ascom synchronization

Issue 05/01 Slide 1.1 http://www.oscilloquartz.com Oscilloquartz SA

The Synchronisation ofTelecommunications Networks


ContentsContentsI. The Need for Synchronisation

II. Characterizing Synchronisation Quality

III. Synchronisation Distribution: General Principles

IV. Synchronisation Distribution: SDH/SONET-basedSolution

1. Elements

2. Architecture

3. Synchronisation Status Message (SSM)

V. Synchronisation Distribution: GPS-based andMixedSolution

I. The Need for Synchronisation

II. Characterizing Synchronisation Quality

III. Synchronisation Distribution: General Principles

IV. Synchronisation Distribution: SDH/SONET-basedSolution

1. Elements

2. Architecture


V. Synchronisation Distribution: GPS-based andMixedSolution


ContentsContents

VI. Synchronisation Distribution: From Co-operating Network

VII. Summary on Standards

VIII. How to Synchronize Mixed TechnologyNetworks

1. Mixed Technology Network Example

2. SDH and SONET Networks

3. The Public Switched Telephony Network

VI. Synchronisation Distribution: From Co-operating Network

VII. Summary on Standards

VIII. How to Synchronize Mixed TechnologyNetworks


2. SDH and SONET Networks

3. The Public Switched Telephony Network


ContentsContents

VIII. How to Synchronize Mixed TechnologyNetworks (cont ’d)

4. ATM Networks

5. Optical Networks

6. GSM and UMTS-FDD Radio Access Networks

7. cdmaOne and cdma2000 Radio Access Networks

VIII. How to Synchronize Mixed TechnologyNetworks (cont ’d)

4. ATM Networks

5. Optical Networks


7. cdmaOne and cdma2000 Radio Access Networks


AbbreviationsAbbreviations» ADM Add Drop Multiplexer

» ATM Asynchronous TransferMode

» BITS Building Integrated TimingSupply

» bit/s bits per second1kbit/s = 1,000 bits/s1Mbit/s = 1,000,000 bits/s1Gbit/s = 1,000,000,000 bit/s

» CBR Constant Bit Rate

» CDV Cell Delay Variation

» DNU Do Not Use

» DXC Digital Cross-Connect

» f frequency

» FIFO First-In First-Out

» GPS Global Positioning System

» GUI Graphical User Interface

» HSC High Stability Clock

» ADM Add Drop Multiplexer

» ATM Asynchronous TransferMode

» BITS Building Integrated TimingSupply

» bit/s bits per second1kbit/s = 1,000 bits/s1Mbit/s = 1,000,000 bits/s1Gbit/s = 1,000,000,000 bit/s

» CBR Constant Bit Rate

» CDV Cell Delay Variation

» DNU Do Not Use

» DXC Digital Cross-Connect

» f frequency

» FIFO First-In First-Out

» GPS Global Positioning System

» GUI Graphical User Interface

» HSC High Stability Clock

» Hz Hertz (cycles per second)1mHz = 0.001Hz1 µHz = 0.000001Hz

» ITU International TelecommunicationsUnion

» k kilo = 1,000

» LOS Loss Of Signal

» LT Line Terminal

» MTIE Maximum Time Interval Error

» NE Network Element

» OC-N Optical Carrier level N

» OS Operating System

» p pointer

» PABX Private Automatic BranchExchange

» PDH Plesiochronous Digital Hierarchy

» PEC Plesiochronous Equipment Clock

» PLL Phase Locked Loop

» Hz Hertz (cycles per second)1mHz = 0.001Hz1 µHz = 0.000001Hz

» ITU International TelecommunicationsUnion

» k kilo = 1,000

» LOS Loss Of Signal

» LT Line Terminal

» MTIE Maximum Time Interval Error

» NE Network Element

» OC-N Optical Carrier level N

» OS Operating System

» p pointer

» PABX Private Automatic BranchExchange

» PDH Plesiochronous Digital Hierarchy

» PEC Plesiochronous Equipment Clock

» PLL Phase Locked Loop


AbbreviationsAbbreviations» PRC Primary Reference Clock

» ps/km/°C pico seconds per kilometreper degree centigrade

» QL Quality Level

» QOS Quality Of Service

» s second1ms = 0.001s

» SASE Stand Alone Synchronous Equipment

» SD Synchronisation Distribution

» SDH Synchronous Digital Hierarchy

» SC SONET Clock

» SEC Synchronous Equipment Clock

» SETG Synchronous Equipment TimingGenerator

» SETS Synchronous Equipment Timing Source

» sin Sine function

» SOH Section OverHead

» SONET Synchronous Optical Network

» SRTS Synchronous Residual TimeStamp

» PRC Primary Reference Clock

» ps/km/°C pico seconds per kilometreper degree centigrade

» QL Quality Level

» QOS Quality Of Service

» s second1ms = 0.001s

» SASE Stand Alone Synchronous Equipment

» SD Synchronisation Distribution

» SDH Synchronous Digital Hierarchy

» SC SONET Clock

» SEC Synchronous Equipment Clock

» SETG Synchronous Equipment TimingGenerator

» SETS Synchronous Equipment Timing Source

» sin Sine function

» SOH Section OverHead

» SONET Synchronous Optical Network

» SRTS Synchronous Residual TimeStamp

» SSM Synchronisation Status Message

» SSU Synchronisation Supply Unit

» STM-N Synchronous Transport Modulelevel N

» STS-N Synchronous Transport Signallevel N

» TF Transfer Function

» TIE Time Interval Error

» TS0 Time Slot Zero

» VBR Variable Bit Rate

» VCO Voltage Controlled Oscillator

» VC Virtual Container

» VT Virtual Tributary

» VTG Virtual Tributary Group

» UN Undefined

» W Wander

» w radians per second

» d variation

» SSM Synchronisation Status Message

» SSU Synchronisation Supply Unit

» STM-N Synchronous Transport Modulelevel N

» STS-N Synchronous Transport Signallevel N

» TF Transfer Function

» TIE Time Interval Error

» TS0 Time Slot Zero

» VBR Variable Bit Rate

» VCO Voltage Controlled Oscillator

» VC Virtual Container

» VT Virtual Tributary

» VTG Virtual Tributary Group

» UN Undefined

» W Wander

» w radians per second

» d variation


The Need for SynchronisationThe Need for Synchronisation


Frequency synchronisationFrequency synchronisation

System A

t

t

Clock signal of system A

Clock signal of system B

System B

TA = 1 / fA

TB = 1 / fB

fA = fB


Phase synchronisationPhase synchronisation

System A

t

t

Clock signal of system A

Clock signal of system B

System B

! !!


Time synchronisationTime synchronisation

System A

t

t

Time signal of system A

Time signal of system B

System B

14/01/0008:34:56

14/01/0008:34:57

14/01/0008:34:55

14/01/0008:34:55

14/01/0008:34:56

14/01/0008:34:57


Where do we need synchronisation?Three examples

Where do we need synchronisation?Three examples

»Public Switched Telephone Networks

»SONET and SDH transport networks

»Cellular mobile telecom networks

»Public Switched Telephone Networks

»SONET and SDH transport networks

»Cellular mobile telecom networks


Public Switched Telephone Network:Synchronous Multiplexing

Public Switched Telephone Network:Synchronous Multiplexing

30 x 64 kbit/s 1 x 2.048 Mbit/s

Multiplexer

30

1

21 2 30


Public Switched Telephone Network:Space-Time Switching

Public Switched Telephone Network:Space-Time Switching

Switch

1 2 3 54 6

1 2 3 54 6

N x 2.048 Mbit/s N x 2.048 Mbit/s


What is a slip?What is a slip?

»A slip occurs when a buffer over- or underflows due to differences in timing

»A slip occurs when a buffer over- or underflows due to differences in timing

Incoming data rate

Outgoing data rate

SlipSlipIncoming data rate


Some services affected by slipsSome services affected by slips» Voice» Uncompressed - only 5% of slips lead to clicks» Compressed - a slip will cause an audible click

» Facsimile» A slip can wipe out several lines

» Modem» A slip can cause several seconds of drop out

» Compressed video» A slip can wipe out several lines» More slips can freeze frames for several

seconds

» Encrypted/compressed data protocol» Slips will reduce transmission throughput

» Voice» Uncompressed - only 5% of slips lead to clicks» Compressed - a slip will cause an audible click

» Facsimile» A slip can wipe out several lines

» Modem» A slip can cause several seconds of drop out

» Compressed video» A slip can wipe out several lines» More slips can freeze frames for several

seconds

» Encrypted/compressed data protocol» Slips will reduce transmission throughput


Slip rate due to frequency deviationSlip rate due to frequency deviation

»Slip rate = fractional freq. dev / frame duration

»For 2 Mbit/s signals, frame duration =125 microseconds :»10-11 = 1 slip in 4.8 months»10-10 = 1 slip in 14.5 days»10-9 = 1 slip in 1.45 days»10-8 = 6.9 slips per day»10-7 = 2.9 slips per hour»10-6 = 28.8 slips per hour»10-5 = 4.8 slips per minute

»Slip rate = fractional freq. dev / frame duration

»For 2 Mbit/s signals, frame duration =125 microseconds :»10-11 = 1 slip in 4.8 months»10-10 = 1 slip in 14.5 days»10-9 = 1 slip in 1.45 days»10-8 = 6.9 slips per day»10-7 = 2.9 slips per hour»10-6 = 28.8 slips per hour»10-5 = 4.8 slips per minute


A customer affected by slips:A customer affected by slips:


SDH/SONET Transport Networks:Nominally Synchronous MultiplexingSDH/SONET Transport Networks:Nominally Synchronous Multiplexing

63 x E1 1 x STM-1

Multiplexer

63

1

2OH 1 2 63


Wander induced by pointer activityWander induced by pointer activity

t

TIE of E1 signal

t

TIE of E1 signal

PRC

NE

NE

NE

SDH/SONET Network

ADME1

ADME1

Synchronisation distributiontrails affected by wander


SDH/SONET Transport Networks:Nominally Synchronous MultiplexingSDH/SONET Transport Networks:Nominally Synchronous Multiplexing

»PDH tributaries need not be synchronous with the SDH aggregates (pointer technique).

»However, relative wander between the incoming PDH tributary and the SDH aggregates induce wander on the outgoing PDH tributary (pointer adjustments!).

» If excessive, this tributary wander causes slips.

»PDH tributaries need not be synchronous with the SDH aggregates (pointer technique).

»However, relative wander between the incoming PDH tributary and the SDH aggregates induce wander on the outgoing PDH tributary (pointer adjustments!).

» If excessive, this tributary wander causes slips.


Cellular Mobile Telecom NetworksCellular Mobile Telecom Networks

BTS

BTS

BTS

BTS

Successful handover requires synchronisation between base transceiver stations (BTS)


Cellular Mobile Telecom NetworksCellular Mobile Telecom Networks

Radio carrier frequencies must be synchronized preciselyin order to prevent cross-talk

Radio spectrum

Frequency


Characterizing Synchronisation Quality

Characterizing Synchronisation Quality


Definition of jitter : ITU-T Rec. G.810Definition of jitter : ITU-T Rec. G.810ÂThe short term variations of the significant

instants of a digital signal from their reference positions in time

ÂGreater than 10Hz in modulation frequency

ÂThe short term variations of the significantinstants of a digital signal from their reference positions in time


IdealIdeal

JitteredJittered

Sampling (reading) pointsSampling (reading) points


Definition of jitter : ITU-T Rec. G.810Definition of jitter : ITU-T Rec. G.810ÂThe short term variations of the significant

instants of a digital signal from their reference positions in time


ÂThe short term variations of the significantinstants of a digital signal from their reference positions in time


IdealIdeal

JitteredJittered

Sampling (reading) pointsSampling (reading) points


Definition of wander : ITU-T Rec. G.810Definition of wander : ITU-T Rec. G.810

·The long term variations of the significantinstants of a digital signal from their reference positions in time

·Less than 10Hz in modulation frequency

·The long term variations of the significantinstants of a digital signal from their reference positions in time

·Less than 10Hz in modulation frequency

IdealIdeal

WanderedWandered

Sampling pointsSampling pointsSampling pointsSampling points


Tref(t)

T(t)

t

x(t) = jitter + wander

Time Error or Phase-Time x(t)Time Error or Phase-Time x(t)


Fractional Frequency Deviation y(t)Fractional Frequency Deviation y(t)

y(t) =

where

ν(t) = actual frequency of the signal

νNOM = specified nominal frequency

y(t) =

where

ν(t) = actual frequency of the signal

νNOM = specified nominal frequency

ν(t) - νNOM

νNOM


Synchronisation Distribution:

General Principles


General Principles


Synchronisation between telecommunication systems : 1


»Master-slave intra-network synchronisation»Master-slave intra-network synchronisation

MasterMasterSystemSystemdatadata SlaveSlave

SystemSystem

clockclock

datadatadata +data + clockclock

transmissiontransmissionlinklink




»The slave system continually adjusts its own clock to the incoming signal»The incoming signal contains both the clock

and data information

»Therefore both the master and slavesystems have the same transmit andreceive rates»There are no slips

»The slave system continually adjusts its own clock to the incoming signal»The incoming signal contains both the clock

and data information

»Therefore both the master and slavesystems have the same transmit andreceive rates»There are no slips




» Inter-network synchronisation» Inter-network synchronisation

NetworkNetwork AASystemSystemdatadata NetworkNetwork BB

SystemSystem

atomic clockatomic clock

datadatadata data

atomic clockatomic clock

transmissiontransmissionlinklink




»Each system is synchronised by separate atomic clocks

»The atomic clocks have nearly the samefree running frequency»There is little difference between the transmit

and receive rates at both ends

»The slip rate is only one in every 72 days»Totally acceptable for inter-national and

intra-national traffic of every kind of services

»Each system is synchronised by separate atomic clocks

»The atomic clocks have nearly the samefree running frequency»There is little difference between the transmit

and receive rates at both ends

»The slip rate is only one in every 72 days»Totally acceptable for inter-national and

intra-national traffic of every kind of services


Physicalsynchronisation network : 1


mastermaster--slaveslavechainschains

PRCPRCPRC

SSU

SynchronisationSynchronisationSupplySupply UnitUnit

SDHSDHEquipment ClockEquipment Clock

PRC =PRC = PrimaryPrimaryReferenceReferenceClockClock

SEC

Telecom equipment clocksTelecom equipment clocks


Physical synchronisationnetwork : 2

Physical synchronisationnetwork : 2

»Not every system in the network can have a direct connection to the master network clock.

»Therefore the telecommunication systemsare synchronised in chains or trees»Each system clock is the master clock of the

subordinate system clocks slaved to it»The chains can be very long or very short

»Not every system in the network can have a direct connection to the master network clock.

»Therefore the telecommunication systemsare synchronised in chains or trees»Each system clock is the master clock of the

subordinate system clocks slaved to it»The chains can be very long or very short




PRC SDPRC SD trailtrail clock qualityclock qualitytraceable backtraceable back

to theto the PRCPRC

PRCPRCPRC

SSU

clock qualityclock qualitytraceable backtraceable back

to theto the SECSEC

SEC SDSEC SD trailtrail SSU SDSSU SD trailtrail

clock qualityclock qualitytraceable backtraceable back

to theto the SSUSSU

SEC


Synchronisation Distribution (SD) Trails

Synchronisation Distribution (SD) Trails

»The clock frequency along a SD trail is theSAME as the head-end, ie PRC, SSU or SEC»SD trails can be very long or very short»There can be hundreds of SD trails in a

synchronisation network

»The clock frequency along a SD trail is theSAME as the head-end, ie PRC, SSU or SEC»SD trails can be very long or very short»There can be hundreds of SD trails in a

synchronisation network


Causes of jitterCauses of jitter

»Phase-noise generated by clock recovery circuits in transmission network elements

»Phase-noise generated by low-quality equipment clocks

»Phase-noise generated by clock recovery circuits in transmission network elements

»Phase-noise generated by low-quality equipment clocks


Causes of wanderCauses of wander

½Temperature variations induce low-frequency phase-noise in network element clocks

½Temperature variations (e.g. between day and night) modify the propagation delays in transmission cables

½Temperature variations induce low-frequency phase-noise in network element clocks

½Temperature variations (e.g. between day and night) modify the propagation delays in transmission cables


Jitter and wander controlJitter and wander control

ÁJitter and wander must be kept below predefined limits called Network Limits

ÁTwo distinct techniques are used for the following jitter and wander components:ÁJitter and wander in the spectral domain above ≈

1 mHzÁWander in the spectral domain below ≈ 1 mHz

ÁJitter and wander must be kept below predefined limits called Network Limits

ÁTwo distinct techniques are used for the following jitter and wander components:ÁJitter and wander in the spectral domain above ≈

1 mHzÁWander in the spectral domain below ≈ 1 mHz


Jitter and wander filteringJitter and wander filtering

ÁRequired to prevent excessive accumulation of jitter and wander in the spectral domain above ≈ 1 mHz

ÁUse very narrow bandwidth (≈ 1 mHz) slaveclocks (SSUs) at intervals on the SD trails

ÁRequired to prevent excessive accumulation of jitter and wander in the spectral domain above ≈ 1 mHz

ÁUse very narrow bandwidth (≈ 1 mHz) slaveclocks (SSUs) at intervals on the SD trails


PLL transfer function (TF)PLL transfer function (TF)» TF = A plot of : Amplitude of output jitter (frequency)

Amplitude of input jitter (frequency)» TF = A plot of : Amplitude of output jitter (frequency)

Amplitude of input jitter (frequency)

--2020

--1010

00

1010

0.001f0.001f 0.01f0.01fffpeakpeak

ff 10f10f 100f100f

modulationmodulationfrequencyfrequency

Gain inGain in dBdB

--33

peakpeak

RollRoll--offoffFilter BandwidthFilter Bandwidth

0.1f0.1fffcutcut--offoff


Wander buffering on input portsWander buffering on input ports

»Wander in the spectral domain below the bandwidth of the SSUs cannot be attenuated

ÁThis low frequency wander accumulates across the synchronisation network

ÁBuffer stores on traffic input ports must be able to absorb at least 18 µs of wander

»Wander in the spectral domain below the bandwidth of the SSUs cannot be attenuated

ÁThis low frequency wander accumulates across the synchronisation network

ÁBuffer stores on traffic input ports must be able to absorb at least 18 µs of wander


Input wander greater than 18µsInput wander greater than 18µs

»The size of the buffer in telecommunications systems is usually just slightly larger than 18µs.

» If the input wander is greater than the size ofthe buffer, then the buffer over- or underflows, thus causing slips.

»To prevent slips the level of wander in the network must be kept below 18 µs.

»The size of the buffer in telecommunications systems is usually just slightly larger than 18µs.

» If the input wander is greater than the size ofthe buffer, then the buffer over- or underflows, thus causing slips.

»To prevent slips the level of wander in the network must be kept below 18 µs.



SDH-Based Solution


SDH-Based Solution


1. Elements1. Elements

Clocks and LinksClocks and Links


ClocksClocks

»Primary Reference Clock (PRC)

»Node Clock or Synchronisation Supply Unit (SSU)»If it is an independent piece of equipment, then

it is called a SASE (Stand-Alone Synchronisation Equipment)

»SDH Equipment Clock (SEC)

»Primary Reference Clock (PRC)

»Node Clock or Synchronisation Supply Unit (SSU)»If it is an independent piece of equipment, then

it is called a SASE (Stand-Alone Synchronisation Equipment)

»SDH Equipment Clock (SEC)


Primary Reference Clock (PRC)Primary Reference Clock (PRC)

»Master clock used to synchronise the entire network with a frequency accuracy of

< 1 x 10-11

»Based on atomic Cesium clocks

»Master clock used to synchronise the entire network with a frequency accuracy of

< 1 x 10-11

»Based on atomic Cesium clocks


PRC ImplementationPRC Implementation

»Autonomous equipment with one or several atomic Cesium clocks

»Radio-controlled clock synchronized to remote atomic Cesium clocks (e.g. Global Positioning System - GPS)

»A combination of the above

»Autonomous equipment with one or several atomic Cesium clocks

»Radio-controlled clock synchronized to remote atomic Cesium clocks (e.g. Global Positioning System - GPS)

»A combination of the above


Node Clock or SynchronisationSupply Unit (SSU)

Node Clock or SynchronisationSupply Unit (SSU)

Reference

Selector

Input

Interface

Input

Interface

Input

InterfaceOutput

Interface

Output

Interface

Output

Interface

Output

InterfaceJitter/Wander

Low-Pass

Filter

Holdover

Memory


Node Clock or Synchronisation Supply Unit (SSU)


»Selects an input reference signal based on»priority table, or»SSM signaling and priority table.

»Attenuates the jitter and wander present at the input via narrow-band (mHz) low-pass filtering.

» If all reference signals are lost, maintains the last phase & frequency as good as it can (holdover mode).

»Selects an input reference signal based on»priority table, or»SSM signaling and priority table.

»Attenuates the jitter and wander present at the input via narrow-band (mHz) low-pass filtering.

» If all reference signals are lost, maintains the last phase & frequency as good as it can (holdover mode).


• SSU Type• Application

Frequencyaccuracy

Holdover frequencydeparture Bandwith

Type I2048 kbit/s based N/A 5E-10 + t x 2E-10 / day 3 mHz

Type II1544 kbit/s based 1.6E-8 1E-10 + t x 1E-10 / day 1 mHz

Type III1544 kbit/s based 4.6E-6 1E-9 + t x 1E-9 / day 1 mHz




SDH network element ’s synchronisation functionSDH network element ’s synchronisation function

STM-Ninput external

timing output(2 MHz or1.5 Mbit/s, or 2 Mbit/s)

externaltiminginput(2 MHz or1.5 Mbit/s or 2Mbit/s)

SynchronousSynchronousEquipmentEquipment

TimingTimingGeneratorGenerator

SelectorSelector BB

NEinternaltiming

Synchronous Equipment Timing Source (SETS)

SelectorSelector CC

PDHinput

SelectorSelector AA

SDH Equipmemt Clock


SDH SEC featuresSDH SEC featuresÀ Input synchronisation signals are :»STM-N aggregates and tributaries»2 Mbit/s tributaries»2 MHz, 2 or 1.5 Mbit/s (non traffic) timing inputs

À Input selection is determined by :»a priority table, that is user definable»Synchronisation Status Message (SSM) on the

STM-N and 2 Mbit/s interfaces

ÀOutput synchronisation signals are :»All STM-N aggregates and tributaries»2 MHz, 2 or 1.5 Mbit/s (non traffic) timing

outputs

À Input synchronisation signals are :»STM-N aggregates and tributaries»2 Mbit/s tributaries»2 MHz, 2 or 1.5 Mbit/s (non traffic) timing inputs

À Input selection is determined by :»a priority table, that is user definable»Synchronisation Status Message (SSM) on the

STM-N and 2 Mbit/s interfaces

ÀOutput synchronisation signals are :»All STM-N aggregates and tributaries»2 MHz, 2 or 1.5 Mbit/s (non traffic) timing

outputs


Interworking between SDH NE and SASE

Interworking between SDH NE and SASE

SDH NESDH NE

cleanedcleanedtraffic traffic &&timingtiming

outputsoutputs

SASESASE

noisy noisy traffic traffic &&timingtiminginputinput

ExternalExternaltimingtiminginputinput

ExternalExternaltiming timing outputoutput


SEC PerformanceSEC Performance

• Type• Application

FrequencyAccuracy

Holdover frequencydeparture Bandwith

Option 12048 kbit/s based 4.6E-6 5E-8 + t x 1E-8 / day 1 .. 10 Hz

Option 21544 kbit/s based 20E-6 5E-8 + t x 5E-7 / day 0.1 Hz


SDH synchronisation link connectionsSDH synchronisation link connections

»Supported by an SDH multiplex section trail

» i.e. the timing information is carried by theSTM-N data rate ( N x 155 Mbits/s)

»SDH regenerator timing generators are not counted as elements of the synchronisation distribution layer, they belong to thesynchronisation link connection


» i.e. the timing information is carried by theSTM-N data rate ( N x 155 Mbits/s)

»SDH regenerator timing generators are not counted as elements of the synchronisation distribution layer, they belong to thesynchronisation link connection


2. Architecture2. Architecture


2.1. Inter-Node Distribution Architecture

2.1. Inter-Node Distribution Architecture


Master-slave principleMaster-slave principle

»A designated master clock is used as a reference frequency generator.

»The frequency generated by the master clock is disseminated to all other clocks which are slaved to the master clock.

»A designated master clock is used as a reference frequency generator.

»The frequency generated by the master clock is disseminated to all other clocks which are slaved to the master clock.


Master-slave principleMaster-slave principle

PRC

SEC

SSU

SEC

SSU

SSU SSU SSU

SEC

SEC

SEC

SEC

SEC

SEC

SEC

SEC

SEC

SEC

= master

= slave

= slave

= slave

= slave

= slave

= slave


Principle of trail redundancyPrinciple of trail redundancy

»Each slave clock should get at least two reference signals form the master clock via geographically separate trails.

»Sometimes it is not possible to fulfill this principle for all nodes of the network (depending on connectivity).

»Each slave clock should get at least two reference signals form the master clock via geographically separate trails.

»Sometimes it is not possible to fulfill this principle for all nodes of the network (depending on connectivity).


Hierachy of clock quality levelsHierachy of clock quality levels

»There is a hierarchy of clock quality levels:

»The higher the clock quality level, the higher the frequency accuracy of the clock

»Frequency accuracy: either overall free-run accuracy or holdover accuracy over a limited time period

»There is a hierarchy of clock quality levels:

»The higher the clock quality level, the higher the frequency accuracy of the clock

»Frequency accuracy: either overall free-run accuracy or holdover accuracy over a limited time period


Clock quality levelsClock quality levels

2048 kbit/s based:

PRC: 1E-11

SSU I: 2E-10/d

SEC 1: 4.6E-6

2048 kbit/s based:

PRC: 1E-11

SSU I: 2E-10/d

SEC 1: 4.6E-6

1544 kbit/s based:

PRC: 1E-11

SSU II: 1.6E-8/1 yr

SSU III/IV: 4.6E-6

SEC 2: 20E-6

1544 kbit/s based:

PRC: 1E-11

SSU II: 1.6E-8/1 yr

SSU III/IV: 4.6E-6

SEC 2: 20E-6


Hierarchical distribution ruleHierarchical distribution rule

»A clock of a given quality level must always (also under failure conditions) take timing (directly or indirectly) from a source clock with the same or higher quality level.

»A clock of a given quality level must always (also under failure conditions) take timing (directly or indirectly) from a source clock with the same or higher quality level.


Synchronisation network with SSMSynchronisation network with SSM

PRC

SEC

SSU

SEC

SSU

SSU SSU SSU

SEC

SEC

SEC

SEC

SEC

SEC

SEC

SEC

Link failure!

Holdover mode!

SEC

SEC


Synchronisation network with SSMSynchronisation network with SSM

»Link failure within a chain of SECs!

»SSM signaling prevents the downstream SSU form following a SEC in holdover mode.

» Instead, the first SSU enters holdover mode and becomes the source clock for the cut off sub-network.

»Link failure within a chain of SECs!

»SSM signaling prevents the downstream SSU form following a SEC in holdover mode.

» Instead, the first SSU enters holdover mode and becomes the source clock for the cut off sub-network.


The control of jitter and wanderThe control of jitter and wander

»SDH requires that jitter and wander be kept below tight network limits.

»This is achieved by inserting narrow-bandwith SSUs in the synchronisation chain (SEC bandiwth is relatively wide).

»Narrow-bandwith SSUs attenuate jitter and wander components that lie outside theSSU bandwith.


»This is achieved by inserting narrow-bandwith SSUs in the synchronisation chain (SEC bandiwth is relatively wide).

»Narrow-bandwith SSUs attenuate jitter and wander components that lie outside theSSU bandwith.


SDH Synchronisation reference chain

SDH Synchronisation reference chain

PRC

SSU

SEC

SEC

SSU

SEC

SEC

SEC

SEC

max. 20 SEC

max. 20 SEC

max. 20 SEC

max. 10 x

max. 60 SEC


SDH synchronisation reference chain

SDH synchronisation reference chain

»For SDH (not SONET!) see ITU-T G.803 or ETSI EN 300 462-2

»The ITU-T/ETSI synchronisation reference chain meets the network limits on jitter andwander:»Not more than 60 SECs in a chain»Not more than 20 SECs between two SSUs»Not more than 10 SSUs in the chain

»For SDH (not SONET!) see ITU-T G.803 or ETSI EN 300 462-2

»The ITU-T/ETSI synchronisation reference chain meets the network limits on jitter andwander:»Not more than 60 SECs in a chain»Not more than 20 SECs between two SSUs»Not more than 10 SSUs in the chain


Summary - 1Summary - 1

SDH/SONET Network

SDH/SONET NE

SDH/SONET NE

PSTN Switch

SSU/SASE

PRC

Ext. timing signal

Ext. timing signal

Ext. timing signal

Synchronisationcarried bySTM-N signals


2.2. Intra-Node Distribution Architecture

2.2. Intra-Node Distribution Architecture


Intra-node star topologyIntra-node star topology

Node boundary

Primary PRC SD trail

PRC SD trail to other nodes PRC SD trail to other nodes

SEC SEC

SSU

Secondary PRC SD trail

SEC SDH equipmentclock

Other equipmentclock


Interworking between SDH NE, SASE, and other equipment

Interworking between SDH NE, SASE, and other equipment

SDH NEcleanedtraffic &timing

outputs

SASE

noisy traffic &timinginput

External timinginput

External timing output

e.g. Switch


Functions of the SSM signaling layerFunctions of the SSM signaling layer

Signal the source clock quality level from clock to clock down the synchronisation chains, in order to

»enable clocks to select the best available reference timing signal

»enable clocks to go into holdover mode if reference timing signals are of low quality

»prevent timig loops in SDH chains and rings

Signal the source clock quality level from clock to clock down the synchronisation chains, in order to

»enable clocks to select the best available reference timing signal

»enable clocks to go into holdover mode if reference timing signals are of low quality

»prevent timig loops in SDH chains and rings


Synchronisation Status MessagesSynchronisation Status Messages

¼The clock source quality level is indicated by the Synchronisation Status Message (SSM):¼QL-PRC = Primary Reference Clock

¼QL-SSU-T = SSU Transit Node Clock

¼QL-SSU-L = SSU Local Node Clock

¼QL-SEC = SDH Equipment Clock

¼QL-DNU = Do not use

¼The clock source quality level is indicated by the Synchronisation Status Message (SSM):¼QL-PRC = Primary Reference Clock

¼QL-SSU-T = SSU Transit Node Clock

¼QL-SSU-L = SSU Local Node Clock

¼QL-SEC = SDH Equipment Clock

¼QL-DNU = Do not use


Transmission Channels for the SSMTransmission Channels for the SSM

¼ The timing quality level carried by STM-N signals is indicated by S1 byte in the STM-N Multiplex Section Over Head (MSOH)

¼ The timing quality level carried over 2048 kbit/s synchronisation signals is indicated in one of the bits Sa4 to Sa8 in Time Slot Zero (TS0).

¼ OC-N signals (SONET): Multiplex SectionOverhead

¼ 1544 kbit/s T1 signals: see ITU-T Rec. G.704

¼ The timing quality level carried by STM-N signals is indicated by S1 byte in the STM-N Multiplex Section Over Head (MSOH)

¼ The timing quality level carried over 2048 kbit/s synchronisation signals is indicated in one of the bits Sa4 to Sa8 in Time Slot Zero (TS0).

¼ OC-N signals (SONET): Multiplex SectionOverhead

¼ 1544 kbit/s T1 signals: see ITU-T Rec. G.704


SSM-based reference selectionSSM-based reference selection

¿ Always select the timing input with the highest QL, and if a number of equal QL timing inputs areavailable, then select the highest priority timing input

» In the locked mode, the output SSMs are set to the selected input SSM, e.g. QL-PRC in => QL-PRC out

À The SSM in the return signal of the selected inputis automatically set to DNU

¿ Always select the timing input with the highest QL, and if a number of equal QL timing inputs areavailable, then select the highest priority timing input

» In the locked mode, the output SSMs are set to the selected input SSM, e.g. QL-PRC in => QL-PRC out

À The SSM in the return signal of the selected inputis automatically set to DNU


If all the timing inputs are badIf all the timing inputs are bad

À The SEC enters hold-over mode:À it remembers the phase and frequency values

of the previously good input, but drifts towardsy = 4.6 x 10-6 (4.6ppm)

À STM-N outputs get SSM = QL-SEC

À 2 MHz ext. timing outputs are squelched (cut off)

À 2 Mbit/s ext. timing outputs get SSM = QL-SEC or AIS = true

À The SEC enters hold-over mode:À it remembers the phase and frequency values

of the previously good input, but drifts towardsy = 4.6 x 10-6 (4.6ppm)

À STM-N outputs get SSM = QL-SEC

À 2 MHz ext. timing outputs are squelched (cut off)

À 2 Mbit/s ext. timing outputs get SSM = QL-SEC or AIS = true


Ring synchronisation with SSM: normal condition

Ring synchronisation with SSM: normal condition

PrimaryPrimary PRCPRC

SSM =QLSSM =QL--PRCPRC

QLQL--PRCPRC

SSM = QLSSM = QL--PRCPRC

SSM = QLSSM = QL--DNUDNU

QLQL--DNUDNU


1111

1111

22

SecondarySecondary PRCPRC

22

SSM = QLSSM = QL--PRCPRCSSM = QLSSM = QL--PRCPRC

22

QLQL--PRCPRC QLQL--PRCPRC

22


Ring synchronisation with SSMprimary PRC section failure

Ring synchronisation with SSMprimary PRC section failure

PrimaryPrimary PRCPRC


QLQL--PRCPRC



QLQL--DNUDNU


1111

1111

22

SecondarySecondary PRCPRC

22SSM = QLSSM = QL--PRCPRC

22

QLQL--PRCPRC QLQL--PRCPRC

22

input 1 = LOSinput 1 = LOS detecteddetectedinput 2 = QLinput 2 = QL--PRC PRC detecteddetectedSECSEC selectsselects input 2input 2

SECSEC selectsselects input 3input 3



GPS-Based and MixedSolutions


GPS-Based and MixedSolutions


The Global Positioning System (GPS)The Global Positioning System (GPS)



» It ’s a global navigation satellite system owned by the US DoD.

»24 satellites broadcast time and their positions (orbits).

»Based on signals from four or more satellites, GPS receivers calculate their position and time.

»Synchronization signals can be derived from that time information.

» It ’s a global navigation satellite system owned by the US DoD.

»24 satellites broadcast time and their positions (orbits).

»Based on signals from four or more satellites, GPS receivers calculate their position and time.

»Synchronization signals can be derived from that time information.



»Orbit and time information that is broadcast on civilian satellite signals can be degraded during times of war (« Selected Availability - S/A»).

»Frequency accuracy of GPS receivers is typically better than 1·10-12 over 24 hours.

»Time error is typically better than 150 ns (when degraded by S/A).

»Orbit and time information that is broadcast on civilian satellite signals can be degraded during times of war (« Selected Availability - S/A»).

»Frequency accuracy of GPS receivers is typically better than 1·10-12 over 24 hours.

»Time error is typically better than 150 ns (when degraded by S/A).



»S/A degradation on timing can be (at least partially) removed by appropriate filteringtechniques.

»Good GPS receivers (good S/A filtering!)comply with the PRC specification of ITU-T Rec. G.811.

»GPS reception is subject to interferenceand can be jammed intentionally.

»S/A degradation on timing can be (at least partially) removed by appropriate filteringtechniques.

»Good GPS receivers (good S/A filtering!)comply with the PRC specification of ITU-T Rec. G.811.

»GPS reception is subject to interferenceand can be jammed intentionally.


Synchronisation distribution architecturesbased on the GPS

Synchronisation distribution architecturesbased on the GPS

1) Master-slave distribution tree plus GPS-receivers in important nodes

2) « Cellular synchronisation distribution »

1) Master-slave distribution tree plus GPS-receivers in important nodes

2) « Cellular synchronisation distribution »


Network example used for the comparisonNetwork example used for the comparison

= SDH node= Multiplex section


Master-slave tree plus GPS-receivers in important nodes


M = Master PRC

S = SSU/SASE

G =GPS-based PRCS

G

G GG

SM

SSS

S


»Conventionnel master-slave distribution tree with a central PRC and SSU/SASEs»GPS-receivers as back-up reference

sources in nodes equiped with SSU/SASEs

»Conventionnel master-slave distribution tree with a central PRC and SSU/SASEs»GPS-receivers as back-up reference

sources in nodes equiped with SSU/SASEs




»Redundancy is provided by the GPS»Therefore master-slave tree need not

necessariliy provide redundant synchronisation distribution, thus simplifying network design»Easier to modify (network evolution) than

with redundant master-slave tree

»Redundancy is provided by the GPS»Therefore master-slave tree need not

necessariliy provide redundant synchronisation distribution, thus simplifying network design»Easier to modify (network evolution) than

with redundant master-slave tree




»Lower risk of creating timing loops than with redundant master-slave tree»Equipment cost somewhat higher than

with redundant master-slave tree (additional GPS-receivers): that ’s the price to pay for all the advantages

»Lower risk of creating timing loops than with redundant master-slave tree»Equipment cost somewhat higher than

with redundant master-slave tree (additional GPS-receivers): that ’s the price to pay for all the advantages




« Cellular synchronisation distribution »« Cellular synchronisation distribution »

G = GPS-based PRC

GG

G G

G G


GPS GPS

1 1 1 1

2 2

2

2

« Cellular synchronisation distribution »:Cell pattern A

« Cellular synchronisation distribution »:Cell pattern A


GPS GPS

1 1 1 1

2 2

2

2

« Cellular synchronisation distribution »:Failure of a GPS-receiver

« Cellular synchronisation distribution »:Failure of a GPS-receiver


GPS GPS

1 1 1 1

2 2

2

2

« Cellular synchronisation distribution »:Link failure

« Cellular synchronisation distribution »:Link failure


1 1 1 1

2 2

2

2

« Cellular synchronisation distribution »:Cell pattern B

« Cellular synchronisation distribution »:Cell pattern B


»Each of the two synchronization inputs of a B-cell must be connected to a different synchronization output of an A-cell

»An A-cell ’s synchronization output may be connected to one or many synchronization inputs of B-cells

»Each of the two synchronization inputs of a B-cell must be connected to a different synchronization output of an A-cell

»An A-cell ’s synchronization output may be connected to one or many synchronization inputs of B-cells

« Cellular synchronisation distribution »:Interconnection rules

« Cellular synchronisation distribution »:Interconnection rules


« Cellular synchronisation distribution »« Cellular synchronisation distribution »

G = GPS-based PRC

GG

G G

G G

B BB

B BA

A

A


»All nodes get two reference signals coming from two GPS receivers located in different sites via geographically separate routes

»Protection against failures of links, clocks and GPS-receivers (including local corruption of the GPS radio signal)

»All nodes get two reference signals coming from two GPS receivers located in different sites via geographically separate routes

»Protection against failures of links, clocks and GPS-receivers (including local corruption of the GPS radio signal)

« Cellular synchronisation distribution »:Protection against failures

« Cellular synchronisation distribution »:Protection against failures


»Cells cannot be looped since A-cells only deliver synchronization, and B-cells only receive synchronization

»Cells are protected against timing loop formation by the SSM protocol

»Cells cannot be looped since A-cells only deliver synchronization, and B-cells only receive synchronization

»Cells are protected against timing loop formation by the SSM protocol

« Cellular synchronisation distribution »:Timing loop prevention

« Cellular synchronisation distribution »:Timing loop prevention


»Simple internal cell structure

»Simple interconnection rules

»Network can be expanded easily by»Adding new cells»Increasing cell length»Spitting a cell into several smaller cells

»Simple internal cell structure

»Simple interconnection rules

»Network can be expanded easily by»Adding new cells»Increasing cell length»Spitting a cell into several smaller cells

« Cellular synchronisation distribution »:Simple and scaleable network design

« Cellular synchronisation distribution »:Simple and scaleable network design


SummarySummary

»The GPS provides a synchronisation source compliant to ITU-T Rec. G.811

»Architectures must provide protection against failure of the GPS radio signal, e.g.:»Master-slave tree plus GPS-receivers in

important nodes»« Cellular synchronisation distribution »

»The GPS provides a synchronisation source compliant to ITU-T Rec. G.811

»Architectures must provide protection against failure of the GPS radio signal, e.g.:»Master-slave tree plus GPS-receivers in

important nodes»« Cellular synchronisation distribution »



From Co-operating Network


From Co-operating Network


Taking synchronisation from a co-operating network

Taking synchronisation from a co-operating network

»There is no PRC in the network»All clocks in the network are slaved to

synchronization signals from a co-operating network.»Under normal operating conditions all slave

clocks operate at the same frequency asthe PRC in the co-operating network.»There are normally no slip for on-net and

off-net traffic to the co-operating network.

»There is no PRC in the network»All clocks in the network are slaved to

synchronization signals from a co-operating network.»Under normal operating conditions all slave

clocks operate at the same frequency asthe PRC in the co-operating network.»There are normally no slip for on-net and

off-net traffic to the co-operating network.


»The clock signals from the co-operating network may be received at only a fewsynchronisation gateway nodes.»The clock signals from the co-operating

network may also be received at every node, or at every sub-network.

»The clock signals from the co-operating network may be received at only a fewsynchronisation gateway nodes.»The clock signals from the co-operating

network may also be received at every node, or at every sub-network.

Taking synchronisation from aco-operating network



PRCPRC

nn nn nn nn nn

nn nn nn nn

nn nn nn nn

COCO--OPERATING NETWORKOPERATING NETWORK




COCO--OPERATING NETWORKOPERATING NETWORK

PRCPRC

nn nn nn

nn nn nn

nn nn nn

nn nn

nn

nn




COCO--OPERATINGOPERATINGNETWORKNETWORK

PRCPRC

nn

nn

nn

nn

nn nn

nn

nn

nn

nn nn

nn nn




»The network’s synchronisation performance is dependent on the qualityof the synchronisation signals from the co-operating network.»There must be an agreement with the co-

operating operator on service level»The cost to lease the synchronisation

signals can be high.

»The network’s synchronisation performance is dependent on the qualityof the synchronisation signals from the co-operating network.»There must be an agreement with the co-

operating operator on service level»The cost to lease the synchronisation

signals can be high.

Critical issuesCritical issues


Agreement on synchronization interfaces

Agreement on synchronization interfaces

»Physical interface specification (e.g. 2 Mbit/s, G.703)»SSM configuration»Guaranteed synchronization quality (e.g.

G.823 Network Limit)»Upstream synchronisation chain length

(number of clocks)»Guaranteed availability of agreed quality

(e.g. 0.9999)

»Physical interface specification (e.g. 2 Mbit/s, G.703)»SSM configuration»Guaranteed synchronization quality (e.g.

G.823 Network Limit)»Upstream synchronisation chain length

(number of clocks)»Guaranteed availability of agreed quality

(e.g. 0.9999)


Agreement onsynchronization interfaces

Agreement onsynchronization interfaces

»Mean Time to Repair in case of failure»Worst case quality degradation in case of

failure (e.g. max. frequency error, max. frequency drift, max. jitter & wander)»Alarming method in case of failure (e.g.

SSM) »Quality monitoring criteria

»Mean Time to Repair in case of failure»Worst case quality degradation in case of

failure (e.g. max. frequency error, max. frequency drift, max. jitter & wander)»Alarming method in case of failure (e.g.

SSM) »Quality monitoring criteria


Summary onStandards

Summary onStandards


Standardisation BodiesStandardisation Bodies

International level : ITU Recommendations

Regional level, Europe: ETSI Legaly binding standards

USA: ANSI Legaly binding standards

Industry level: e.g. TIA Industry standards

Company level: e.g. Bellcore Internal standards

International level : ITU Recommendations

Regional level, Europe: ETSI Legaly binding standards

USA: ANSI Legaly binding standards

Industry level: e.g. TIA Industry standards

Company level: e.g. Bellcore Internal standards

ITU : International Telecommunication UnionETSI : European Telecommunications Standards InstituteANSI : American National Standards InstituteTIA: Telecommunication Industry Association


ITU-T Rec. concerning synch.ITU-T Rec. concerning synch.

» G.703 Electrical characteristics of digital interfaces

» G.781 Synchronisation layer function of SDH NEs

» G.783 SDH equipment functional blocks

» G.803 Architecture of SDH transport networks

» G.810 Definitions and terminology

» G.811 Specification for PRCs

» G.812 Specification for SSUs

» G.703 Electrical characteristics of digital interfaces

» G.781 Synchronisation layer function of SDH NEs

» G.783 SDH equipment functional blocks

» G.803 Architecture of SDH transport networks

» G.810 Definitions and terminology

» G.811 Specification for PRCs

» G.812 Specification for SSUs


ITU-T Rec. concerning synch.ITU-T Rec. concerning synch.

» G.813 Specification for SECs

» G.823 Jitter and wander control in 2048 kbit/s-based PDH networks (incl. synch. interfaces)


» G.825 Jitter and wander control in SDH networks

» G.8251 Jitter and wander control in the OTN

» G.813 Specification for SECs



» G.825 Jitter and wander control in SDH networks

» G.8251 Jitter and wander control in the OTN


Cross-Reference TableCross-Reference Table

Definitions Network Clocks

Architecture NetworkLimits

SSM PRC SSU/BITS SEC

ITU-T G.810 G.803 PDH:G.823G.824

G.704G.832

G.811 G.812 G.813G.781

SDH:G.825

ETSI ETS 300 462-1 ETS 300 462-2 sync.:ETS 300 462-3

ETS 417-6-1ETS 300 147

ETS 300 462-6 ETS 300 462-4 ETS 300 462-5ETS 417-6-1

non sync.:ETS 302 084

ANSI T1.101 T1.101 T1.101T1.105.03

T1.105.09

Telcordia GR-436-CORE(1)

GR-253-COREGR-499-COREGR-253-CORE

GR-253-CORE GR-2830-CORE GR-378-COREGR-1244-CORE

GR-253-CORE

Note (1): Synchronisation Planning


How to Synchronize Mixed Technology Networks

How to Synchronize Mixed Technology Networks


Core Network LayersCore Network Layers

SDH or SONET

OTN

IP64 kbit/s

E1 or DS1 ATM

ATMAdaption

layer

MPLS


Access Network LayersAccess Network Layers

IP

UMTS-FDD RAN ISDN xDSL Ethernet

Compressed Voice

64 kbit/s Voice

IP

Wireless access network Wired access network


Symbols and TerminolgySymbols and Terminolgy

PSTN switch

IP Router

IP Label Switching Router

ATM Switch

PSTN switch

IP Router

IP Label Switching Router

ATM SwitchATM

PSTN

IP LSR

IP



UMTS Base Station

UMTS Radio Network Controller

Mobile Switching Center

Mobile Packet Router

UMTS Base Station

UMTS Radio Network Controller

Mobile Switching Center

Mobile Packet RouterMPR

MSC

RNC

C

BS



SDH/SONET Add-Drop-Multiplexer

SDH/SONET Crossconnect

Wavelength Add-Drop-Multiplexer

Wavelength Crossconnect

SDH/SONET Add-Drop-Multiplexer

SDH/SONET Crossconnect

Wavelength Add-Drop-Multiplexer

Wavelength CrossconnectOTN

OTN

SDH

SDH


Core Network LayersCore Network Layers

S

P

S

P

SS

M

O

S

O

IP

A

O

IP

A

O

IP

A

ATM

SDH/SONET

OTN

IP/PSTN

O

O

S


Wired Access NetworkWired Access Network

PSTN

Modem

DSLAM

Ethernet LAN

xDSL

E1 or DS1

64 kbit/s Voice

PBX

Core Network

ATM

IP

ATM

IP


ATM

Wireless Access NetworkWireless Access Network

BS

BS

S S

MSC

MPR

RNC

C

S

PSTN

IP Network

SDH/SONET

N x E1/DS1

N x E1/DS1


2. SDH and SONET Networks2. SDH and SONET Networks


Synchronization requirementsSynchronization requirements

»Frequency synchronization

»Frequency accuracy:»Normal operation: 1E-11»Failure conditions: 4.6E-6 (SDH Eqmt. Clock)

2E-5 (SONET Minimum Clock)

»Jitter and Wander:»ITU-T G.825 (Network Limits)»ANSI T1.101, § 7.2 and 7.3 (MTIE, TDEV, phase

transients)


»Frequency accuracy:»Normal operation: 1E-11»Failure conditions: 4.6E-6 (SDH Eqmt. Clock)

2E-5 (SONET Minimum Clock)

»Jitter and Wander:»ITU-T G.825 (Network Limits)»ANSI T1.101, § 7.2 and 7.3 (MTIE, TDEV, phase

transients)


Synchronization distribution architecture


»Tree of slave clocks locked to a master clock(master-slave distribution)

» Independent synchronization islands synchronized by GPS-receivers or other local PRCs

»A combination of master-slave distribution and local PRCs in important nodes

»Tree of slave clocks locked to a master clock(master-slave distribution)

» Independent synchronization islands synchronized by GPS-receivers or other local PRCs

»A combination of master-slave distribution and local PRCs in important nodes


Synchronization issuesSynchronization issues

»Jitter and wander levels more critical than frequency accuracy (therefore need for low-bandwith slave clocks in master-slave networks).

»Establishment and maintenace of a network plan with trail redundancy

»Avoiding timing loops

»Jitter and wander levels more critical than frequency accuracy (therefore need for low-bandwith slave clocks in master-slave networks).

»Establishment and maintenace of a network plan with trail redundancy

»Avoiding timing loops


3. The Public Switched Telephone Network

3. The Public Switched Telephone Network




»Frequency accuracy:»Normal operation: 1E-11»Failure conditions: depends on clock level or

stratum

»Jitter and Wander:»ITU-T Rec. G.823 & G.824 (Network Limits)»ANSI T1.101, § 7.2 (MTIE, TDEV, phase

transients)


»Frequency accuracy:»Normal operation: 1E-11»Failure conditions: depends on clock level or

stratum

»Jitter and Wander:»ITU-T Rec. G.823 & G.824 (Network Limits)»ANSI T1.101, § 7.2 (MTIE, TDEV, phase

transients)




» Synchronization distribution based on PDH transport network (E1 or DS1 signals as synchronization carriers).

» Synchronization distribution based on SDH/SONET transport network (STM-n or OC-n signals as synchronization carriers).

» Synchronization distribution based on the GPS

» Mixed solutions

» Synchronization distribution based on PDH transport network (E1 or DS1 signals as synchronization carriers).

» Synchronization distribution based on SDH/SONET transport network (STM-n or OC-n signals as synchronization carriers).

» Synchronization distribution based on the GPS

» Mixed solutions



»Frequency accuracy more critical than jitter and wander levels (therefore need for clocks with good holdover stability).

»The SDH/SONET network is not transparent for timing of E1 and DS1 signals (wander induced by SDH/SONET pointer adjustments).

»Frequency accuracy more critical than jitter and wander levels (therefore need for clocks with good holdover stability).

»The SDH/SONET network is not transparent for timing of E1 and DS1 signals (wander induced by SDH/SONET pointer adjustments).


4. ATM Networks4. ATM Networks




»Frequency accuracy for synchronous physical layer: »Normal operation: 1E-11»Failure conditions: 2E-5

»Frequency accuracy for asynchronous physical layer: »Failure conditions: 2E-5

»Jitter & wander: ITU-T G.825 Network Limits


»Frequency accuracy for synchronous physical layer: »Normal operation: 1E-11»Failure conditions: 2E-5

»Frequency accuracy for asynchronous physical layer: »Failure conditions: 2E-5

»Jitter & wander: ITU-T G.825 Network Limits




»Take synchronization from SONET-based synchronization distribution network

»Master-slave distribution tree using ATM traffic signals as synchronization carriers

»Take synchronization from GPS-receivers

»Combinations of the above techniques

»Take synchronization from SONET-based synchronization distribution network

»Master-slave distribution tree using ATM traffic signals as synchronization carriers

»Take synchronization from GPS-receivers

»Combinations of the above techniques


Synchronization from the SONET networkSynchronization from the SONET network

PRSPRS

ATMATMNENE

ATMATMNENEATMATM

NENE

ATMATMNENE

ATMATMNENE

ClockClock

ClockClock

ClockClock

ClockClock

ClcokClcok

ClockClock

ATMATMNENE

SONET-basedsynch.distribution

ATMnetwork


Master-slave synchronization distributionMaster-slave synchronization distribution

PRSPRS

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMnetwork


GPS-receiversGPS-receivers

ATMnetwork

ATMATMNENE

GPSGPS

ATMATMNENE

GPSGPS

ATMATMNENE

GPSGPS

ATMATMNENE

GPSGPSATMATMNENE

GPSGPS


Mixed solutionMixed solution

PRSPRS

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMATMNENE

ATMATMNENE

ClockClock

ATMATMNENE

SONET-basedsynch.distribution

ATMnetwork

GPSGPS



»SDH/SONET-based synchronization not always available (e.g. in case of ATM overON)

»How can we achieve protection against reference failures in cases where SDH/SONET-based synchronization is not available?

»SDH/SONET-based synchronization not always available (e.g. in case of ATM overON)

»How can we achieve protection against reference failures in cases where SDH/SONET-based synchronization is not available?


5. Optical Networks5. Optical Networks


Today’s Optical Networks (ON)Today’s Optical Networks (ON)

»Transmission: optical fibre

»Multiplexing: (Dense) Wavelength Division Multiplexing (WDM, DWDM)

»Switching: circuit switching, where lightpaths are the circuits

»Control plane:»ASON (Automatically Switched Optical Network)

by ITU-T (G.8080/Y.1304)

»GMPLS (Generalized Multiprotocol Label Switching) by IETF (draft-ietf-mpls-generalized-signaling-07)

»Transmission: optical fibre

»Multiplexing: (Dense) Wavelength Division Multiplexing (WDM, DWDM)

»Switching: circuit switching, where lightpaths are the circuits

»Control plane:»ASON (Automatically Switched Optical Network)

by ITU-T (G.8080/Y.1304)

»GMPLS (Generalized Multiprotocol Label Switching) by IETF (draft-ietf-mpls-generalized-signaling-07)


Optical Transport NetworksOptical Transport Networks

The remaining slides are about today’s lightpath switched ONs as standardized by

ITU-T; they are called Optical Transport Networks (OTN)

(ITU-T Rec. G.709, G.871/Y.1301, G.872)

The remaining slides are about today’s lightpath switched ONs as standardized by

ITU-T; they are called Optical Transport Networks (OTN)

(ITU-T Rec. G.709, G.871/Y.1301, G.872)


General OTN FeaturesGeneral OTN Features

»New transport networking layer

»DWDM, each wavelength transports an Optical Channel (Och)

»Lightpath switching

»Data-rate per wavelength:»2.5 Gbit/s or 10 Gbit/s or 40 Gbit/s

»Service transparency for clients

»Asynchronous and bit-synchronous payload mappings

»New transport networking layer

»DWDM, each wavelength transports an Optical Channel (Och)

»Lightpath switching

»Data-rate per wavelength:»2.5 Gbit/s or 10 Gbit/s or 40 Gbit/s

»Service transparency for clients

»Asynchronous and bit-synchronous payload mappings


OTN Network LayersOTN Network Layers

+OPU/ODU/OTU Overhead

+OMS Overhead

OCh Overhead

Client Information1

+

+OTS/COMMS Overhead

Client InformationN

OC

h la

yer

net

wo

rkO

MS

laye

rn

etw

ork

OT

S la

yer

net

wo

rk

wavelengthN

wavelength1

wavelength0

opticalfibre

......

OCh1

OMU

OTM

OSC

OTU1

OChN

OTUN


OTN AbbreviationsOTN AbbreviationsOCh Optical Channel

OPU Optical Channel Payload Unit

ODU Optical Channel Data Unit

OTU Optical Channel Transport Unit

OMS Optical Multiplex Section

OMU Optical Multiplex Unit

OTS Optical Transmission Section

OTM Optical Transport Module

OCh Optical Channel

OPU Optical Channel Payload Unit

ODU Optical Channel Data Unit

OTU Optical Channel Transport Unit

OMS Optical Multiplex Section

OMU Optical Multiplex Unit

OTS Optical Transmission Section

OTM Optical Transport Module


OTN AbbreviationsOTN Abbreviations

OSC Optical Supervisory Channel

COMMS Management Communications

OADM Optical Add-Drop Multiplexer

OXC Optical Cross-Connect

LC Link Connection

REG Regenerator

OSC Optical Supervisory Channel

COMMS Management Communications

OADM Optical Add-Drop Multiplexer

OXC Optical Cross-Connect

LC Link Connection

REG Regenerator


OTN ArchitectureOTN Architecture

OADM OADMOXCREG REG

Client (e.g. SDH) Link Connection

OCh Link Connection (LC)

OTS LC

OCh Link Connection (LC)

OTS LCOTS LCOTS LC

OMS LC OMS LCOMS LCOMS LC

OChmatrix


Synchronisation IssuesSynchronisation Issues

»The OTN is required to limit jitter and wander accumulation.»Hence there are Network Limit specifications

(ITU-T Rec, G.8251).

»The OTN itself is not required to transportsynchronisation.»No requirement for different optical channels to

by synchronous.

»The OTN is required to limit jitter and wander accumulation.»Hence there are Network Limit specifications

(ITU-T Rec, G.8251).

»The OTN itself is not required to transportsynchronisation.»No requirement for different optical channels to

by synchronous.


Synchronisation IssuesSynchronisation Issues

»The OTN is required to allow the transport of synchronisation via SDH/SONET client connections.»SDH/SONET client connections must meet

G.825 Network Limits after transport over the OTN.»A synchronisation reference chain which meets

G.825 Network Limits is proposed in ITU-T Rec. G.8251.

»The OTN is required to allow the transport of synchronisation via SDH/SONET client connections.»SDH/SONET client connections must meet

G.825 Network Limits after transport over the OTN.»A synchronisation reference chain which meets

G.825 Network Limits is proposed in ITU-T Rec. G.8251.




»The SDH multiplex section trail may be supported by an optical channel of the OTN (Optical Transport Network)


»The SDH multiplex section trail may be supported by an optical channel of the OTN (Optical Transport Network)


SDHSDH

OTNOTN

SDH/SONET

OTN

PRCSynchronizationSSU


SDH multiplex section trail

Synch. link connection

Optical trail


The control of jitter and wanderThe control of jitter and wander


»This is achieved by inserting narrow-bandwith SSUs in the synchronisation chain.

»Narrow-bandwith SSUs attenuate jitter andwander components that lie outside theSSU bandwith.


»This is achieved by inserting narrow-bandwith SSUs in the synchronisation chain.

»Narrow-bandwith SSUs attenuate jitter andwander components that lie outside theSSU bandwith.


Synchronisation reference chain for SDH over OTN


OTNIsland

OTNIsland

PRC

SSU

SSU

SEC

SEC

typ. 10 OTN NEs

typ. 10 OTN NEs

max. 20 SEC

x n (n < 10) max. n x 10 OTN NEs




» ITU-T Rec. G.8251 for SDH over Optical Transport Network (OTN):

»Not more than 10 OTN islands in the chain»Insertion of an SSU after each OTN island»Not more than 10 OTN NEs per OTN island

(may be redivided freely over the OTN islands)»Not more than 20 SECs in the pure SDH tail

» ITU-T Rec. G.8251 for SDH over Optical Transport Network (OTN):

»Not more than 10 OTN islands in the chain»Insertion of an SSU after each OTN island»Not more than 10 OTN NEs per OTN island

(may be redivided freely over the OTN islands)»Not more than 20 SECs in the pure SDH tail




OTNIsland

PRC

SSU

SSU

SEC

SEC

max. 20 SECs

typ. 10 OTN NEs

max. 20 SEC

n x SDHm x OTNn + m < 10

max. (60 SECs+ m x 10 OTN NEs)

SEC

SEC




»There may be a mix of pure SDH islands and SDH-over-OTN islands.»Not more than 10 islands alltogether»Insertion of an SSU after each island»Not more than 20 SECs in a pure SDH island»Not more than 60 SECs in the entire chain

»There may be a mix of pure SDH islands and SDH-over-OTN islands.»Not more than 10 islands alltogether»Insertion of an SSU after each island»Not more than 20 SECs in a pure SDH island»Not more than 60 SECs in the entire chain


GSM Network ArchitectureGSM Network Architecture

MSC

BSC

MS

BTS

BTS

BTS

TransportNetwork

» MSC: Mobile Switching Center

» BSC: Base Station Controller

» BTS: Base Transceiver Station

» MS: Mobile Station (handy)


» BSC: Base Station Controller

» BTS: Base Transceiver Station

» MS: Mobile Station (handy)


UMTS Network ArchitectureUMTS Network Architecture

MSC

BS

UE

BS

BS

BS

TransportNetwork


» RNC: Radio Network Controller

» BS: Base Station

» UE: User Equipment (handy)


» RNC: Radio Network Controller

» BS: Base Station

» UE: User Equipment (handy)


Relevant GSM StandardsRelevant GSM Standards

»ETS 300 577: Digital cellular telecommunications systems (Phase 2+); Radio transmission and reception

»ETS 300 912: Digital cellular telecommunications systems (Phase 2+); Radio subsystem synchronization (GSM 05.10)

»ETS 300 577: Digital cellular telecommunications systems (Phase 2+); Radio transmission and reception

»ETS 300 912: Digital cellular telecommunications systems (Phase 2+); Radio subsystem synchronization (GSM 05.10)


Relevant UMTS-FDD StandardsRelevant UMTS-FDD Standards

»3G TS 25.401: …, UTRAN OverallDescription.

»3G TS 25.402: …, Synchronisation in UTRAN Stage 2.

»3G TS 25.305: ..., Stage 2 Functional Specification of Location Services in UTRAN.

»3G TS 25.401: …, UTRAN OverallDescription.

»3G TS 25.402: …, Synchronisation in UTRAN Stage 2.

»3G TS 25.305: ..., Stage 2 Functional Specification of Location Services in UTRAN.


GSM Synchronisation Requirements(at the radio interface)

GSM Synchronisation Requirements(at the radio interface)

» Max. frequency error: 0.05 ppm for observation intervals down to 0.577 ms

» Max. phase error over a 0.577 ms long « burst »:»20 degrees peak-to-peak»5 degrees RMS

» « Burst » timing error: 1 microsecond over 0.5 s observation interval

» Holdover capability to keep within 0.05 ppm max. freq. error over the repair time

» Max. frequency error: 0.05 ppm for observation intervals down to 0.577 ms

» Max. phase error over a 0.577 ms long « burst »:»20 degrees peak-to-peak»5 degrees RMS

» « Burst » timing error: 1 microsecond over 0.5 s observation interval

» Holdover capability to keep within 0.05 ppm max. freq. error over the repair time


UMTS Synchronisation Requirements(at the radio interface)

UMTS Synchronisation Requirements(at the radio interface)

» Max. frequency error: 0.05 ppm

» Holdover capability to keep within 0.05 ppm max.freq. error over the repair time

» Max. frequency error: 0.05 ppm

» Holdover capability to keep within 0.05 ppm max.freq. error over the repair time


Synchronisation Network Architectures

Synchronisation Network Architectures

»PDH-based solution

»SDH-based solution

»GPS-based solution

»PDH-based solution

»SDH-based solution

»GPS-based solution


PDH-Based SynchronisationPDH-Based Synchronisation

MSC BSC

MS

BTS

PDHMUX

PDHMUX

E1 E1

Switch

PRC

PDH transport network

PDHMUX

PDHMUX

E1 E1

PDHMUX

PDHMUX

E1 E1


SDH-Based SynchronisationSDH-Based Synchronisation

MSC BSC

MS

BTS

E1 E1

PRC

SDH transport network

SDH NE

E1 E1E1 E1

RET

SDH NESDH NE SDH NE

SYN

C

SYN

C

SYN

CRET = retiming


GPS-based SynchronisationGPS-based Synchronisation

MSC

BSC MS

BTS

BTS

BTS

TransportNetwork

GPS satellites


The problem with PDH over SDHThe problem with PDH over SDH

IMPORTANT:

»PDH path layers supported by SDH path layers are not suitable for transportingsynchronisation.

IMPORTANT:

»PDH path layers supported by SDH path layers are not suitable for transportingsynchronisation.


Wander from pointer adjustmentsWander from pointer adjustments¸ The payload is asynchronously mapped using bit

stuffing, but since its position is referenced to theSTM-N frame, each pointer adjustment causes a phase step to the PDH tributary desynchroniser

TRIBUTARYTRAFFIC SONET/SDH OUTPUTSIGNAL PAYLOAD WANDER

1.5Mbit/s VC-11, VT1.5 4.63µs2Mbit/s VC-12, VT2 3.47µs34,45Mbit/s VC-3, STS-1 0.16µs140Mbit/s VC-4 0.16µs

· Therefore unlike PDH, SONET and SDH are NOTsuitable to transport 1.5Mbit/s or 2Mbit/s SD trails

¸ The payload is asynchronously mapped using bitstuffing, but since its position is referenced to theSTM-N frame, each pointer adjustment causes a phase step to the PDH tributary desynchroniser

TRIBUTARYTRAFFIC SONET/SDH OUTPUTSIGNAL PAYLOAD WANDER

1.5Mbit/s VC-11, VT1.5 4.63µs2Mbit/s VC-12, VT2 3.47µs34,45Mbit/s VC-3, STS-1 0.16µs140Mbit/s VC-4 0.16µs

· Therefore unlike PDH, SONET and SDH are NOTsuitable to transport 1.5Mbit/s or 2Mbit/s SD trails


Typical 2Mbit/s output TIE plot for a VC-12 pointer adjustmentTypical 2Mbit/s output TIE plot for a VC-12 pointer adjustment

1E1E--66

2E2E--66

3E3E--66

4E4E--66

11 22 33 44 55 66 7700

00Time (s)Time (s)

3.1E3.1E--663.6E3.6E--66

dd = 1E= 1E--77dtdt

== FrequencyFrequency offsetoffset

TIE (s)TIE (s)


Retiming IRetiming I

ClockClockrecoveryrecovery

BufferBuffermemorymemorywritewrite

E1 or DS1 data inE1 or DS1 data in

EquipmentEquipmentclockclock

readread

RetimedRetimeddatadataoutout

Timing signalTiming signal


Retiming IIRetiming II»Retiming is applied on E1 or DS1 traffic signals

affected by excessive wander.

»The long-term frequency (data rate) of the traffic signal must locked to the network PRC.

»The retiming buffer transmits the incoming traffic at the data rate of the timing signal, thus removing the excessive wander.

»Retiming is used when E1/DS1 traffic signals transported over SDH/SONET are used as synchronisation links (e.g. GSM BST).

»Retiming is applied on E1 or DS1 traffic signals affected by excessive wander.

»The long-term frequency (data rate) of the traffic signal must locked to the network PRC.

»The retiming buffer transmits the incoming traffic at the data rate of the timing signal, thus removing the excessive wander.

»Retiming is used when E1/DS1 traffic signals transported over SDH/SONET are used as synchronisation links (e.g. GSM BST).


7. cdmaOne, cdma2000 and UMTS-TDD Radio Access

Networks

7. cdmaOne, cdma2000 and UMTS-TDD Radio Access

Networks


cdmaOne/2000 Synchronization Requirements

cdmaOne/2000 Synchronization Requirements

»Phase synchronization

»Frequency accuracy of base station transmit carrier: 5E-8

»Phase-time accuracy of base station time base:»Normal operation: 3 µs»Reference failure: 10 µs over 8 h failure period



»Phase-time accuracy of base station time base:»Normal operation: 3 µs»Reference failure: 10 µs over 8 h failure period


UMTS-TDD Synchronization Requirements

UMTS-TDD Synchronization Requirements



»Phase-time accuracy of base station time base in normal operation: 1.25 µs



»Phase-time accuracy of base station time base in normal operation: 1.25 µs




»A GPS-receiver in each base station

»A GPS-receiver per cluster of cells ormicrocells

»A GPS-receiver in each base station

»A GPS-receiver per cluster of cells ormicrocells


A GPS-receiver in each base stationA GPS-receiver in each base station

TransportNetwork

MSC

RNC

CBS

GPS

BS

GPS


A GPS-receiver per cluster of cells or microcellsA GPS-receiver per cluster of cells or microcells

TransportNetwork

MSC

BSC

C

GPS

BS

BS

Time transfer signal & protocol



»So far the GPS is the only practical source for phase synchronization - it may be accessed directly or indirectly (via other signals carrying GPS time or phase).

»Redundant GPS-receivers and GPS-antennas do not protect against degradation of the GPS radio signal due to interference or jamming.

»So far the GPS is the only practical source for phase synchronization - it may be accessed directly or indirectly (via other signals carrying GPS time or phase).

»Redundant GPS-receivers and GPS-antennas do not protect against degradation of the GPS radio signal due to interference or jamming.


Thank you

ascom synchronization

Documents

synchronisation

synchronisation

mobile switching

sdhsonet client

cellular synchronisation

mixed technology

short term

primary reference