................................................... 2
REFERENCES .................................................................................................................. 3
ABBREVIATIONS ............................................................................................................ 3
1 SYNCHRONIZATION NETWORK ARCHITECTURE........................................................... 3
1.1 National synchronization network ........................................................................ 4
1.1.1 SYNCHRONOUS METHOD .......................................................................................4
1.1.2 PSEUDO-SYNCHRONOUS METHOD ..........................................................................5
2 CONTROL OF JITTER AND WANDER WITHIN SYNCHRONIZATION NETWORKS ............. 6
3 TIMING CHARACTERISTICS OF SLAVE CLOCKS ............................................................ 6
4 TIMING CHARACTERISTICS OF PRIMARY REFERENCE CLOCKS..................................... 6
5 OPERATION AND MAINTENANCE OF SYNCHRONIZATION NETWORKS.......................... 6
5.1 Management of synchronization trails .................................................................. 6
5.2 Maintenance of synchronization trails .................................................................. 6
5.3 Performance monitoring ...................................................................................... 6
6 PLANNING OF SYNCHRONIZATION NETWORKS ........................................................... 7
7 SEC AS NODE CLOCK ................................................................................................... 9
8 SYNCHRONIZING THE TRAFFIC NETWORKS................................................................. 9
GFI 9501 2 (10)

FOREWORD

This Guidelines for implementation, Synchronization of the digital telecommunication

network contains clarifications and recommended option selections for ETSI standards related to
synchronization and also some additional recommendations related to synchronization and not
covered by the ETSI standards. The aim of this document is to ensure the due synchronization of
the digital telecommunications networks in Finland.

The overall technical provisions are given in FICORA Technical regulation 26 C/2003 M. This
guidance document gives more detailed recommendations on the use of standards and on the
implementation issues.

This guideline document has been prepared by the members of the national standardization group
for synchronization.
GFI 9501 3 (10)

REFERENCES

ETSI EN 300 462-1-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 1-1: Definitions and terminology for
synchronization networks

ETSI EN 300 462-2-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 2-1: Synchronization network architecture

ETSI EN 300 462-3-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 3-1: The control of jitter and wander within
synchronization networks

ETSI EN 300 462-4-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 4-1: Timing characteristics of slave clocks
suitable for synchronization supply to Synchronous Digital Hierarchy
(SDH) and Plesiochronous Digital Hierarchy (PDH) equipment

ETSI EN 300 462-5-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 5-1: Timing characteristics of slave clocks
suitable for operation in Synchronous Digital Hierarchy (SDH ) equipment

ETSI EN 300 462-6-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 6-1: Timing characteristics of primary refe-
rence clocks

ETSI EN 300 462-7-1 Transmission and Multiplexing (TM); Generic requirements for synch-
ronization networks; Part 7-1: Timing characteristics of slave clocks
suitable for synchronization supply to equipment
In local node applications

ETSI EG 201 793 Transmission and Multiplexing (TM); Synchronization network

engineering

ABBREVIATIONS

GPS Global Positioning System

NE Network Element
PDH Plesiochronous Digital Hierarchy
PRC Primary Reference Clock
SASE Stand Alone Synchronization Equipment
SDH Synchronous Digital Hierarchy
SEC SDH Equipment Clock
SSM Synchronization Status Message
SSU Synchronization Supply Unit
STM-N Synchronous Transport Module N

1 SYNCHRONIZATION NETWORK ARCHITECTURE

General requirements for the synchronization network architecture are specified in the ETSI
standard EN 300 462-2-1. Both synchronous and pseudo-synchronous modes can be used in the
national network.
GFI 9501 4 (10)

1.1 National synchronization network

Additional information concerning the national synchronization network is given below.

The national telecommunication network consists of networks owned by separate operators.

Digital exchanges, SDH cross connects and stand alone synchronization equipment (SASE) can
serve as nodes of the synchronization network (SSU). The distribution of timing can be done by
using 2.048 MHz or 2.048 Mbit/s or STM-N signal. Each operator network operates synchronously.
Both synchronous and pseudo-synchronous method are possible between the networks of separate
operators.

1.1.1 SYNCHRONOUS METHOD

The principle of the synchronization network structure in the case of synchronous method is shown
in figure 1.

NATIONAL
PRIMARY
REFERENCE
CLOCK (PRC)
SYSTEM

OWN CESIUM-SOURCE/

2 1 LORAN C/
GPS-SATELLITE/
OTHER COUNTRY
3

1
SECONDARY PRIMARY
CLOCK CLOCK
2

1. HIERARCHY LEVEL
------------------------------------------------
2. HIERARCHY LEVEL

REGIONAL NODE

OPERATOR 1 OPERATOR 2 *) OPERATOR 3 *)

*) The same structure as operator 1.

Figure 1. National synchronization network structure in case of synchronous method

The national synchronization network is based on the national primary reference clock (PRC)
system. The system consists of two totally separate reference clocks, which fulfill the requirements
of ETSI EN 300 462-6-1. The synchronization signal of the operator network is derived from the
national PRC system.

The owner of the national PRC system and the other operators agree the principles which are to be
applied in the maintenance of the system and in the sharing of the costs caused by it. For the time
being the owner of the national PRC system is Sonera. The PRC system consists of two separate
clocks, one with two cesium sources and a SASE and the other with one cesium source and a
SASE. The SASEs correspond with the primary and secondary clocks in Sonera’s network(see
figure 1).
GFI 9501 5 (10)

The synchronization network of the operator is connected to the PRC system using the principle
that the primary clock of the network derives the primary and secondary synchronization signal
from different outputs of the PRC system and uses physically separated transmission systems for
the transfer of the signal (1 and 2 in Figure 1).

The secondary clock of the network is a synchronization node, which derives its primary and
secondary synchronization signals from the primary clock of the network. The two synchronization
signals are transferred via independent transmission systems (① and ② in Figure 1).

The primary clock of the network may have a third alternative synchronization source, e.g. own
cesium reference, GPS satellite connection, Loran C radio network connection or connection to the
network of a foreign operator. In failure situations the networks may then function temporarily
pseudo-synchronously (3 in Figure 1).

A connection to the national PRC system is defined as the third alternative synchronization source
for the secondary clock of the network (③ in Figure 1).

The lower level nodes in the synchronization networks shall have access to the primary and the
secondary clock of the network (regional node in Figure 1).

Each operator can define if the telephone, mobile, data etc. networks are separate master-slave
systems which have on the highest level connections to the primary and secondary clocks of the
network or they form together one master-slave system. It is recommended that the latter
principle is used i.e. the nodes of the separate networks are synchronized (from the
synchronization network) mainly in the regional level.

All operators may not have direct connections from the national PRC system. Transfer of the
synchronization signals via a network owned by another operator may thus be necessary. In such
a case the parties agree the related technical and economical facts. The synchronization trails
starting from the PRC system shall be checked for conformity with the standards (refer to chapter
6).

In the networks of small regional operators and service providers the above described structure
may not be required. They may derive the synchronization signals from another operator’s
synchronization network. In such a case the parties agree the related technical and economical
facts. The synchronization networks of the parties as a whole shall meet the above-described
structure and other related standards.

It is important that the transfer of synchronization between the networks of separate operators is
configured in one direction only. It shall not be fed backwards at any point. Otherwise there is a
danger of timing loop.

Each operator and service provider is responsible for proper planning of the synchronization within
its own network.

1.1.2 PSEUDO-SYNCHRONOUS METHOD

The principles defined in the synchronous method are applied also in case of an operator using the
pseudo-synchronous method. The only difference is that the equivalent of the national PRC system
must in this case be operated and maintained by the operator itself.
GFI 9501 6 (10)

2 CONTROL OF JITTER AND WANDER WITHIN SYNCHRONIZATION NETWORKS

The specifications for jitter and wander within synchronization networks are contained in the ETSI
standard EN 300 462-3-1.

3 TIMING CHARACTERISTICS OF SLAVE CLOCKS

The specifications for timing characteristics of slave clocks are contained in the ETSI standards
EN 300 462-4-1, EN 300 462-5-1 and EN 300 462-7-1.

4 TIMING CHARACTERISTICS OF PRIMARY REFERENCE CLOCKS

The specifications for timing characteristics of primary reference clocks are contained in the ETSI
standard EN 300 462-6-1.

5 OPERATION AND MAINTENANCE OF SYNCHRONIZATION NETWORKS

5.1 Management of synchronization trails

From all the synchronization trails entering a node it is possible to create and delete
synchronization trails with a software command. Furthermore it is possible to change the priority
of synchronization trails.

The operational mode of the synchronization node is set automatically or manually.

The existing synchronization trails and their priorities can be listed with a software command.

Management operations to the synchronization trails are accomplished without interfering

operation of normal traffic handling.

5.2 Maintenance of synchronization trails

Synchronization trails monitoring is included in the alarm system. Alarms are generated at least in
the following events:

- change to the stand-by clock unit

- change of synchronization trail
- loss of synchronization trail

The event of the recovery should also be recorded.

If all the synchronization trails entering a node are lost the node shall operate in hold-over mode.

5.3 Performance monitoring

Some types of synchronization equipment (e.g. SASE) are able to online measure the wander of
their input signals. By setting thresholds a deteriorated synchronization trail can be detected and
disabled as a reference. This feature may also be used to monitor the synchronism of some
important NE in the node by measuring its output signal. The measurement results (either
threshold crossing events or complete measurement data files) may be collected by the
management system.
GFI 9501 7 (10)

6 PLANNING OF SYNCHRONIZATION NETWORKS

The planning of synchronization in the networks of the operators is fully independent to each other
with the exception of one operator providing another operator with synchronization signals (refer
to chapter 1.1.1).

The architecture of the synchronization network of an operator shall be in accordance with the
ETSI standard EN 300 462-2-1. The ETSI standard EG 201 793 deals with practical issues of
planning the synchronization network.

The clock reference signals are distributed between levels of the hierarchy via distribution trails
offered by normal SDH or PDH transmission systems. No special transport network for the
distribution of synchronizing signals is used. It shall be noted that a 2.048 Mbit/s signal crossing
SDH network shall not be used for timing distribution in the synchronization network.

The main points of the synchronization network architecture are explained next.

PRC: Primary Reference Clock PRC

SSU SSU

SSU SSU SSU SSU

Figure 2. Synchronization network topology

The topology of the hierarchical synchronization network is tree-like as shown in figure 2.

The synchronization architecture requires that the timing of all network element clocks are
traceable to a PRC and hence the principal structure is the synchronization network reference chain
as shown in figure 3. Timing is distributed via master-slave synchronization from the PRC to all
clocks in the chain.

To ensure the correct operation of the synchronization network it is important that clocks of lower
hierarchical level only accept timing from clocks of the same or higher hierarchical level and that
timing loops are avoided.

The distribution network shall be designed so that the requirements for the hierarchical network
reference chain (described below) will be met even under fault conditions.
GFI 9501 8 (10)

PRC

N x SECs

Slave
1st

K-1th Slave

For worst-case scenario

calculation purposes: K = 10
N = 20 with restriction
N x SECs that the total number of SECs
is limited to 60.

Kth Slave SSU

N x SECs

Figure 3. Synchronization network reference chain

In general, the quality of timing will deteriorate as the number of synchronized clocks in tandem
increases and hence for practical synchronization network design, the number of network elements
in tandem should be minimized. Based on theoretical calculations it is recommended that the
longest chain should not exceed 10 SSUs and 20 SECs interconnecting any SSUs with restriction
that the total number of SECs is limited to 60 (refer to figure 3).

It is preferable that all SSUs and SECs are able to recover timing from at least two synchronization
trails. The slave clock shall reconfigure to recover timing from an alternative trail if the original
trail fails. Where possible synchronization trails should be provided over diversely routed paths.

In the event of a failure of synchronization distribution, all network elements will seek to recover
timing from the highest hierarchical level clock source available. To effect this, both SSUs and
SECs may have to reconfigure and recover timing from one of their alternate synchronization
trails.

SSM and squelching may be used on SDH trails for correct reference transfer between the SSUs.
The use of SSM also makes it possible to recover timing for the SEC clocks in the chain from the
opposite direction if the signal in the original direction fails.

A general procedure in planning the synchronization network may be as follows.

If the synchronous method is used:

GFI 9501 9 (10)

Find out the connections to the national PRC-system

Plan the locations for SSUs
Plan the synchronization trails

If the pseudo-synchronous method is used:

Plan the PRC system

Plan the locations for SSUs
Plan the synchronization trails

When planning the placing for SSUs the importance of the node locations for the traffic networks to
be synchronized and the synchronization network itself is considered. The maximum number of
SEC clocks between two SSUs has also to be taken into account (refer to figure 3).

When planning the synchronization trails first the transmission systems for the transfer of
synchronization are selected. Secondly the timing configuration of the selected systems is planned
in detail.

After the detailed planning for the synchronization of a sub-network is finished the presence of
potential timing loops is checked. A simple method for that is to follow through all physical loops
(clockwise and counter-clockwise) and make sure that the reference signal loop is not closed. The
priorities of the references are ignored in the checking procedure.

7 SEC AS NODE CLOCK

As shown in figure 3 the synchronization is transported via chains of cascaded SEC clocks. E.g. for
economic reasons some nodes may be configured without SSU. The major SDH NE of the node
(which implements an SEC) may be used as the node clock. The output of this major SDH NE is
distributed via STM-N interfaces to other SDH NEs and via 2.048 MHz or 2.048 Mbit/s interface to
the other equipment.

A suitable large number of reference inputs (STM-N) should be available to such a node in order to
ensure that the SEC node almost never loses all references.

The configuration with SSU shall be considered as an alternative.

8 SYNCHRONIZING THE TRAFFIC NETWORKS

The synchronization network is a network to provide all types of telecommunication traffic

networks with reference timing signals.

As recommended before the separate networks are synchronized mainly in the regional level, i.e.
the network elements of the separate networks are synchronized from the nearest synchronization
network nodes.
The possible signals for synchronizing the equipment, which are co-located with the node clock,
are usually 2.048 MHz or 2.048 Mbit/s clock signals. However some equipment may not have
dedicated 2.048 MHz or 2.048 Mbit/s inputs for timing and therefore can not be synchronized
directly from the node clock. An acceptable alternative is to synchronize such equipment with a
synchronous traffic signal e.g. a directly connected 2.048 Mbit/s signal from the co-located digital
exchange or an STM-N signal from an SDH NE.

Some equipment at the edge of a network may be situated beyond the reach of the
synchronization network. The only practical way to synchronize such equipment is to use a traffic
signal e.g. 2.048 Mbit/s signal from a digital exchange. In this case crossing an SDH network has
to be accepted.
The distance between the exchange and the equipment to be synchronized is usually small, so that
only a single SDH island may be crossed and the probability of TU-12 pointer adjustments is very
GFI 9501 10 (10)

low. Thus re-timing the 2.048 Mbit/s signal would hardly result in better network performance but
might be the solution if problems with longer trails appear.