2024年2月7日发(作者:)

Technical Specification

MEF 3

Circuit Emulation Service Definitions, Framework

and Requirements in Metro Ethernet Networks

April 13, 2004MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 1 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Disclaimer

The information in this publication is freely available for reproduction and use by any recipient and is believed to be

accurate as of its publication date. Such information is subject to change without notice and the Metro Ethernet

Forum (MEF) is not responsible for any errors. The MEF does not assume responsibility to update or correct any

information in this publication. No representation or warranty, expressed or implied, is made by the MEF

concerning the completeness, accuracy, or applicability of any information contained herein and no liability of any

kind shall be assumed by the MEF as a result of reliance upon such information.

The information contained herein is intended to be used without modification by the recipient or user of this

document. The MEF is not responsible or liable for any modifications to this document made by any other party.

The receipt or any use of this document or its contents does not in any way create, by implication or otherwise:

(a) any express or implied license or right to or under any patent, copyright, trademark or trade secret rights held or

claimed by any MEF member company which are or may be associated with the ideas, techniques, concepts or

expressions contained herein; nor

(b) any warranty or representation that any MEF member companies will announce any product(s) and/or service(s)

related thereto, or if such announcements are made, that such announced product(s) and/or service(s) embody

any or all of the ideas, technologies, or concepts contained herein; nor

(c) any form of relationship between any MEF member companies and the recipient or user of this document.

Implementation or use of specific Metro Ethernet standards or recommendations and MEF specifications will be

voluntary, and no company shall be obliged to implement them by virtue of participation in the Metro Ethernet

Forum. The MEF is a non-profit international organization accelerating industry cooperation on Metro Ethernet

technology. The MEF does not, expressly or otherwise, endorse or promote any specific products or services.

© The Metro Ethernet Forum 2004. All Rights Reserved.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 2 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Table of Contents

1.

2.

3.

4.

5.

6.

8

8

10

.10

MEF 10

CIRCUIT EMULATION 11

6.1 TDM

LINE

SERVICE

(T-LINE)............................................................................................................................11

6.1.1 Operational Modes of a .12

6.1.1.1 Unstructured 12

6.1.1.2 Structured 13

6.1.1.3 13

6.1.2 Bandwidth Provisioning for a 14

6.1.2.1 Bandwidth allocation at 100 kbit/14

6.1.2.2 14

6.1.2.3 15

6.2 TDM

ACCESS

LINE

SERVICE

(TALS)................................................................................................................15

6.2.1 Operational Modes of a 16

6.3

6.4

7.

CUSTOMER

OPERATED

17

MIXED-MODE

CES

.18

CIRCUIT EMULATION 19

7.1 GENERAL

19

7.1.1 Use of .19

7.2 SERVICE

INTERFACE

.19

7.2.1 Examples of TDM 20

7.2.2 TDM Service Processor (TSP)..................................................................................................................20

7.2.3 Circuit Emulation Inter-working Function (CES IWF)...........................................................................21

7.2.3.1 PDH Circuit 22

7.2.3.2 SONET/SDH Circuit .22

7.2.4 Emulated Circuit De/Multiplexing Function (ECDX).............................................................................23

7.2.5 Ethernet Flow Termination Function (EFT)............................................................................................23

7.2.6 .23

7.23

7.3.1 CES Interworking Function- 25

7.3.2 Synchronous IWF and 27

7.3.2.1 Synchronous IWF 27

7.3.2.2 Synchronous IWF and .27

7.3.3 Asynchronous IWF and 27

7.3.3.1 Asynchronous IWF, 28

7.3.3.2 Asynchronous IWF, 28

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 3 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.3.4 28

7.3.4.1 Single Service-Provider 29

7.3.4.2 Multi Service-Provider 31

7.3.4.3 Private (customer owned) 34

7.4 PERFORMANCE

MONITORING AND

35

7.4.1 Facility 35

7.4.35

7.4.2.1 35

7.4.2.2 36

7.4.2.3 Buffer Underflow .36

7.4.2.4 Alarms in 36

7.4.3 .36

7.5 SERVICE

36

7.5.1 Errors within the MEN causing TDM 37

7.5.1.1 37

7.5.1.2 Frame Delay and 37

7.5.1.3 37

7.5.1.4 Frame Error Ratio and 37

7.5.2 Relationship to TDM service 38

7.5.3 Errored Seconds Requirement for 38

7.5.4 Severely Errored Seconds Requirement for 39

7.5.5 Service Impairments for SONET/41

7.5.5.1 Performance Objectives for the SONET/41

7.5.5.2 MEN 42

7.5.5.3 Summary & .45

7.5.6 45

7.6 TDM

46

7.46

7.7.1 Provider 46

7.7.2 Customer 47

7.48

7.8.1 Scenario 1 – dual .48

7.8.2 Scenario 2 – dual 48

7.8.3 Scenario 3 – Single 49

7.8.4 Scenario 4 – Single to dual 49

7.9

7.10

7.11

8.

SERVICE

50

50

50

CIRCUIT EMULATION 51

8.1 STRUCTURED

DS1/E1

NX64 KBIT/S

.51

8.1.1 51

8.1.2 51

8.1.3 51

8.1.52

8.1.5 Jitter .52

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 4 of 65

8.1.6

8.1.7

8.1.8

8.1.9

8.1.10

8.1.11

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Facility 52

Bit 52

52

.53

Lost and Out of 53

Buffer 53

8.2 DS1/E1

UNSTRUCTURED

53

8.2.54

8.2.54

8.2.3 Jitter .54

8.2.4 Facility 54

8.2.54

8.2.6 Lost and Out of 55

8.2.7 Buffer 55

8.3 UNSTRUCTURED

DS3/E3

55

8.3.55

8.3.56

8.3.3 Jitter .56

8.3.56

8.3.5 Lost and Out of 56

8.3.6 Buffer 57

8.4 SONET/SDH

57

8.4.1 Framing and 57

8.4.58

8.4.3 59

8.4.4 Jitter .60

8.4.60

8.4.6 Buffer 60

8.5 GENERAL

61

8.5.61

8.5.2 Frame Loss 61

8.5.61

9. 62

.62

ETHERNET

FRAME

62

ETHERNET

FRAME

63

NETWORK

63

BANDWIDTH

63

9.1

9.2

9.3

9.4

9.5

10. 64

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 5 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

List of Figures

Figure 1: TDM Line Service over Metro 11

Figure 2: Possible TDM Virtual Private 12

Figure 3: Example Multi-point to Point T-Line 14

Figure 4: Multiplexing across a single Ethernet 15

Figure 5: Handoff of a multiplexed trunk to an 16

Figure 6: Possible TDM 17

Figure 7: Customer Operated CES over a Metropolitan Ethernet Service (e.g. E-Line)..............................................17

Figure 8: 18

Figure 9: Functional Elements and 19

Figure 10: Functional Elements of SONET/SDH 22

Figure 11: Interworking 23

Figure 12: Timing distribution architecture showing 24

Figure 13: CES IWF Synchronization 26

Figure 14: Synchronization Administration for a Single 30

Figure 15: Synchronization Administration for a Multi 31

Figure 16: Synchronization Administration for a 34

Figure 17: Allowed Frame Error Ratio to meet 39

Figure 18: Relationship between Packing and FER for 0.01% 40

Figure 19: Relationship between Packing and FER for 0.01% SES based on 10 BER on end to end TDM data

41

Figure 20: Provider Controlled 47

Figure 21: Customer Controlled 47

Figure 22: Dual .48

Figure 23: Dual 49

Figure 24: Single 49

Figure 25: Single to dual .49

-3MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 6 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

List of Tables

Table 1: TDM .20

Table 2: CES TDM 22

Table 3: Timing 24

Table 4: CE Synchronization modes and Expected 25

Table 5: CES IWF 26

Table 6: Timing Options per .28

Table 7: CES Services and supporting Synchronization .29

Table 8: Timing Options Single-Service Provider .30

Table 9: Timing Options Multi-Service Provider 32

Table 10: Timing Options for a 34

Table 11: Example conversion from ES to FER for Ethernet .38

Table 12: Sparse FER objectives for MEN [G.826].........................................................................................................43

Table 13: Background FER and BER objectives for MEN [G.826]................................................................................43

Table 14: Maximal FER accounted by SESR [G.826].....................................................................................................44

Table 15: MEN FER Requirements for SDH/.45

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 7 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

1. Abstract

The emulation of TDM circuits over a Metro Ethernet Network (known as Circuit Emulation Services, or CES) is a

useful technique to allow MEN service providers to offer TDM services to customers. This document outlines the

types of the TDM services that can be offered over a MEN, and the requirements of such services. It covers both

PDH services (e.g. Nx64 kbit/s, T1, E1, T3, E3) and SONET/SDH services (STS-1, STS-3, STS-3c, STS-12, STS-12c and European equivalents).

The document contains three sections:

-

-

-

a set of service definitions for Circuit Emulation Services (CES) in the Metro Ethernet Forum context,

a framework section explaining issues with the implementation of such services

a set of requirements needed for providing such services in Metro Ethernet Networks (MEN).

Separate implementation agreements for PDH and SONET/SDH services will be produced which cover the

implementation of these services in an inter-operable manner. This document is intended as a reference against

which any Implementation Agreement can be compared to ensure that all the service requirements have been met.

2. Terminology

Term

AIS

ANSI

APS

BER

BBER

BLSR

CAS

CCS

CE

CES

CR

CRC

CUET

DSTM

DTM

E-Line

ECDX

ECID

EFT

ES

ESR

MEF 3

Definition

Alarm Indication Signal

American National Standards Institute

Automatic Protection Switching

Bit Error Ratio

Background Block Error Ratio

Bi-directional Line Switched Ring

Channel Associated Signaling

Common Channel Signaling

Customer Equipment

Circuit Emulation Services

Clock Recovery.

Cyclic Redundancy Check

Customer Unit Edge Terminal

Derived System Timing Mode.

The IWF system clock is derived from the incoming Ethernet frames.

Derived Timing Mode.

The recovered service clock is derived from the incoming Ethernet frames.

Ethernet Line Service

Emulated Circuit De/Multiplexing Function

Emulated Circuit Identifier

Ethernet Flow Termination

Errored Second

Errored Second Ratio

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 8 of 65

Term

ESF

EVC

FDL

FER

HOVC

IA

IETF

ITU-T

IWF

LOH

LOPS

LOS

LOVC

LTE

MIB

MEN

NE

NNI

PDH

PE

PLL

PRS

PSTN

PWE3

RDI

SDH

SES

SESR

SF

SLA

SLS

SONET

SSM

TALS

TDM

T-Line

TSP

Definition

Extended Super Frame

Ethernet Virtual Circuit

Facility Data Link

Frame Error Ratio

Higher Order Virtual Container

Implementation Agreement

Internet Engineering Task Force

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

International Telecommunication Union – Telecommunication standardization sector

Inter-Working Function

Line OverHead

Loss Of Packet Synchronization

Loss Of Signal

Lower Order Virtual Container

Line Termination Equipment

Management Information Base

Metro Ethernet Networks

Network Equipment

Network to Network Interface

Plesiochronous Digital Hierarchy.

PDH refers to the DS1/DS2/DS3 and E1/E3/E4 family of signals

Provider Edge

Phase Locked Loop

Primary Reference Source

Public Switched Telephone Network

Pseudo-Wire Emulation Edge to Edge (an IETF working group)

Remote Defect Indication

Synchronous Digital Hierarchy

Severely Errored Second

Severely Errored Second Ratio

Super Frame

Service Level Agreement

Service Level Specification

Synchronous Optical Network

Synchronization Status Message

TDM Access Line Service

Time Division Multiplexing

(examples of TDM services include Nx64 kbit/s, T1, T3, E1, E3, OC-3, STM-1, OC-12, STM-4)

TDM Line Service

TDM Service Processing

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

MEF 3

Page 9 of 65

Term

UNI

VC

VT

Mapper

Mapping

Definition

User Network Interface

Virtual Container

Virtual Tributary

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

A device or logic that implements a mapping function

The process of associating each bit transmitted by the service into the SDH/SONET payload

structure that carries the service. For example, mapping a DS1 service into SONET VT-1.5 /

SDH VC-11 associate each bit of the DS1 with a location in the VT-1.5/VC-11

Transmit one or more channels over a single channel

A part of the SDH/SONET overhead that locates a floating payload structure. Frequency

differences between SDH/SONET network elements or between the payload and the

SDH/SONET equipment can be accommodated by adjusting the pointer value.

Multiplex

Pointer

3. Scope

The scope of this document is to address the particular requirements of transport over MEN for edge-to-edge-emulation of circuits carrying time division multiplexed (TDM) digital signals. It makes references to requirements

and specifications produced by other standards organizations (notably the ITU, ANSI, IETF PWE3 and ATM

Forum), and adapts these to address the specific needs of MEN. It is not in the scope of this document to duplicate

any work of other related standardization bodies.

The document covers the definition of such services, a framework in which the service provision can be understood,

and also the requirements for the defined services. The actual implementation of these services in an inter-operable

manner will be the subject of separate Implementation Agreements, and hence is outside the scope of this document.

4. Compliance Levels

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD

NOT”, “RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in

[RFC 2119]. All key words must be use upper case, bold text.

5. MEF Document Roadmap

MEF draft specifications and the document roadmap are available in the members area on the Metro Ethernet Forum

website at /.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 10 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

6. Circuit Emulation Service Definition

Ethernet CES provides emulation of TDM services, such as Nx64 kbit/s, T1, E1, T3, E3, OC-3 and OC-12, across a

Metropolitan Ethernet Network. The object of this is to allow MEN service providers to offer TDM services to

customers. Hence it allows MEN service providers to extend their reach and addressable customer base. For

example, the use of CES enables metro Ethernet transport networks to connect to PBX’s on customer premises and

deliver TDM voice traffic along side of data traffic on Ethernet.

The CES is based on a point-to-point connection between two Inter Working Functions (IWF). Essentially, CES

uses the MEN as an intermediate network (or ‘virtual wire’) between two TDM networks. This is handled as an

application of the Ethernet service, using the interworking function to interface the applications layer onto the

Ethernet services layer.

6.1 TDM

LINE

SERVICE

(T-LINE)

The TDM Line (T-Line) service provides TDM interfaces to customers (Nx64 kbit/s, T1, E1, T3, E3, OC-3, OC-12

etc.), but transfers the data across the Metropolitan Ethernet Network (MEN) instead of a traditional circuit switched

TDM network. From the customer perspective, this TDM service is the same as any other TDM service, and the

service definition is given by the relevant ITU-T and ANSI standards pertaining to that service.

From the provider’s perspective, two CES interworking functions are provided to interface the TDM service to the

Ethernet network. The CES interworking functions are connected via the Metro Ethernet Network (MEN) using

point-to-point Ethernet Virtual Connections (EVCs) as illustrated in Figure 1. (For the purposes of CES, the service

provider owning the CES interworking functions is viewed as the “customer” of the MEN, providing the ability to

use an Ethernet UNI, and the performance and service attributes associated with this concept).

The TDM Service Processor (TSP) block shown in Figure 1 consists of any TDM grooming function that may be

required to convert the TDM service offered to the customer into a form that the CES IWF can accept (see section

7.2.2). For example, the TSP may be a Framer device, converting a fractional DS1 service offered to the customer

into a Nx64 kbit/s service for transport over the MEN. The operation of the TSP is outside the scope of this

document.

The TSP and the CES IWF may physically reside in the Provider Edge (PE) unit at the provider’s nearest point-of-presence, or in a service provider owned box in a customer location (e.g. a multi-tenant unit). From the architecture

perspective, there is no difference between these alternatives.

TDMCECES(optional)IWFEFTTSPService Provider NetworkCESTSP(optional)IWFEFTTDMCETDM LinkTDM LinkMENEVCUNI-CUNI-NUNI-NUNI-CETH Access LinkETH Access LinkTDM SubscriberEthernet UNIDemarcationEthernet UNITDM SubscriberDemarcation

Figure 1: TDM Line Service over Metro Ethernet Networks

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 11 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

6.1.1 Operational Modes of a T-Line Service

The basic T-Line service is a point to point, constant bit rate service, similar to the traditional leased line type of

TDM service. However, service multiplexing may occur ahead of the CES inter-working functions, (e.g.

aggregation of multiple emulated T1 lines into a single T3 or OC-3 link), creating a multi-point to point or even a

multi-point to multi-point configuration (see Figure 2 below).

This service multiplexing is carried out using standard TDM multiplexing techniques, and is considered as part of

the TSP block, rather than the CES inter-working function. The TDM interface at the input of the CES inter-working function is the same as that output from the CES IWF at the opposite end of the emulated link. It is the

TSP that may be used to multiplex (or demultiplex) that TDM service into the actual TDM service provided to the

customer. This allows a TDM service to a customer to be provided as a collection of emulated services at lower

rates.

Point-to-point EVCsTDMCETSP /CES IWFT3TSP /CES IWFT3TDMCETSP /CES IWFT1TDMCEMENTSP /CES IWFT1TSP /CES IWFOC3TDMCETDMCE

Figure 2: Possible TDM Virtual Private Line Configurations

Hence there are three possible modes of operation of a T-Line service:

i.

ii.

iii.

Unstructured Emulation mode (also known as “structure-agnostic” emulation)

Structured Emulation Mode (also known as “structure-aware” emulation)

Multiplexing mode

Modes (i) and (ii) are point-to-point connections. Mode (iii) permits multi-point-to-point and multi-point-to-multi-point configurations, although all of these modes are operated over simple point-to-point EVCs in the MEN.

6.1.1.1 Unstructured Emulation mode

In Unstructured Emulation Mode, a service is provided between two service-end-points that use the same interface

type. Traffic entering the MEN on one end-point leaves the network at the other end-point and vice-versa.

The MEN must maintain the bit integrity, timing and other client-payload specific characteristics of the transported

traffic without causing any degradation that would exceed the requirements for the given service as defined in the

Requirements section of this document. All the management, monitoring and other functions related to that specific

flow must be performed without changing or altering the service payload information or capacity.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 12 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Examples where unstructured emulation mode could be implemented are leased line services or any other transfer-delay sensitive (real time) applications. The specific transport rates and interface characteristics of this service are

defined in the Requirements section of this document.

6.1.1.2 Structured Emulation mode

In Structured Emulation Mode, a service that is provided between two service-end-points that use the same interface

type. Traffic entering the MEN on one end-point is handled as overhead and payload. The overhead is terminated

at the near end point, while the payload traffic is transported transparently to the other end. At the far end point the

payload is mapped into a new overhead of the same type as at the near end point.

The MEN must maintain the bit integrity, timing information and other client-payload specific characteristics of the

transported traffic without causing any degradation that would exceed the requirements for the given service as

defined in the Requirements section of this document. All the management, monitoring and other functions related

to that specific flow must be performed without changing or altering the service payload information or capacity.

An example of such a service is the transport of OC-3 when the SOH is terminated at both ends and the STS-1

payloads are transported transparently over the MEN. A second example is a fractional DS1 service, where the

framing bit and unused channels are stripped, and the used channels transported across the MEN as an Nx64 kbit/s

service. The specific transport rates and interface characteristics are defined in the Requirements section of this

document.

6.1.1.3 Multiplexing mode

In the multiplexing mode, multiple lower rate transparent services are multiplexed at a specific service-end-point of

the MEN into a higher digital hierarchy. Similarly, a higher rate service may be de-composed into several lower rate

services. For example, a customer may have several sites – a head office with a full DS1 connection, and several

satellites with fractional DS1 connections, as shown in Figure 3. The same architecture can be used for multiplexing

of other rate services, e.g. several full DS1 services onto a single DS3, or multiplexing of VT-1.5s into an STS-1.

The specific transport rates and interface characteristics are defined in the Requirements section of this document.

Service multiplexing is typically performed in the TDM domain as part of a TSP, not in the Ethernet domain. A

fractional TDM link going into the MEN comes out as a fractional TDM link, or at least as a payload containing

solely that fractional link. Multiplexing and de-multiplexing is performed outside the CES IWF, as part of any

native TDM signal processing, as shown in Figure 3. Therefore the customer service is multiplexed, but the

emulated service (i.e. the service handled by the IWF) is structured.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 13 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Fractional DS1 payloadto be multiplexedTDMmultiplexingfunction(TSP)Head OfficeTDMCEFractional DS1 linkTDMCESatellitesiteCES IWFCES IWFMENFull DS1 linkCES IWFPoint-to-point EVCsFractional DS1 linksTDMCETDMCESatellite sites

Figure 3: Example Multi-point to Point T-Line Multiplexed service

6.1.2 Bandwidth Provisioning for a T-Line Service

The Metro Ethernet service provider will need to allocate sufficient bandwidth within the network to carry the T-Line service. This may require very fine granularity of bandwidth provision to allow efficient allocation. For

example, an N x 64 kbit/s service may have very low bandwidth requirements where N is small. This could result in

very inefficient provisioning if the smallest unit increment of bandwidth provision is 1 Mbit/s. The following

sections outline three possible schemes to provision the bandwidth efficiently for low data rate CES.

6.1.2.1 Bandwidth allocation at 100 kbit/s granularity

The bandwidth required to be able to offer emulation of Nx64 kbit/s services increases with N in steps of 64kbit/s,

plus an overhead for encapsulation headers. Therefore, in order to be able to allocate MEN bandwidth efficiently

for such services, the bandwidth needs to be provisioned in similar step sizes. It is recommended that to achieve

reasonable efficiency, the granularity of service provisioning should be at 100 kbit/s or smaller.

However, most existing equipment only allows for bandwidth provision at multiples of 1 Mbit/s (or 1.5 Mbit/s for

much SONET-based Ethernet equipment), so this fine level of granularity is not guaranteed to be available.

Therefore alternative means of allocating bandwidth may need to be used to provide a reasonable level of efficiency,

such as described in sections 6.1.2.2 and 6.1.2.3 below.

6.1.2.2 TDM multiplexing

Multiple TDM services between the same two IWFs may be multiplexed together in the TDM domain before being

carried across the MEN. De-multiplexing at the far end is also accomplished in the TDM domain. For example,

several fractional DS1 services could be multiplexed onto a single full DS1 link before transmission across the MEN.

Bandwidth can then be provisioned for the full link, rather than individually for each fractional customer service.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 14 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Similarly, the same architecture can be used for multiplexing of other rate service, e.g. several full DS1 services

onto a single DS3 or SONET link.

6.1.2.3 Ethernet multiplexing

Another alternative is to multiplex several circuits across a single Ethernet Virtual Connection, as shown in Figure 4.

TDM links(may be full or fractional)TDMCETDMCETDMCETDMCETDMCECES IWFCES IWFTDMCEMENCircuit emulated linksCustomer sites(may or may not belong tothe same customer)TDMCETDMCECES IWFPoint-to-point EVCsTDMCETDMCE

Figure 4: Multiplexing across a single Ethernet Virtual Connection

All TDM circuit emulation between any two given IWFs is handled across a single point-to-point EVC. Individual

circuit emulated links are carried across that EVC using Ethernet multiplexing.. Customer separation is maintained

using this multiplexing.

The total bandwidth of circuit emulation traffic between two points is known in the case of constant bit rate, always

on service. Therefore the amount of traffic across that EVC is known and can be provisioned accordingly. This

allows efficient bandwidth provisioning. For example, 5 fractional TDM services at 384 kbit/s could share a single

2 Mbit/s EVC, and don’t have to be allocated 1 Mbit/s each.

6.2 TDM

ACCESS

LINE

SERVICE

(TALS)

The second type of TDM service that can be offered over Metropolitan Ethernet Networks is where the MEN

service provider hands off one (or both) ends of the network to a second network (e.g. the PSTN). For example,

Figure 5 shows the case where the customer interface is TDM, and the hand-off to an external network (in this case

the PSTN) is via some form of TDM or SONET trunk line (e.g. OC-3). The prime use of this type of service is

where the MEN is used as an access network onto the second network. With TALS service the customer-facing

IWF can be located on the customer’s site but is still under the control of the MEN operator.

Multiple customer services may be multiplexed up onto a single trunk to be handed off to the second network

(Figure 6). As with T-Line service, this multiplexing is implemented using conventional TDM techniques in a

native signal processing block. In some instances, both ends of the emulated TDM service may be within the

service provider network, e.g. backhaul of ATM or SONET services.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 15 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

As with T-Line service, the customer-facing TSP and CES IWF may physically reside in the Provider Edge (PE)

unit at the provider’s nearest point-of-presence, or in a service provider owned box in a customer location (e.g. a

multi-tenant unit). From the architecture perspective, there is no difference between these alternatives.

TDMCECES(optional)IWFEFTTSPService Provider NetworkCESIWFEFTTSPMUXPSTNTDM LinkTDM Trunk LinkMENEVCUNI-CUNI-NUNI-NUNI-CETH Access LinkETH Access Link

TDM SubscriberEthernet UNIDemarcationEthernet UNINetworkDemarcationFigure 5: Handoff of a multiplexed trunk to an external network

6.2.1 Operational Modes of a TALS Service

The TALS service is essentially very similar to the multiplexed, multi-point-to-point T-Line service. The two

services use the MEN in the same manner (see Figure 6). The only difference is that the final multiplexed service

(e.g. OC-3) is handed off to another network rather than an end customer. As such, it may have some performance

requirements deriving from the requirements of the second network not present in the T-Line service.

The MEN must maintain the bit integrity, timing and other client-payload specific characteristics of the transported

traffic without causing any degradation that would exceed the requirements for the given service as defined in the

Requirements section of this document. All the management, monitoring and other functions related to that specific

flow must be performed without changing or altering the service payload information or capacity.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 16 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Point-to-point EVCsTDMCETSP /CES IWFT3TSP /CES IWFTDMCETDMCETSP /CES IWFT1MENT3TDMCETSP /CES IWFT1TSP /CES IWFOC3Multiple subscribershanded off toexternal networkPSTN

Figure 6: Possible TDM Handoff Configurations

6.3 CUSTOMER

OPERATED

CES

It is also possible for customers to operate the Circuit Emulation Service themselves across an E-Line service (see

Figure 7). However, in order to operate CES at a reasonable level of quality, the service provider will need to offer

a “premium SLA”, with tighter definitions of parameters such as packet delay, variation in packet delay, and packet

loss. The requirements section of this document covers the parameters that need to be controlled, and the

appropriate range of values over which CES can be operated with acceptable quality.

Some customers may choose to operate CES across a standard E-Line service with no special SLA for cost reasons.

In this case, the level of quality experienced is entirely the customer’s own responsibility. Since from the service

provider’s perspective this is purely an Ethernet service, the definition of this service is considered to be outside

the scope of this document, other than to document possible parameters of a service level specification for CES-capable E-Line service.

TDMCECES(optional)IWFEFTTSPService Provider NetworkCESTSPIWF(optional)EFTTDMCETDM LinkTDM LinkMENEVCUNI-CUNI-NUNI-NUNI-CCUETETH Access LinkETH Access LinkCUET

Ethernet UNICustomer DemarcationEthernet UNICustomer DemarcationFigure 7: Customer Operated CES over a Metropolitan Ethernet Service (e.g. E-Line)

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 17 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

6.4 MIXED-MODE

CES

OPERATION

A further scenario is a “mixed mode service”, where the customer interface is Ethernet at one side, and TDM at the

other. Therefore the customer is providing their own interworking function at one end of the service, and this must

inter-operate with the service provider’s interworking function at the far end (Figure 8).

This service may be operated using the same methods as the other services described in this document. However, it

may create significant issues as far as troubleshooting links between the service provider and the customer. For

example it may be difficult to determine whether a fault is in the customer’s own equipment, or in the service

provider’s equipment. Service providers offering mixed-mode services should be aware of the potential issues

involved.

Resolution of these types of issues, as well as any operations and management issues involved with running between

IWFs in different administrative domains are for further study. The MEF do not plan any further work on the

definition of such a service at present.

TDMCECES(optional)IWFEFTTSPService Provider NetworkCESTSPIWF(optional)EFTTDMCETDM LinkTDM LinkMENEVCUNI-CUNI-NUNI-NUNI-CCUETETH Access LinkETH Access LinkCUETTDM SubscriberDemarcation

Ethernet UNICustomer DemarcationFigure 8: Mixed-Mode Service

Ethernet UNIMEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 18 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7. Circuit Emulation Service Framework

This section is intended to provide a framework within which the actual requirements for the Circuit Emulation

Service can be understood.

7.1 GENERAL

PRINCIPLES

In most service provider networks, SONET/SDH has been used as the technology for transporting customers’ PDH

or SONET/SDH traffic. The purpose of the CES is to allow the MEN to serve not only as a transport of Ethernet

and data services, but also as a transport of customer’s TDM traffic.

The CES solution in the MEN should therefore make the MEN behave as a standard SONET/SDH and/or PDH

network, as seen from the customer’s perspective. The intention is that the CES customer should be able to use the

same (legacy) equipment, regardless of whether the traffic is carried by a standard SONET/SDH or PDH network, or

by a MEN using CES.

7.1.1 Use of External Standards

Section 3.2 of the “Architectural Principles of the Internet” [RFC 1958] states “If a previous design, in the Internet

context or elsewhere, has successfully solved the same problem, choose the same solution unless there is a good

technical reason not to.”

Much of the material in the next two sections has been adapted from the ATM specification for circuit emulation

[ATM CES], since this covers broadly similar requirements. It also references the various applicable standards from

the ITU-T or ANSI for TDM services.

7.2 SERVICE

INTERFACE

TYPES

There are two basic service interfaces in the TDM domain. These are shown in Figure 9, and are defined as follows:

1) TDM Service Interface: The TDM service that is handed off to the customer or TDM network operator. TDM

services fall into two categories, unstructured and structured.

2) Circuit Emulation Service Interface (CES TDM Interface): The actual circuit service that is emulated between

interworking functions through the MEN.

For unstructured TDM service, the CES Interface carries all information provided by the TDM Service Interface

transparently. The service is emulated in its entirety by the IWF, including any framing structure or overhead

present.

For structured TDM service, the TDM service interface is operated on by the TSP (TDM Service Processor) to

produce the service that is to be emulated across the MEN. A single structured TDM service may be decomposed

into one or many CES flows, or two or more structured TDM services may also be combined to create a single CES

flow.

TDM ServiceInterfaceCustomerTDMServiceTSP(optional)(e.g. framing,mux/demux)CES TDMInterfaceCES IWF TypeTDMIWFPacketizationCES PayloadCES IWFAdaptedPayloadECDXEFTAdds DA, SA,and CRC fieldsEthernetInterfaceIWFAdds ECIDand Length/Type field

Figure 9: Functional Elements and Interface Types

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 19 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.2.1 Examples of TDM Service Interfaces

Table 1 shows examples of possible TDM service interfaces. These service interfaces are defined as standardized

PDH, SONET or SDH interfaces.

Unstructured TDM services are either emulated as is by the CES IWF, or mapped by the TSP onto the CES interface

without modifying their content. These services preserve the framing structure as well as SONET/SDH overhead as

input to the TDM service interface. These services provide point-to-point connectivity across a MEN.

Structured services may be either demultiplexed into any of their defined service granularities or multiplexed with

other structured TDM services. For example, an OC-3 structured service can be demultiplexed into three separate

STS-1 signals. Each of these signals can then be circuit emulated by different IWF across the MEN to one or more

service end-points. At these IWFs, these STS-1 can be reassembled with other STS-1s (from different originating

services) to form an entirely new OC-3 signal (provided each originating site uses a PRS level clock). Structured

service can be used to provide point-to-point, point-to-multipoint, and multipoint-to-multipoint TDM services.

TDM Service

Interface

DS1

DS3

E1

E3

OC-1

OC-3

OC-3c

STM-1

STM-1c

OC-12

OC-12c

STM-4

STM-4c

Unstructured

TDM Service

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Structured TDM

Service

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Yes

Structured TDM Service Granularity

Nx64 kbit/s

DS1, Nx64 kbit/s

Nx64 kbit/s

E1, Nx64 kbit/s, DS0

STS-1, VT-1.5, VT-2

STS-1, VT-1.5, VT-2

STS-3c

VC-11 (DS1), VC-12 (E1), VC-3 (DS3, E3, other)

VC-4, VC-3, VC-11, VC-12

VT-1.5 (DS1), VT-2 (E1),

STS-1 (DS3, E3, other), STS-3c

STS-12c

VC-11 (DS1), VC-12 (E1),

VC-3 (DS3, E3, other), VC-4

VC-4-4c

Table 1: TDM Service Interfaces

7.2.2 TDM Service Processor (TSP)

The TDM Service Processor is defined as a function, operating in the TDM domain, that:

a. modifies the bit rate or contents of the service to be emulated (e.g. overhead termination, removal of unused

channels in a fractional service)

b. multiplexes two or more TDM services into a single service to be emulated, and de-multiplexes the composite

service into its component services at the remote end as required (e.g. multiplexing of several Nx64kbit/s

services onto a larger Nx64kbit/s, or multiplexing of several DS1 services onto a DS3)

c. transparently maps the TDM service onto a different service to be emulated (e.g. asynchronous mapping of a

DS1 onto a VT-1.5)

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 20 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

It may make use of standard or proprietary techniques, and is not considered part of the inter-working function.

Therefore the requirements in this document and the related implementation agreements do not cover the operation

of any TSP function, which is considered part of an equipment vendors own value added domain.

By this definition therefore, a T1/E1 Framer would be considered a TSP (terminates the framing bits, breaks the

service down into an Nx64kbit/s service), whereas an LIU would not (bit rate and data to be emulated are the same

as in the TDM service offered to the customer). Another example of a TSP would be a DS1 to VT-1.5 mapper.

7.2.3 Circuit Emulation Inter-working Function (CES IWF)

Circuit Emulation Service is defined as an application service in terms of layered network model defined in the

MEN Architecture framework document. It uses the MEN as an intermediate network (or “virtual wire”) between

two TDM networks. The Circuit Emulation Interworking Function is therefore defined as the adaptation function

that interfaces the CES application to the Ethernet layer.

The CES IWF is responsible for all the functions required for an emulated service to function, e.g. data packetization

and de-packetization (including any padding required to meet the minimum Ethernet frame size), sequencing,

synchronization, TDM signaling, alarms and performance monitoring. Listed in Table 2 are the defined interface

types required to support circuit emulation over the MEN. These services are divided by payload type, CES

Interface type, CES rate, and IWF type. These interface definitions are illustrated in Figure 9.

Payload Type

PDH

CES TDM Interface Type

NDS0

DS1

E1

E3

DS3

SONET/SDH STS-1

STS-3

STS-3c

STM-1

STS-12

STS-12c

STM-4

SONET/SDH Tributary VT-1.5

VT-2

VC-11

VC-12

VC-3

STS-1P

STS-3P

VC-4

STS-3cP

STS-12P

STS-12cP

CES Rate

N x 64 kbit/s

1.544 Mbit/s

2048 kbit/s

34368 kbit/s

44.736 Mbit/s

51.84 Mbit/s

155 Mbit/s

155 Mbit/s

155 Mbit/s

622.08 Mbit/s

622.08 Mbit/s

622.08 Mbit/s

1.728 Mbit/s

2.176 Mbits

1.728 Mbit/s

2176 kbit/s

48384 kbit/s

50.112 Mbit/s

150.336 Mbit/s

150.336 Mbit/s

150.336 Mbit/s

601.344 Mbit/s

601.344 Mbit/s

IWF Type

NDS0-CE

DS1-CE

E1-CE

E3-CE

DS3-CE

STS-1-CE

STS-3-CE

STS-3c-CE

STM-1-CE

STS-12-CE

STS-12c-CE

STM-4-CE

VT-1.5-CE

VT-2-CE

VC-11-CE

VC-12-CE

VC-3-CE

STS-1P-CE

STS-3P-CE

VC-4-CE

STS-3cP-CE

STS-12P-CE

STS-12cP-CE

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 21 of 65

Payload Type CES TDM Interface Type

VC-4-4

VC-4-4cP

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

CES Rate

601.344Mbit/s

601.344Mbit/s

IWF Type

VC-4-4P-CE

VC-4-4cP-CE

Table 2: CES TDM Interface Definition

7.2.3.1 PDH Circuit Emulation Service

PDH emulation services provide transport of PDH services through the MEN. Managed by the IWF, as defined in

Table 2, these services can be used to support structured or unstructured TDM service interfaces. As structured

TDM services, the input TDM service can be demultiplexed into its service granularity as defined in Table 1. This

service granularity includes DS0, Nx64 kbit/s, DS1, E1, DS3 or E3. Likewise, multiple structured TDM services

can be multiplexed into a single service and circuit emulated across the MEN as this new service. For example, a

multiple Nx64 kbit/s services can be multiplexed into a DS1 which is circuit emulated across the MEN and output as

a DS1 structured service.. These interface rates and IWFs are presented in Table 2.

7.2.3.2 SONET/SDH Circuit Emulation Service

SONET/SDH emulation services provide transport of SONET/SDH services through the MEN. Managed by the

IWF, as defined in Table 2, these services can be used to support structured or unstructured TDM service interfaces.

As structured TDM services, the input TDM service can be demultiplexed into its service granularity as defined in

Table 1. This service granularity includes STS-1, STM-1 or the appropriate virtual container. Likewise, multiple

structured TDM services can be multiplexed into a single service and circuit emulated across the MEN as this new

service. These interface rates and IWFs are presented in Table 2.

SONET/SDH emulation service provides emulation of the following services:

1. Higher order Virtual Container (HOVC): VC-3/STS-1, VC-4/STS-3c, VC-4-4c/STS-12c

2. Lower order Virtual Container (LOVC): VC-11/VT-1.5, VC-12/VT-2, none/VT-3, VC-2/VT-6, VC-3/none

SONET terminology refers to HOVC as Synchronous Payload Envelope (SPE) and for LOVC as Virtual Tributary

(VT).

Figure 10 shows an example of the functional blocks used in SONET/SDH structured emulation service. For

SONET/SDH interface, a SONET/SDH framer TSP is used to terminate and handle the section and line overheads.

The SONET/SDH Virtual containers (SPEs) are extracted and passed to the SONET/SDH IWF(s). Each virtual

container (SPE) can be sent to a different destination IWF. The ECDX and EFT are used to add the required

multiplexing and Ethernet headers and trailers as described in Figure 9. For PDH interfaces, a SONET/SDH mapper

TSP is used to map the PDH signal (T1/E1/T3/E3) into SONET/SDH virtual containers (SPE/VT). Each virtual

container can be sent to a different destination IWF, and either groomed to a SONET/SDH interface or mapped back

to a PDH interface.

TSPMapper To CEPDHIWFIWFIWFTSPFramer To MENECDXEFTSDH

Figure 10: Functional Elements of SONET/SDH emulation service

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 22 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.2.4 Emulated Circuit De/Multiplexing Function (ECDX)

The Emulated Circuit De-multiplexer (ECDX) is a function, operating in the packet domain, that:

a. Selects one (or more) IWF as the final destination of each Ethernet frame received from the MEN based on an

Emulated Circuit Identifier (ECID) attribute within the frame header.

b. Prepends an ECID attribute for each Ethernet frame sent to the MEN to allow circuit de-multiplexing on the

MEN egress.

c. Assigns the length/type field to each Ethernet frame sent to the MEN.

By this definition, equipment supporting 8 DS1 TDM service interfaces would be able to multiplex 8 DS1 interface

onto a single EVC. Each emulated circuit would be identified by a unique ECID attribute. The ECDX function

would examine the ECID attribute of received Ethernet frames and compare it to its local ECID to IWF association.

According to the look-up result the ECDX would pass the circuit emulation payload to the relevant IWF for

processing.

7.2.5 Ethernet Flow Termination Function (EFT)

A termination function is defined as: “A transport processing function which accepts adapted characteristic

information from a client layer network at its input, adds information to allow the associated trail to be monitored

(if supported) and presents the characteristic information of the layer network at its output(s) (and vice versa)”

In the context used here, an Ethernet Flow Termination function takes an adapted payload from the ECDX (the

MAC client information field), along with a Length/Type attribute describing it as CES payload. It adds the MAC

Destination and Source addresses, and finally the frame check sequence.

In the CE-bound direction, the EFT takes in an Ethernet frame from the MEN. It determines whether it contains CES

payload from the Length/Type field, and forwards it to its associated ECDX function, for passing to the appropriate

CES IWF.

7.2.6 Direction terminology

For each direction of an emulated circuit, there is a pair of CES interworking functions. The MEN-bound IWF

handles the packetization of the TDM data, encapsulation into Ethernet frames and forwarding into the Ethernet

network. The corresponding CE-bound IWF extracts the TDM data from the Ethernet frames, and recreates the

TDM service. Each direction of the service is handled separately by second pair of IWFs. Hence for a given MEN-bound IWF, the corresponding CE-bound IWF is at the other side of the MEN.

CES TDMInterfaceCES EthernetInterfaceCES EthernetInterfaceCES TDMInterfaceMEN-boundIWF 1CE-boundIWF 1MENCE-boundIWF 2MEN-boundIWF 2MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 23 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

MENTSPCE1CES

IWFPHYME1MEnPHYCES

IWFTSPCE2SDCE1SDTDM1SDIWF1SDED0SDED1SDEDmSDEDnSDIWF2SDTDM2SDCE2Circuit Emulation ServiceTDM Emulation Service

Figure 12: Timing distribution architecture showing clock domains

The following notations are used in Figure 12:

Attribute

CES IWF

CEn

TSP

Description

Circuit emulation service interworking function.

Customer edge devices terminating/originating TDM circuits

The TDM Service Processor performs all TDM functions necessary to support structured or

unstructured TDM emulation service. These functions are performed in the TDM domain using

standardized techniques.

Metro Ethernet devices providing Circuit Emulation Service transport in the MEN

Physical interface terminating Ethernet traffic

The TDM-end service circuits

The CES IWF providing edge-to-edge-emulation for the TDM circuit

The synchronization domain used by Ethernet network elements (NEs) in the MEN cloud. This

timing information is used to transport packets between Ethernet NEs providing the CES. Since each

of these devices operates independently, the synchronization mode of each device must be

provisioned separately. Typical timing modes include line, external and free-run.

The synchronization domain used by CES IWF. This timing is used to transport the circuit emulated

service across the MEN. This synchronization domain is specific to the CES interface types as shown

in Table 2 (CES Interface Definition). Details of this synchronization domain are contained in the IA

reference as shown in Table 2 (CES Interface Definition)..

The synchronization domain used by the TSP. This timing is used to transport TDM signals between

the TSP and CE. This synchronization domain supports through timing mode.

The synchronization domain used by the CE devices to establish the TDM service clock. This timing

is used to transport TDM signals between the CE and the TSP. This synchronization domain supports

external, line, or free-run timing modes.

Table 3: Timing Distribution Definitions

One of the objectives of edge-to-edge-emulation of a TDM circuit is to preserve the service clock with a perfomance

level as specified in the relevant Implementation Agreements. For example, the performance objective could be to

MEF 3

MEn

PHY

SDEdn

SDIWFn

SDTDMn

SDCen

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 24 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

meet the G.823 Traffic Interface specification. The service clock can be either generated at a CE via external timing

mode, or recovered from the TDM bit stream via line/loop timing mode. It should be noted that loop timing mode is

a recovery process where timing is extracted from only one available input bit stream. Line timing mode is a

recovery process where timing is extracted from one of several input bit streams. Since the number of available

timing inputs is based on network design and system architecture, this document will use “line timing” when

describing timing modes where either line or loop timing are used.

Typical timing modes for a point-to-point CES connection are shown in Table 4. These timing modes are defined in

[T1.101] and [G.812].

Timing CE1 Timing Mode CE2 Timing Mode Comments

Option

1 External to PRS

traceable source

External to PRS

traceable source

Line timing mode

External to PRS

traceable source

Line timing mode

CE1 and CE2 will operate in a Plesiochronous mode. If

connected to other TDM circuits (in other synchronization

domains) network slip-rate objectives will be met.

CE2 will have the same average frequency as CE1. If

connection to other TDM circuits (in other synchronization

domains) network slip-rate objectives will be met.

CE1 will have the same average frequency as CE2. If

connection to other TDM circuits (in other synchronization

domains) network slip-rate objectives will be met.

CE2 will have the same average frequency as CE1. If

connection to other TDM circuits (in other synchronization

domains) undesirable slip performance may result.

CE1 will have the same average frequency as CE2. If

connection to other TDM circuits (in other synchronization

domains) undesirable slip performance may result.

CE1 and CE2 synchronization domains will operate

independently. TDM data will experience slips between

CE1 and CE2 and TDM circuits in other synchronization

domains.

2

3 External to PRS

traceable source

Line timing mode 4 Free-run mode

5 Line timing mode Free-run mode

6 Free-run mode Free-run mode

Table 4: CE Synchronization modes and Expected Timing performance

When the TDM circuit is transported via CES, this continuous signal is broken into packets at the MEN-bound IWF

of the CES connection and reassembled into a continuous signal at the CE-bound IWF of the CES connection. In

essence, the continuous frequency of the TDM service clock is disrupted when the signal is mapped into packets.

In order to recover the service clock frequency at the egress of the CES connection, the interworking function must

employ a process that is specific to the CES interface type. The description and requirements of the IWF service

clock recovery are contained in the Implementation Agreement for that service.

7.3.1 CES Interworking Function- Synchronization Description

The synchronization block diagram for the CES Interworking function is shown in Figure 13.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 25 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

External TimingReferenceCES IWFETFDataCLKTo To MENR

Figure 13: CES IWF Synchronization Reference Model

The following notations are used in Figure 13:

Attribute

IWF

Processor

TDM Line

Ethernet

Line

EXT

FR

SDIWF

ETF

Description

Performs all data, addressing and timing functions necessary to support circuit emulation over the

MEN.

A line timing option may be used to provide the IWF with a timing reference. TDM Line timing will

extract timing from either the TDM service SDTDM.

Ethernet Line timing will extract timing from either the MEN SDED or from the arriving Ethernet

frames.

An External timing option may be used to provide the IWF with a timing reference

An internal timing reference may be used to provide the IWF with a timing reference. This internal

timing reference may either be a free-running oscillator or an oscillator in holdover mode.

The synchronization domain for the IWF. The definition and requirements for the SDIWF

may be

found in the IA for the specific IWF per table 2 (CES Interface Definition)

Physical interface terminating Ethernet traffic

Table 5: CES IWF Synchronization Definitions

The main goal of the CES IWF is to preserve the service clock of the TDM service through the MEN. The CES

IWF must preserve the service clock to the accuracy defined in section 7 of this document even under the levels of

MEN jitter defined in section 9.2.

The operation and definition of the interworking functions are specific to the CES interface type as shown in Table 2

(CES Interface Definition Table). Specific information about operation and requirements for specific interworking

functions is contained in the implementation agreements as indicated in Table2 (CES Interface Definition Table).

The IWF can use a variety of timing inputs to use as a reference for service clock recovery. A list of the five

available timing inputs used by the CES IWF are shown below:

1. Line timing (from the CE) – This mode is used to recover timing from a CE. In order for this timing mode to

PRS traceable, the CE must be provisioned to recover and transmit PRS traceable timing. Line timing should

be used if synchronization messaging is available (e.g., SONET or SDH line timing sources).

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 26 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

2. Line timing (from the MEN) – This mode is used to recover timing from the MEN. In order for this timing

mode to be PRS traceable, the adjacent Ethernet NE (in the MEN) must be provisioned to recover and transmit

PRS traceable timing. Line timing should be used if synchronization messaging is available (e.g., SONET or

SDH line timing sources). Line timing can also include deriving a clock from incoming Ethernet packets (e.g.

use of adaptive or differential clock recovery methods).

3. External Timing – This mode is used to recover PRS traceable timing from a co-located building integrated

timing supply (BITS). Timing is typically sent to the TSP/IWF as an all ones DS1 or E1 with framing.

Synchronization status messaging (SSMs) may also be transmitted over the DS1 link using a “blue signal” (all

ones without framing) or via SSMs on the ESF data link as defined by ANSI T1.101.

4. Free-Run – This mode is considered a standalone mode. It should only be used when a suitable line or external

timing reference is not available. The frequency and stability of this timing mode is determined by the

TSP/IWF’s internal oscillator.

5. Holdover – This mode is usually considered a backup or protection timing mode. Holdover mode may be

initiated when the external or line timing references have been lost due to a failure. This failure is indicated to

the CES IWF via a loss of frame (LOF), loss of signal (LOS) or SSM indication. Unlike free-run, this timing

mode relies on the TSP/IWF’s internal oscillator that has been trained by an external timing reference (e.g., PRS

traceable timing source). See ANSI T1.101 for performance specifications of this and other timing modes.

Holdover may also be used when there is a cessation of activity on the MEN-bound TDM interface (e.g. due to

the normal operation of a variable bit rate IWF).

7.3.2 Synchronous IWF and Associated Tributaries

Synchronous IWFs require that the IWF synchronization domain (SDIWF) at both the MEN-bound IWF and the CE-bound IWF use common clock synchronization (e.g. timing that is traceable to the same physical clock).

Unless specific mechanisms have been put in place to ensure that a common clock is distributed between both IWFs,

it should be assumed that the IWFs are NOT synchronous. It is not reasonable to assume the IWFs have suitable

clocking information unless it has been specifically provided for.

Tributaries to the IWFs consist of those TDM services that originate at the CE and are terminated at the IWF. The

service clock of each tributary may be either synchronous (PRS traceable) or asynchronous (not PRS traceable).

7.3.2.1 Synchronous IWF and Tributaries

Synchronous IWF: In this scenario, the synchronization domains at the MEN-bound and CE-bound IWFs (SDIWF)

use common clock synchronization (e.g. timing that is traceable to the same physical clock).

Synchronous Tributary: The synchronization domains of the CE devices (SDCen) require that at least one of the CE

devices use external timing that is PRS traceable. The other CE, in the CES connection, may be either line-timed to

the originating CE or externally-timed to a PRS traceable source.

7.3.2.2 Synchronous IWF and Asynchronous Tributaries

Synchronous IWF: In this scenario, the synchronization domains at the MEN-bound and CE-bound IWFs (SDIWF)

use common clock synchronization (e.g. timing that is traceable to the same physical clock).

Asynchronous Tributary: The synchronization domains of the CE devices (SDCen) require that at least one of the

CE devices be externally-timed to a non-PRS traceable source or use its internal free-running oscillator. The other

CE can be either line-timed, be externally-timed, or operate on its internal free-running oscillator. Of these options,

only the line-timed option will provide slip-free performance under non-fault conditions (see Table 4).

7.3.3 Asynchronous IWF and Associated Tributaries

In the Asynchronous network, the synchronization domains of all IWFs (SDIWF) do not use a common source of

timing. For example, the IWFs may receive plesiochronous timing as defined in [T1.101]. Tributaries to the IWF

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 27 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

consist of TDM services that originate at the CE and are terminated at the IWF. At the TDM Service Processor,

the service clock of each TDM tributary (SDCen) is terminated. This service clock will be used to support any

functions necessary to create structured or unstructured TDM emulation services. After these services are created,

the service clock along with the TDM data are then input to the CES IWF.

7.3.3.1 Asynchronous IWF, Asynchronous Tributaries

Asynchronous IWF: In this scenario, not all of the synchronization domains of the IWF devices (SDIWF) use

common clock synchronization (e.g. timing from the same physical clock). Timing distribution for the IWF may be

configured with each either line-timed to the MEN, line timing to a dedicated CE, or externally-timed to a PRS

traceable source.

Asynchronous Tributary: The synchronization domains of the CE devices (SDCen) require that at least one of the

CE devices be externally timed to a non-PRS traceable source or use its internal free-running oscillator. The other

CE can be either line-timed, externally-timed, or operate on its internal free-running oscillator. Of these options,

only the line timed option will provide slip-free performance under non-fault conditions (see Table 4).

7.3.3.2 Asynchronous IWF, Synchronous Tributaries

Asynchronous IWF: In this scenario, not all of the synchronization domains of the IWFs (SDIWF)

use common

clock synchronization (e.g. timing from the same physical clock). Timing distribution for the IWFs may be

configured with each either line timed to the MEN, line-timing to a dedicated CE, or externally-timed to a PRS

traceable source.

Synchronous Tributary: The synchronization domains of the CE devices (SDCen) require that at least one of the CE

devices use external timing that is PRS traceable. The other CE, in the CES connection, may be either line-timed to

the originating CE or externally-timed to a PRS traceable source.

7.3.4 Synchronization Administration

Proper synchronization administration is crucial to support the transport of TDM emulation and circuit emulation

services. Each synchronization domain in the CES needs to be considered and provisioned independently. Table 6

provides a summary of typically available options per synchronization domain.

CE

SDCE

External

Line

Free-Run

IWF

SDIWF

Per IA

TSP

SDTDM

External

Line – CE

Line-MEN

Free-Run

Table 6: Timing Options per Synchronization Domain

Proper synchronization administration begins with the following concepts:

• Separate and Diverse – Synchronization equipment and paths should follow separate and diverse paths. This

reduces the chance that a single point of failure will cause service disruption. This philosophy extends to

equipment placement, cable routing, powering sources and network element configuration.

Synchronization Trail – All synchronization flows have a defined source and end. The context of this flow is

based on a logical flow rather than a topology (which could be linear, ring, mesh, or a combination of these).

The flow can even span multiple transport technologies including Ethernet, SONET and SDH. The source of a

synchronization trail should be from an independent reference that is either PRS traceable or from a free-running oscillator. The synchronization trail ends on a specific device that does not further propagate the source

timing information. Such a terminating device is necessary to prevent timing loops, which will cause transport

errors in the physical layer. Intermediate equipment in the synchronization trail can use timing information

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

MEN

SDED

PRS Traceable

Non-PRS Traceable

MEF 3

Page 28 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

derived from the source in a daisy-chained configuration. It should be noted that upstream timing events

(protection switching, holdover events, clock failures, etc.,) will be propagated to downstream equipment in this

configuration. For these cases, downstream equipment will need to have pre-determined fault modes to deal

with these circumstances when they arise. Such fault modes may include protection switching to use a different

line-timed reference or entry into holdover.

• Service Clock Preservation – The service clock is defined at the timing source used to generate the client

signal. In a CES transport network, the service clock frequency must be preserved in order to have error-free

transport of the physical layer. That is to say that the average bit rate of the transmitting and receiving

Customer Edge devices must be the same. Otherwise, the transport capability of the CES will be compromised.

Synchronization Traceability – In synchronization network planning, it is important to know where a timing

source originates, how it flows in the network, and its quality. The concept of traceability provides the answers

to these quantities. Synchronization flows can be described in terms of source traceability. That is to say, that

timing originates at a physical location and is used by equipment in a synchronization trail either by external or

line timing options. Common clock distribution is an example of a source traceable clock scheme. In the case

of frequency traceability, the quality of a timing signal can be specified. For example, the frequency of

synchronization sources may be specified as being PRS traceable or accurate to within 1 e-11 or (.00001 parts

per million). Plesiochronous clock distribution is an example of a frequency traceable clock scheme.

Three synchronization administration models are presented in the following sections. These models have been

chosen to illustrate how they may be used to accommodate the CES definitions presented in Section 6. These

models are:

Single Service Provider Owned Network – A single service provider controls the entire transport

synchronization trail. The service synchronization trail is separate.

Multi-Service Provider Owned Network – Multiple service providers control the transport synchronization

trail. The service synchronization trail is separate.

Private (customer owned) Network – A customer controls the entire transport and service synchronization

trails.

Table 7 summarizes these synchronization administration models and shows how they map to the CES service

definitions of Section 6.

CES Service

TDM Virtual Private Line

Service

TDM Access Line Service

Customer operated CES

over Metro Ethernet Service

Sync Admin Model Notes

Single Service Provider Can also be used with Private Network model.

Model

Single Service Provider Handoff to PSTN does not retime the payload (DS1 or

Model DS3). Therefore, there is no new sync trail.

Multi-Service Provider

Model

IWF at CES egress may be owned by a different service

provider.

Table 7: CES Services and supporting Synchronization Administration Models

7.3.4.1 Single Service-Provider Owned Network

The synchronization administration for a single service-provider owned network is illustrated in Figure 14.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 29 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

SDCEServicesync trailTransportsync trailCE1SDIWFSDTDMIWFTSPTransport TimingSDIWFSDTDMIWFTSPSDCECE2MENSDEDService TimingFigure 14: Synchronization Administration for a Single Service-Provider Network

A summary of available timing options for the Single-Service Provider owned network is presented on Table 8.

Device

CE

Timing

Domain

SDCE

Timing Options

External

Notes

Customer Option: Requires a collocated BITS or SSU to supply a

timing input to the CE. May be used as a source for the

synchronization trail.

Customer Option: May be used at the end of a synchronization trail.

Customer Option: May be used as the source for the synchronization

trail.

Service Provider Option:. Timing recovery for CES IWF must match

the IWF type. Details can be found in the IA as specified on Table 2

(CES Interface Definition).

Service Provider Option: Requires a collocated BITS or SSU to

supply a timing input to the PE.

Not Used since CE and TSP are on different synchronization trails.

Service Provider Option: Used if the IWF can provide timing.

Consistent with the service needs of the TSP. Options and capabilities

for a specific IWF is specified per the appropriate IA.

Service Provider Option: Used only if there is no suitable line or

external timing source.

Service Provider Option: Requires that all elements use PRS

traceable timing

Service Provider Option: Requires that one or more elements not use

PRS traceable timing.

Line

Free-Run

IWF SDIWF

See IA

TSP SDTDM External

Line – CE

Line-MEN

Free-Run

MEN SDED

PRS Traceable

Non-PRS

Traceable

Table 8: Timing Options Single-Service Provider Owned Network

Note that in the Single-Service Provider configuration, the synchronization trails for service-timing and transport-timing are separate. Separation of these synchronization trails is typically done for administrative and liability

reasons. An end customer will not be able to manage a service provider’s timing equipment. Likewise, a service

provider would typically not use timing from a CE due to the reliability and liability issues associated with a single

point of failure.

Separate synchronization trails also means that the physical timing source used to derive the service clock is not the

same as that used in the transport network (TSP/IWF and MEN). This is not to say that the source for the service

and transport cannot be PRS traceable, simply that one is not the source for the other.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 30 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.3.4.1.1 Service Timing – Single Service Provider Network

The Customer Edge synchronization domain (SDCE) is owned and maintained by the customer. Timing for the CE

(SDCE) requires that at least one CE be the source of timing. Referring to Table 4, any of the 6 options listed can be

used. The choice of which option to use will depend on economic and service needs. A PRS traceable source can

be highly reliable and accurate but requires physical placement and substantial engineering support. It is for this

reason that option 1 may be more expensive to administer than options 2 and 3.

Options 4 and 5 rely on the CE’s internal oscillator to provide a frequency source for the service clock. In this case,

due to the relaxed need for additional timing equipment (BITS or SSUs) this option may be less expensive to

administer than option 1 and require less engineering support.

Option 6 may yield undesirable slip performance due to the frequency difference between the free-running clocks.

This option is generally the least expensive but also has the lowest overall performance.

7.3.4.1.2 Transport Timing – Single Service-Provider Network

Timing for the single service-provider network assumes that the service provider owns the TSP/IWF and all of the

Ethernet devices in the MEN. The equipment will encompass synchronization domains: SDIWF, SDTDM

, and SDED.

For this case, it is up to the service provider to perform the synchronization administration of all transport and

synchronization equipment.

Regardless of synchronization options that the service provider may use, as listed in Table 5, it is the ability of the

service provider’s network to accurately transport the service clock that is most important. Therefore, the operation

of the CES interworking function is key to transport timing.

For the CES interworking function (SDIWF

) must use timing that is consistent with the IWF type.

The TSP synchronization domain (SDTDM

) provides a foundation for the CE IWF. PRS traceable timing may be

available via externally timing the TSP/IWF or line timing (from the MEN). It should be noted that the use of

synchronization messaging (SSM) between the MEN and TSP/IWF is required to always ensure that the TSP/IWF is

receiving PRS traceable timing. SSMs facilitate fault or protection switching from upstream failures. If SSMs are

not available, then external timing is the preferred timing mode for PRS traceable timing.

7.3.4.2 Multi Service-Provider Owned Network

Synchronization administration for a multi service-provider owned network is illustrated in Figure 15.

Servicesync trailTransportsync trailSDCECE1PSTNSDTDMProvider - ATransport TimingProvider - BTransport TimingSDIWFSDTDMIWFTSPMENSDEDProvider - CTransport TimingSDIWFSDTDMIWFTSPSDCECE2Service Timing

Figure 15: Synchronization Administration for a Multi Service-Provider Network

A summary of available timing options for a multi service-provider owned network is presented on Table 9.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 31 of 65

Device Timing

Domain

SDCE

Timing Options

External

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Notes

Customer Option: Requires a collocated BITS or SSU to supply a

timing input to the CE. May be used as a source for the

synchronization trail.

Customer Option: May be used at the end of a synchronization trail.

Customer Option: May be used as the source for the synchronization

trail.

Service Provider Option: Requires a collocated BITS or SSU to

supply a timing input to the PE.

Customer Option: May be used to recover timing from the received

client signal. This mode should only be used if synchronization

messaging is available (e.g. SONET or SDH line timing sources)..

Service Provider Option: Used only if the connecting TSP equipment

is PRS traceable. This mode should only be used if synchronization

messaging is available (e.g. SONET or SDH line timing sources)

Service Provider Option: Used only if there is no suitable line or

external timing source.

Service Provider Option: Timing recovery for CES IWF must match

the IWF type. Details can be found in the IA as specified on Table 2

(CES Interface Definition).

Service Provider Option: Requires a collocated BITS or SSU to

supply a timing input to the PE.

Customer Option: May be used to recover timing from the received

client signal. This mode should only be used if synchronization

messaging is available (e.g. SONET or SDH line timing sources).

Service Provider Option: Used only if the connecting PSTN

equipment is PRS traceable. This mode should only be used if

synchronization messaging is available (e.g. SONET or SDH line

timing sources)

Service Provider Option: Used if the IWF can provide timing.

Consistent with the service needs of the TSP. Options and capabilities

for a specific IWF is specified per the appropriate IA.

Service Provider Option: Used only if there is no suitable line or

external timing source.

Service Provider Option: Requires that all elements use PRS

traceable timing

Service Provider Option: Requires that one or more elements not use

PRS traceable timing.

CE

Line

Free-Run

PSTN SDTDM

External

Line – CE

Line-TSP

Free-Run

IWF SDIWF

See IA

TSP SDTDM External

Line – CE

Line – PSTN

Line-MEN

Free-Run

MEN SDED

PRS Traceable

Non-PRS

Traceable

Table 9: Timing Options Multi-Service Provider Owned Network

Note that in the Multi-Service Provider configuration, the synchronization trails for service-timing and transport-timing are separate. Also notice that the transport timing can be TDM based (PSTN) or Ethernet based (MEN).

Separation of these synchronization trails is typically done for administrative and liability reasons. An end customer

will not be able to manage a service provider’s timing equipment. Likewise, a service provider would typically not

use timing from a CE due to the reliability and liability issues associated with a single point of failure.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 32 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Separate synchronization trails also means that the physical timing source used to derive the service clock is not the

same as that used in the transport network (TSP/IWF and MEN). This is not to say that the source for the service

and transport cannot be PRS traceable, simply that one is not the source for the other.

7.3.4.2.1 Service Timing – Multi Service Provider Network

The Customer Edge synchronization domain (SDCE) is owned and maintained by the customer. Timing for the CE

(SDCE) requires that at least one CE be the source of timing. Referring to Table 6, any of the 6 options listed can be

used. The choice of which option to use will depend on economic and service needs. A PRS traceable source can

be highly reliable and accurate but requires physical placement and substantial engineering support. It is for this

reason that option 1 may be more expensive to administer than options 2 and 3.

Options 4 and 5 rely on the CE’s internal oscillator to provide a frequency source for the service clock. In this case,

due to the relaxed need for additional timing equipment (BITS or SSUs) this option may be less expensive to

administer than option 1 and require less engineering support.

Option 6 may yield undesirable slip performance due to the frequency difference between the free-running clocks.

This option is generally the least expensive but also has the lowest overall performance.

7.3.4.2.2 Transport Timing – Multi Service-Provider Network

Timing for the multi service-provider network assumes that no one service-provider owns the PSTN, TSP/IWF and

all of the Ethernet devices in the MEN. The equipment will encompass synchronization domains: SDIWF, SDTDM,

and SDED.

For this case, it is up to each service provider to perform the synchronization administration of all

transport and synchronization equipment in their domain.

Regardless of synchronization options that the service provider may use, as listed in Table 14, it is the ability of the

service provider’s network to accurately transport the service clock that is most important. Therefore, the operation

of the CES interworking function is key to transport timing.

The PSTN synchronization domain (SDTDM) preserves the service clock through the PSTN. PRS traceable timing

may be provided to all of the PSTN’s network elements (NEs) via externally by BITS or line timing from adjacent

NEs. It should be noted that the use of synchronization messaging (SSM) between NEs may be required to ensure

that each NE is receiving PRS traceable timing. SSMs facilitate fault or protection switching from upstream failures.

If SSMs are not available, then external timing is the preferred timing mode for PRS traceable timing.

The TSP synchronization domain (SDTDM

) provides a foundation for the CE IWF. PRS traceable timing may be

available via externally timing the TSP/IWF or line timing (from the MEN). It should be noted that the use of

synchronization messaging (SSM) between the MEN and TSP/IWF is required to always ensure that the TSP/IWF is

receiving PRS traceable timing. SSMs facilitate fault or protection switching from upstream failures. If SSMs are

not available, then external timing is the preferred timing mode for PRS traceable timing.

For the CES interworking function (SDIWF

) must use timing that is consistent with the IWF type as specified in the

appropriate IA.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 33 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.3.4.3 Private (customer owned) Network

Synchronization administration for a private network is illustrated in Figure 16.

SDIWFSDTSPIWFTSPTransport TimingSDIWFSDTSPIWFTSPSDCEServicesync trailTransportsync trailCE1SDCECE2MENSDEDService Timing

Figure 16: Synchronization Administration for a Private Network

A summary of available timing options for a private network is presented on Table 10.

Device

CE

Timing

Domain

SDCE

Timing

Options

External

Notes

Customer Option: Requires a collocated BITS or SSU to supply a

timing input to the CE. The most accurate means of timing

distribution.

Customer Option: May be used at the end of a synchronization trail.

Customer Option: May be used as the source for the synchronization

trail.

Service Provider Option: Timing recovery for CES IWF must match

the IWF type. Details can be found in the IA as specified on Table 2

( CES Interface Definition).

Service Provider Option: Requires a collocated BITS or SSU to

supply a timing input to the TSP. The most accurate means of timing

distribution.

Customer Option: May be used to recover timing from the received

client signal. This mode should only be used if synchronization

messaging is available (e.g. SONET or SDH line timing sources).

Service Provider Option: Used if the IWF can provide timing.

Consistent with the service needs of the TSP. Options and capabilities

for a specific IWF is specified per the appropriate IA.

Service Provider Option: Used only if there is no suitable line or

external timing source.

Service Provider Option: Requires that all elements use PRS

traceable timing

Service Provider Option: Requires that one or more elements not use

PRS traceable timing.

Line

Free-Run

IWF SDIWF

See IA

TSP SDTDM

External

Line – CE

Line-MEN

Free-Run

MEN SDED

PRS

Traceable

Non-PRS

Traceable

Table 10: Timing Options for a Private Network

Note that in this configuration, the synchronization trails for service-timing and transport-timing overlap. This is

allowed because the customer owns and operates the equipment that supports all of the synchronization domains. In

this case, the customer and service provider are the same entity, so combining synchronization trails should not

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 34 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

create any problems. The rules for timing, as with the single and multi service-provider cases, should still be

followed.

7.3.4.3.1 Service Timing – Private Network

The Customer Edge synchronization domain (SDCE ) is owned and maintained by the customer, just as the single

service provider case. In fact, all timing aspects related to private networks are identical to those of the single

service-provider case.

7.3.4.3.2 Transport Timing – Private Network

Timing for the private network assumes that the TSP/IWF and Ethernet devices in the MEN all owned by the same

customer. This equipment encompass synchronization domains: SDIWF, SDTDM, and SDED.

The options for

configuring this equipment are shown in Table 10.

The timing options for the Private Network are the same as those of the single service-provider case with the

exception of the line-CE mode for the TSP/IWF. In this mode, the TSP/IWF are allowed to use synchronization

from client signal as sent by the CE. In order to ensure the best quality timing, it is recommended that the CE

sourcing this client signal use PRS traceable timing. It should be noted that the use of synchronization messaging

(SSM) between the CE and TSP/IWF are required to always ensure that they are receiving PRS traceable timing.

SSMs facilitate fault or protection switching from upstream failures. If SSMs are not available, then external timing

is the preferred timing mode for PRS traceable timing.

7.4 PERFORMANCE

MONITORING AND

ALARMS

The CES should allow the operation of the normal mechanisms for monitoring the performance of the TDM service,

at the level of the TDM interface type. Generic performance monitoring requirements are specified in [G.826].

Performance monitoring in line with [G.826] is one-way, and occurs between two end-points. Specific requirements

for particular services are documented in the Requirements section.

7.4.1 Facility Data Link

The Circuit Emulation Service will carry any signal that meets the bit rate requirement specified in Section 8. In the

particular case of DS1 circuits using ESF framing, a Facility Data Link (FDL) may be present in the signal. The

DS1 ESF Facility Data Link is used to carry once-per-second Performance Report Messages as described in

[T1.403]. These messages carry information on numbers of CRCs, framing errors, line code violations and other

impairments detected over the last second.

The CES IWF is allowed to monitor the FDL, and to modify the relative position of the FDL with respect to the

TDM payload, but must not change messages carried by the FDL, or insert new FDL messages. For example, the

Interworking Function may be required to monitor Performance Report Messages as described in [T1.403]. The

means of carrying the FDL is to be defined in the Implementation Agreement.

7.4.2 Alarms

7.4.2.1 Unstructured Service

For unstructured services, all alarms received at the input of the Service Interface are carried through to the output

Service Interface without modification, since they are embedded in the data on the wire. In addition to this, the IWF

can detect a loss of signal (LOS) at the IWF Service Interface. Upon detection of LOS, the IWF is required to notify

the IWF at the opposite end of the CES service.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 35 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.4.2.2 Structured Service

For structured Nx64 service, the alarm status is not necessarily propagated within the data. Several kinds of alarms

can be detected at the point where the Service Interface is received by the IWF. The definition of alarm states is

given in [T1.403] for DS1 and [G.704] for E1. Some alarm situations require that an alarm condition detected at the

point where the TDM Service Interface is received by the MEN-bound IWF be propagated downstream to the CE-bound IWF responsible for reproducing the bit stream.

7.4.2.3 Buffer Underflow and Overflow

The IWF at the egress of the MEN will require a buffer in which the re-assembled data stream is stored before it is

transmitted out the Service Interface. The size of this buffer will be implementation dependent, but it must be large

enough to accommodate expected MEN frame jitter, while small enough to not introduce excessive delay in the

emulated circuit. This buffer will be subject to overflow or underflow if slight clocking differences exist between

the upstream and downstream IWF, or in the presence of unexpectedly large network jitter. In the case of an

underflow, or “data starvation” condition, data will have to be inserted into the TDM stream until a new Ethernet

frame has arrived. The data to be inserted is implementation-dependent.

Under some circumstances, such as a failure in the MEN network carrying the emulated service, the flow of

Ethernet frames to the re-assembly unit will stop for an extended period. This is effectively the same as a LOS

condition on the TDM network. If this condition persists for a long period, this should be signaled to the

downstream TDM equipment using a Trunk Conditioning procedure. For most applications, implementors are

advised to use a 2.5 ± 0.5 s integration period, in a manner analogous to that used to integrate Loss of Signal to

declare red alarm state.

Although not required as part of this specification, implementors may wish to consult Bellcore GR-1113-CORE and

ETSI ETS 300 353 Annex D for advice on the handling of various fault conditions.

7.4.2.4 Alarms in CCS Signaling

This revision of this CES document does not provide any specific support for Common Channel Signaling (CCS)

systems. CCS is supported in that it is transported transparently over the CES. However, this specification does not

address a reaction of a CES IWF to an incoming AIS signal when CCS is used. It should be noted that the presence

of the AIS signal will disrupt any CCS data link carried over that same link. This disruption will cause the signaling

system that uses the CCS link to declare an alarm. In this case, no further conditioning is required.

Similarly, a second alarming condition (in CCS systems) is not addressed by this document. The situation is if the

VC that carries the CCS link is different than the VC that carries the DS0s controlled by that CCS link. There is

currently no means of conveying an alarm on that VC which carries the DS0s to the CCS system. This is similar to

failure case in TDM networking where only some of the DS0s fail. In TDM networking, this case is sometimes

handled by network management actions and other times by DS0 testing systems used just as calls are being

established. But since this version of this document does not specify support for CCS signaling systems, this

situation is left for further study.

7.4.3 End-to-End Delay

End-to-end delay requirements are application-specific. End-to-end delay requirements are therefore beyond the

scope of this specification. However, it should be noted that excess delay could have adverse effects on some traffic

types, such as voice.

7.5

SERVICE

IMPAIRMENT

This section addresses impairments to the emulated TDM service caused by errors within the MEN. Principally it

addresses the relationship between the performance parameters of the underlying Metro Ethernet Network (MEN)

transport to the defined service impairment metrics for the TDM service being emulated – errored seconds (ES) and

severely errored seconds (SES).

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 36 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.5.1 Errors within the MEN causing TDM service impairment

There are three performance parameters defined for an EVC that have an impact on the TDM service impairment

metrics: Frame Loss, Frame Delay and Frame Jitter. In addition any bit errors induced in the data flow across the

MEN will also have an impact on the TDM service, although there is no performance metric defined to measure this

within the MEN.

7.5.1.1 Frame Loss

Frame loss is defined as the percentage of in-profile frames (“green” frames that are within CIR/CBS) not reliably

delivered over a measurement interval T.

Frame LossT = (1 – frames delivered to destination / total frames sent to destination) x 100

Frame loss will cause a burst of bit errors in the re-constructed TDM stream.

7.5.1.2 Frame Delay and Frame Jitter

Frame Delay is defined as the maximum delay measured for a percentile (P) of successfully delivered frames over a

measurement interval T. Frame Jitter can then be derived from the Frame Delay measurement using the maximum

and minimum of Frame Delay samples over same measurement interval T and percentile (P). Frame Jitter can be

calculated as follows:

Frame JitterT,P = Frame DelayT,P(Max. measured delay value – Min. measured delay value)

Typically the Frame Jitter is used to size the re-assembly buffer in the IWF at the egress of the MEN (see section

7.4.2.3). This buffer is sometimes referred to as the “jitter buffer”. However, the Frame Jitter figure is calculated on

a percentile of frame delay values, where the percentile is defined as part of the Service Level Specification.

Therefore a small percentage of frames will arrive outside the specified jitter level. Depending on the size of the

jitter buffer, these frames may either arrive too late to be played out, or too early to be accommodated within the

buffer. Such frames will then be discarded, and must be considered lost for the purposes of reconstructing the TDM

service.

7.5.1.3 Bit Errors

Bit errors induced in the MEN will normally be detected by a CRC error in the frame check sequence. Therefore,

the errors will cause the whole frame to be discarded. On some occasions a frame containing bit errors may still

yield a correct CRC value, but this is expected to be extremely rare. Therefore, as far as the IWF is concerned, any

bit errors in the data stream will appear as lost frames.

7.5.1.4 Frame Error Ratio and IWF behaviour

The collective sum of all the above errors (frame loss, excess frame jitter and bit errors) can be aggregated into a

single measure termed “Frame Error Ratio” (FER), defined as the total of all effects leading to the loss of or

discarding of a frame. For the purposes of TDM emulation, an Ethernet frame is deemed to be errored if it:

a. fails to arrive at the egress IWF

b. arrives too late to be played out

c. arrives too early to be accommodated in the jitter buffer

d. arrives with bit errors causing the frame to be discarded

In order to maintain timing integrity, the IWF must then insert an equivalent number of octets of data into the

reconstructed stream. The data to be inserted is application and/or implementation dependent. If the frame should

subsequently arrive, it should be discarded.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 37 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.5.2 Relationship to TDM service impairment metrics

For DS1, fractional DS1, and DS3 there are requirements in ANSI [T1.510] for SES and ES related to the access

portion of a TDM link (which relates to the metro portion of the link). When voice was the predominant application,

one could, assuming no error control, derive some reasonable requirements for BER since an end-to-end PCM-encoded voice signal would tolerate a 10-6 BER in the TDM circuit. (Source: SR-TSV-00275, Iss. 1, March 1991).

Although it should be noted that with Extended Super Frame (ESF) a loss of a single bit within a frame will cause a

block error to be detected by the error detection code resulting in a ES even if the voice quality was not adversely

effected. The following discussion of ES and SES assumes an absence of error control in the MEN.

For the European standards, there are equivalent metrics defined in [G.826]. The Errored Second Ratio (ESR) and

Severely Errored Second Ratio (SESR) are defined across an international link rather than a metro, with guidelines

for calculating the requirement on the national portion of the link. [G.826] does not break down the figures further

to give the requirement on the metropolitan portion. In the discussion below, the national figures are used to define

an upper bound on the MEN performance requirement.

7.5.3 Errored Seconds Requirement for PDH Circuits

An ES is a one-second interval with one or more bit errors. Each 1.544 Mbits/s (including Nx64kbits/s) and 44.736

Mbits/s channel requires that, over 30 or more consecutive days, fewer than 0.25% and 0.125%, respectively, of the

seconds are errored seconds (Source: [T1.510]). The conversion of ES for PDH circuits to bit error ratio (BER) and

frame error ratio (FER) for Ethernet CES virtual circuits is dependant on factors such as the number of TDM frames

packed into the CES Ethernet frame (packing density) and the size of the CES header attached to each Ethernet CES

frame.

Conversion of the Errored Seconds requirements to an Ethernet Frame Error Ratio is possible based on knowledge

of the number of TDM frames packed into each Ethernet frame containing CES. Figure 17 shows the allowed

Frame Error Ratio (FER) for given packing densities. This FER is for the Ethernet Virtual Circuit (EVC) associated

with the circuit emulated service, assuming that only the associated CES frames are used in the frame loss

calculation. Values of FER for specific packing densities (8 T1 or E1 frames and 2 T3 or E3 frames) are given in

Table 11.

The FER is derived from the Errored Seconds objective by the following equation:

FER = %ES /(100 * CES frame rate).........[1]

Where: CES frame rate = TDM data rate / (N * bits per TDM frame)

N = number of TDM frames per CES frame

Bits per TDM Frame = 193 for 1.544 Mbits/s; 4760 for 44.736 Mbits/s

This figure has to include all possible sources of lost or errored Ethernet frames, as described in section 7.5.1. It

does not allow breakdown into individual causes of error.

%ES (note 1, 2)

FER (note 3, 4)

1.544 Mbits/s

(T1)

0.25

2.5x10-6

1.544 Mbits/s

(T1 enhanced)

0.0625

6.3x10-7

44.736 Mbits/s

(T3)

0.125

2.7x10-7

2.048 Mbit/s

(E1)

0.7

7.0x10-6

34.368 Mbit/s

(E3)

1.3125

3.3x10-6

Table 11: Example conversion from ES to FER for Ethernet Virtual Circuit

Note 1 T1, T2 enhanced and T2, reference [T1.101] (Access segment of the reference model)

Note 2 E1, E3 reference [G.826], using 17.5% of the international objective as defined in section 7.2. This

represents the national objective for errored seconds, and hence is higher than would be allocated to the

access segment, but provides an upper bound.

Note 3 See Figure 17 for details of FER curves vs packing density

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 38 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Note 4 FER values represent the packetization of 8 T1 or E1 frames and 2 T3 or E3 frames in an Ethernet frame

1.00E-041.544 Mb/s1.544 Mb/s Enhanced44.736 Mb/s2.048 Mb/s (note 1)34.368 Mb/s (note 1)Allowed

FER1.00E-051.00E-061.00E-3035Packing (TDM frames per CES frame)

Figure 17: Allowed Frame Error Ratio to meet ES objectives

Note 1 As with the table above, E1 and E3 figures use the national objective for ES. This is higher than would be

expected for a metro network, and hence lower frame error ratio is required in a real MEN. The exact value

is for further study.

7.5.4 Severely Errored Seconds Requirement for PDH Circuits

A severely errored second (SES) is defined as a period of time, during which bits are transferred from a source to a

destination, where > 30% of the blocks received are errored, or at least one severely disturbed period occurred. A

severely disturbed period occurs when, over a period of time equivalent to four contiguous blocks or 1 ms, which

ever is larger, all contiguous blocks are affected by a high bit error density of >10-2. This definition applies for a

specified block size.

Note: Where a suitable block is not available, an alternate definition can be used: - A specified period of time during

which bits are transferred from a source to a destination, where a BER worse than 10-3 occurs [T1.503].

The relationship between block size and CES frame size will greatly influence the allowed FER. The block size

related to an Extended Super Frame (ESF) is 193 bits x 24 whereas a DS3 block is 4760 bits. For ESF a number of

CES Frames may be required to form a block and any loss of a CES frame will impact the whole block. For DS3 a

single CES frame can carry multiple blocks and the loss of a CES Frame will impact all of the blocks in the CES

Frame. Figure 18 shows the relationship between packing and allowed FER to meet the SES requirement of <30%

of the blocks received being errored for a SES requirement of 0.01%. It should be noted that for DS3 services only

2 DS3 frames can be packed in a CES Ethernet frame. The requirement for SES is 0.01% for the Access portion as

defined in [T1.510] for all PDH services (1.544 Mbits/s, 1.544 Mbits/s enhanced and 44.736 Mbits/s).

In the case where errors are Poisson in nature and not correlated then equations 2 through 4 can be used to calculate

FER for SES. Results are plotted in Figure 18.

If Fs/N – Int(Fs/N) = 0 then

%SES * BE * N

FER = --------------- ....................[2]

Fs

Where: N

MEF 3

= number of TDM frames per Ethernet CES frame

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 39 of 65

Fs

FER

%SES

BE

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

= number of TDM frames in a block (e.g. 24 for ESF)

= FER limit for %SES

= percent severe error second limit = 0.01%

= the block error limit = 30%

If Fs/N – Int(Fs/N) <> 0 then

%SES * BE

FER = ------------------------

(Fs/N+1 – GCD(Fs,N)/N)

Where: N

Fs

FER

%SES

BE

GCD

..........[3]

= number of TDM frames per Ethernet CES frame

= number of TDM frames in a block (e.g. 24 for ESF)

= FER limit for %SES

= percent severely errored seconds limit = 0.01%

= the block error limit = 30%

= Greatest Common Divisor function

If there are one or more blocks packed in an Ethernet CES Frame and the blocks do not span Ethernet CES Frame

boundaries then:

FER = %SES * BE............................[4]

Where: FER

%SES

BE

= FER limit for 0.01% SES

= percent severely errored seconds limit = 0.01%

= the block error limit = 30%

1.00E-031.00E-04Allowed

FERDS1 ESF1.00E-05DS-3E1 (note 1)E3 (note 1)1.00E-061.00E-Packing (TDM Frames per CES Frame)

Figure 18: Relationship between Packing and FER for 0.01% SES

Note 1 As with the ES figures, E1 and E3 figures use the national objective for SES. This is higher than would be

expected for a metro network, and hence lower frame error ration is required in a real MEN. The exact

value is for further study.

Note 2 The discontinuities in the plots are due to the boundaries where a block does not span across an Ethernet

CES frame.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 40 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

In the case where no block definition is available then the SES definition given in ANSI T1.510-1999 can be used.

-3The resultant FER limits are as follows: 10 BER on the TDM data stream is equivalent to a TDM frame loss ratio

of 19% for T1. The resultant FER is given by:

FER = %SES * 0.19/N

Where: N

FER

%SES

....................[5]

= number of TDM frames per Ethernet CES frame

= FER limit for 0.01% SES

= percent severely errored seconds limit = 0.01%

The resultant FER is given in Figure 19.

It is recommended that where a block is clearly defined, such as for ESF or DS3, that the requirements given Figure

-318 are used, and where no block requirement is given that Figure 19 be used. It should be noted that a 10 BER on

a TDM data link carrying ESFs would result in 100% of the blocks being errored.

1.00E-04Allowed

FER1.00E-05DS-1 based on 10^-31.00E-061.00E-Packing (TDM Frames per CES Frame)

Figure 19: Relationship between Packing and FER for 0.01% SES based on 10 BER on end to end TDM

data stream

-37.5.5 Service Impairments for SONET/SDH Circuits

This section provides an analysis of the Frame Error Ratio (FER) requirements from the Metro Ethernet Network

(MEN) in order to support SDH/SONET Circuit Emulation Services (CES).

7.5.5.1 Performance Objectives for the SONET/SDH network

The error performance limits used in this document are taken from ITU recommendations [M2101.1] and [G.826].

[M2101.1] describes a Hypothetical Reference Model (HRM) for performance of the international path. The

allocation of the performance objectives per segment of the path in percentage of the end to end allocation is

provided in Table 2A of [M2101.1]. For the purpose of calculation of performance objectives for the Metro

connections, 2% of the end to end allocation is taken, which correspond to a segment with diameter less than 500

km.

7.5.5.1.1 ES/SES evaluation criteria

[M2101.1] gives the guidelines for ES/SES evaluation criteria for LOVC in Table B.1 and for HOVC in table B.2.

From the guidelines specified in these tables, the interpretation of ES and SES for packet loss and BER effects in

SDH/SONET CES can be derived, and are discussed in the next section.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 41 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.5.5.1.2 Performance Objectives in [G.826]

Performance objectives as defined in [G.826] are measured in ESR, SESR and BBER. The definitions of these three

terms are:

ESR: Error Second Ratio; The ratio of Error Second (ES) to total seconds during a fixed measurement interval.

SESR: Severe Error Second Ratio: The ratio of SES to total seconds in available time during a fixed measurement

interval.

BBER: Background Block Error Ratio: The ratio of Background Block Errors (BBE) to total blocks in available

time during a fixed measurement interval. The count of total blocks excludes all blocks during SESs.

Block sizes used for SDH path performance monitoring is provided in Table C.1 of [G.826].

The End to End error performance objectives for international digital Hypothetical Reference Model at or above the

primary rate is given in Table 1 of [G.826].

The recommended time period for evaluation of the performance objectives is one month.

7.5.5.2 MEN Performance Objectives

This section discusses the interpretation of the requirement specified in the previous section to the Metro Ethernet

Networks. The requirements from the MEN are specified in maximum Frame Error Ratio (FER) provided by the

MEN for the various SDH/SONET services and rates.

The SDH/SONET service provided over the MEN should stand within the limits of the ESR, SESR and BBER

requirements specified in [G.826] and the ES and SES requirements specified in [M2101.1]. These requirements can

be directly tested for each emulated circuit using standard SDH/SONET and PDH testers.

The [G.826] requirements are specified in ESR, SESR and BBER. The MEN requirements are specified in BER and

FER. The derivation requires assumptions on the distribution of BER and FER within the MEN.

1. ESR requirements derive FER MEN requirements assuming that the errors does not occur in bursts, rather each

error occur in a different measurement interval (separated by at least one second). Since errors can occur in

bursts, the actual FER measured in MEN may be higher than the requirements derived from ESR and still meet

[G.826] requirements.

2. BBER measures the rate of blocks received in error, without counting the blocks received in SES intervals.

Since the number of blocks per packet is known, MEN FER requirements can be directly deduced. The FER

requirements calculated from ESR is usually stricter than the requirements calculated from BBER. Therefore

FER measured in the MEN may be higher than the FER requirement derived from BBER, and still meet [G.826]

requirements, as some of the errors would account for SES.

3. SESR requirements do not directly derive FER requirements. SESR requirement limits the number and length

of bursts of errors. Rather, the maximal FER that can be accounted to SES is calculated.

7.5.5.2.1 Error Second Objective

Table 12 summarizes the FER objectives for MEN derived from ESR requirements defined in [G.826]. Note that

these requirements correspond to FER assuming that only one error occurs each second. Performance objectives for

burst of errors or multiple packet drops are covered in BBER objective section.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 42 of 65

PathVC-11VC-12VC-3VC-4VC-4-4cPackingpkt/blks14141331.89Block Sizebits8328321125168Ratepps2240Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

P Sizebits488106648576ESRG.8260.040.040.040.040.050.050.160.160.16ESRG.826 MEN8.00E-048.00E-048.00E-048.00E-041.00E-031.00E-033.20E-033.20E-033.20E-03CESFER4.00E-071.00E-074.00E-071.00E-071.25E-074.17E-081.33E-072.22E-074.44E-08

Table 12: Sparse FER objectives for MEN [G.826]

Where:

Path:

Packing:

The SDH/SONET LOVC or HOVC service.

Defines the number of packets carrying a single block. Packing values in orange represent the default

packing values for SDH/SONET emulation.

Rate of the channel in packets per second

The Ethernet frame size in bits. The overhead used for calculation of the CES BER ration comprise of

14 Ethernet header bytes, 4 VLAN tag bytes, 4 bytes for circuit multiplexing label 4 bytes for CES

control word and 4 bytes for Ethernet CRC, altogether 28 bytes. Where applicable (VC-11 and VC-12

8k flows), padding to minimal Ethernet is added as well.

Block Size: Block sizes in bits as defined in [G.826]

Rate:

P Size:

ESR [G.826]: The end to end performance objective for international link

ESR MEN: 2% of [G.826] specified end to end performance objective, suitable for the MEN segment.

CES FER:

The maximal Frame Error Ratio within the CES Ethernet frames. This is derived from the following

equation:

CES FER = ESR MEN / Rate

7.5.5.2.2 BBER Objectives

Assuming that no Severe Error Seconds occur in a measurement interval, the BBER ratio provide a direct upper

limit for Frame Error Ratio, as the relation between blocks and packets is fixed given by the packing ratio.

Burst of errors may cause severe error seconds. When a severe error second occur, the frames within this second

(both good and errored) are not counted in these requirements.

PathVC-11VC-12VC-3VC-4VC-4-4cPackingpkt/blks14141331.89Block Size bits8328321125168Ratepps2240P Sizebits488106648576BBERG.8260.00020.00020.00020.00020.00020.00020.00020.00020.0001BBERG.826 MEN4.00E-064.00E-064.00E-064.00E-064.00E-064.00E-064.00E-064.00E-062.00E-06CES

BFER2.00E-062.00E-062.00E-062.00E-062.00E-062.00E-062.00E-062.00E-061.00E-06

Table 13: Background FER and BER objectives for MEN [G.826]

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 43 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

Where:

Path:

Packing:

The SDH/SONET LOVC or HOVC service.

Defines the number of packets carrying a single block. Packing values in orange represent the default

packing values for SDH/SONET emulation.

Rate of the channel in packets per second

The Ethernet frame size in bits. The overhead used for calculation of the CES BER ration comprise of

14 Ethernet header bytes, 4 VLAN tag bytes, 4 bytes for circuit multiplexing label 4 bytes for CES

control word and 4 bytes for Ethernet CRC, altogether 28 bytes. Where applicable (VC-11 and VC-12

8k flows), padding to minimal Ethernet is added as well.

Backgournd Block Error Ratio (BBER) objectives as defined in [G.826]

Block Size: Block sizes in bits as defined in [G.826]

Rate:

P Size:

BBER:

BBER MEN: BBER objectives for MEN, 2% of BBER end to end requirement

CES BFER: The maximal background FER for the CES frames. BFER accounts for both single as well as small

bursts of frame errors. CES BFER does not count frames in Severe Error Seconds, e.g. during large

bursts of FER. This is derived from the following equation:

CES BFER = BBER MEN / 2

A packet does not carry more than one block. A packet may carry segments from two blocks. To account for the

latter, the 2 factor is added.

7.5.5.2.3 Severe Error Second Objectives

The criteria for SES are defined in table B1 and B2 of [M2101.1]. In particular, one of the SES criterions is errors in

over 300 blocks during one second. In SDH/SONET emulation this corresponds to an error or loss of about 300

frames during one second interval. We refer to these cases as connectivity problems within the MEN. MEN

connectivity problems may be failure of a switch or a link that leads to loosing a series of packets.

CES service running over MEN with Frame Error Ratio specified in Table 14 can still stand within [G.826]

requirements, if most errors occur within single intervals, i.e. within the SES intervals.

PathVC-11VC-12VC-3VC-4VC-4-4cPackingpkt/blks14141331.89Block Sizebits8328321125168Ratepps2240P Sizebits488106648576SESRG.8260.0020.0020.0020.0020.0020.0020.0020.0020.002SESRG.826 MEN4.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-05CES

SESFER4.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-05

Table 14: Maximal FER accounted by SESR [G.826]

Where:

Path:

Packing:

The SDH/SONET LOVC or HOVC service.

Defines the number of packets carrying a single block. Packing values in orange represent the default

packing values for SDH/SONET emulation.

Rate of the channel in packets per second

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Block Size: Block sizes in bits as defined in [G.826]

Rate:

MEF 3

Page 44 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

P Size: The Ethernet frame size in bits. The overhead used for calculation of the CES BER ration comprise of

14 Ethernet header bytes, 4 VLAN tag bytes, 4 bytes for circuit multiplexing label 4 bytes for CES

control word and 4 bytes for Ethernet CRC, altogether 28 bytes. Where applicable (VC-11 and VC-12

8k flows), padding to minimal Ethernet is added as well.

CES SESFER The maximal Frame Error Ratio within the CES Ethernet frames that can be accounted for SES. This

is derived from the following equation:

CES SESFER = SESR MEN

7.5.5.3 Summary & Discussion

Table 15 describes the FER MEN requirements. The actual FER can range between CES FER and CES SESFER

and [G.826] Requirements would still be met.

PathVC-11VC-12VC-3VC-4VC-4-4cRatepps2240CESFER4.00E-071.00E-074.00E-071.00E-071.25E-074.17E-081.33E-072.22E-074.44E-08CES

BFER2.00E-062.00E-062.00E-062.00E-062.00E-062.00E-062.00E-062.00E-061.00E-06CES

SESFER4.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-054.00E-05

Table 15: MEN FER Requirements for SDH/SONET CES

MEN that can not stand within the FER requirements can still run SDH/SONET Circuit Emulation Services. The

implication would be that larger than 2% of the international end to end ESR, BBER and SESR requirements would

be used by the MEN. A similar approach is taken for other technologies in [M2101.1] (Satellite links).

7.5.6 Availability Requirements

Network unavailability is specified in [T1.510] for DS1 and DS3, [T1.514] for SONET, and in [G.827] for E1, E3

and SDH circuits. The “unavailable state” is entered at the start of a period of 10 consecutive Severely Errored

Seconds (SES). The “available state” is resumed at the start of a period of 10 consecutive seconds of which none are

severely errored.

Translating this to give an understanding of the availability of an EVC used to carry an emulated circuit requires

knowledge of the level of errors in a given second that will result in a SES. One difficulty is that there are several

ways of defining a SES.

[T1.503] gives the following three definitions of SES:

Either

Or

Or

Either

Or

>= 30% of blocks received definition 1

BER > 10-2 for four definition 2

BER > 10-3 (where no suitable blocks are defined).........definition 3

>= 30% of blocks received are errored ............same as definition 1

one or more defects (i.e. LOS, AIS, LOF).......................definition 4

[G.826] gives two definitions of a SES:

Taking the first definition, if all the errored CESoETH frames are consecutive, then this translates into a simple level

of 30% or errored CESoETH frames per second. If the errors are randomly or evenly distributed, then the level of

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 45 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

errors will depend on the number of blocks contained in each CESoETH frame, and the exact distribution of the

errors (e.g. has the error hit a new block, or hit an already errored block?).

However, the other definitions place much stricter requirements on the FER. Definition 2 requires only 4 errored

CESoETH frames to yield a SES (assuming each packet hits a different block). However, with the assumption that

errored frames are consecutive, an FER of around 1% would be required to yield a SES.

Definition 3 translates simply into an FER of >0.1% to yield a SES (i.e. > 1 errored CESoETH frame in 1000).

Definition 4 is the strictest, since it requires only one errored CESoETH frame to yield an SES (since each lost

frame may result in a temporary defect, e.g. AIS). However, given a CESoETH frame rate in the region of 1000

frames per second, it is similar to definition 3.

Therefore in the worst case, one errored CESoETH frame per second for 10 consecutive seconds may cause

unavailability of the TDM service. With the use of suitable frame error concealment techniques in the CES IWF,

(e.g. to prevent loss of multiframe synchronization), loss of availability of the TDM service may be reduced.

7.6 TDM

SIGNALING

It is not normally required for Circuit Emulation Service to intercept or process TDM signaling, e.g. Channel

Associated Signaling (CAS) or Common Channel Signaling (CCS). Signaling is embedded in the TDM data stream,

and hence it is carried end-to-end across the emulated circuit. The signaling can be extracted by the end equipment

from the data that has been transported across the MEN.

The exception is Nx64 kbit/s service where CAS signaling is supported. Nx64 kbit/s Service with CAS requires

direct recognition and manipulation of the signaling bits by the CES IWF. This mode is necessary to support

multiplexed Nx64 kbit/s applications requiring DS1 Robbed Bit Signaling or E1 CAS support, where the association

between the signaling bits and the channel may otherwise be lost.

7.7 LOOPBACKS

Loopbacks should be supported at the level of the TDM interface type. Loopback of internal multiplexed levels

(e.g., a DS1 level inside an OC-3 line) is outside the scope of this document.

7.7.1 Provider Controlled Loopbacks

It is suggested that provider-controlled loopbacks should be supported in the manner illustrated in Figure 20, where

the direction of loopbacks is from PE1 (near end) towards PE2 (remote end). Loopback for PE2 to PE1 is the

reverse of Figure 20. The aim of the loopback is to isolate the fault with respect to PE1. Therefore, four loopbacks

at the system level are suggested:

1) a PE1-CE1 facility loopback, which should be located at a point within the provider’s network as close as

practicable to the Network Demarcation, for example, at a final span line repeater or SmartJack;

2) a PE1 terminal loopback of the TDM interface;

3) a PE1 terminal loopback of the MEN interface; and

4) a PE1-PE2 MEN loopback.

The loopback at the Network Demarcation must comply with the appropriate specification for the TDM interface

type (e.g. ANSI [T1.403] for a DS1 interface). Refer to [T1.107] for more information on how to handle loopbacks.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 46 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

CE 1PE 1MENPE 2CE 2 1 2 3 4NetworkDemarcationNetworkDemarcation

Figure 20: Provider Controlled Loopback Points

7.7.2 Customer Controlled Loopbacks

It is also possible for customers to use loopbacks to verify the operation of either their own equipment, or of the

service being provided. The aim of the loopback is to isolate the fault with respect to a particular item of customer

equipment (e.g. CE1), while treating the service provider network as a “black box”. Therefore, two loopbacks are

suggested, as illustrated in Figure 21:

1) Local loopback at CE1

2) Remote loopback at CE2

The required loopbacks for CE2 are the reverse of Figure 21.

CE 1PE 1MENPE 2CE 2 1 2NetworkDemarcationNetworkDemarcation

Figure 21: Customer Controlled Loopback Points

Customer controlled loopbacks are often initiated under manual control, although signaling may be provided to

operate the remote loopback. The nature of such signaling is implementation specific. Customers are not normally

able to control the provider loopbacks shown in Figure 20, except via a service call to the provider concerned.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 47 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

7.8 PROTECTION

TDM networks usually include protection mechanisms for service recovery in less than 50ms. The CES solution

should fit into such a protection scheme. This means the following:

A CES service running over a MEN should have capabilities for sub-50ms protection within the MEN part.

A failure of the TDM interface connecting a TDM legacy network to the MEN should cause a switchover

to an alternative TDM interface. This switch should interact both with the MEN and with the legacy

network, so that both sides of the interface start working with the alternative interface.

The following scenarios may be relevant for CES protection:

7.8.1 Scenario 1 – dual unprotected services

In such topologies a failure of one of the interfaces connecting between the TDM and MEN, or inside the MEN will

cause the TDM network to work with the alternative CES connection. This may use 1+1 protection, 1:1 protection

or 1:N protection, but is outside the scope of the CES solution.

CES unprotected connection 1CECESIWFCESIWFCETDMCECESIWFMENCES unprotected connection 2CESIWFTDMCE

Figure 22: Dual unprotected services

The CES service is a non-protected service. Protection is established through the mechanisms of the TDM network

connected on both sides of the MEN. For example, a SONET network connected on both sides of the CES

connection could perform the protection switching based on either APS (Automatic Protection Switching) or BLSR

(Bi-directional Line Switched Ring). Since the protection related information is carried in the line overhead (LOH),

there are two possibilities:

The CES service is not terminating the line layer. The protection information is carried across the MEN to

the other side.

The CES service is terminating the line layer. Upon failure of the path in the MEN, the PE updates the

relevant protection bytes in the LOH within the requirements of SONET/SDH, so that the SONET/SDH

equipment can make the switching in time. This may be by asserting AIS/RDI bits for the line.

7.8.2 Scenario 2 – dual protected services

This scenario is similar to the previous one; however, each CES connection is protected inside the MEN.

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 48 of 65

Circuit Emulation Service Definitions, Framework and

Requirements in Metro Ethernet Networks

CECESIWFCES protected connection 1CESIWFCETDMCECESIWFMENCES protected connection 2CESIWFTDMCE

Figure 23: Dual protected services

In case of a failure in the TDM interface, the behavior is similar to the previous scenario.

In case of a failure inside the MEN, both the MEN internal mechanisms may trigger bypass, and the TDM network

may initiate a switch to the alternate TDM interfaces as well. This requires careful design in order to avoid races

between the TDM and the MEN protection mechanisms.

7.8.3 Scenario 3 – Single protected service

In this scenario, a single TDM interface is used on each side of the MEN. In this case, the CES connection in the

MEN needs to be protected using MEN protection mechanisms. A failure inside the MEN would be bypassed in

less than 50ms, so that service is maintained.

CES protected connectionCECESIWFCESIWFCETDMMENTDM

Figure 24: Single protected service

This scenario does not cover a case of TDM interface failure. The end-to-end protection mechanisms require

detecting the liveliness of an end-to-end path. This is done generically through an OAM mechanism. Note that in

the case of CES, traffic is constantly flowing through the active path (unlike Ethernet connections where there may

be silent periods on the connection). Receiving traffic at the end of the path can therefore be used for detecting of

the path liveliness, and trigger switch-over in case of failure.

7.8.4 Scenario 4 – Single to dual interface service

CESIWFCESIWFCESIWFCES protected connectionCE 1TDMCE 2CE 3MENCES+MuxCES+MuxCE ATDMCE B

Figure 25: Single to dual interface connection

MEF 3

© The Metro Ethernet Forum 2004. Any reproduction of this document, or any portion thereof,

shall contain the following statement: "Reproduced with permission of the Metro Ethernet Forum."

No user of this document is authorized to modify any of the information contained herein.

Page 49 of 65