+ All Categories
Home > Documents > Ethernet Over Transport

Ethernet Over Transport

Date post: 09-Apr-2018
Category:
Upload: anovarebooks
View: 221 times
Download: 0 times
Share this document with a friend

of 36

Transcript
  • 8/8/2019 Ethernet Over Transport

    1/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 1

    Ethernet over Transport

    Technology White Paper

    by Steve Gorshe, Principal Engineer

    Issue No. 1: July, 2005

    PMC-Sierra, Inc.

  • 8/8/2019 Ethernet Over Transport

    2/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 2

    Legal Information

    Copyright

    Copyright 2005 PMC-Sierra, Inc. All rights reserved.

    The information in this document is proprietary and confidential to PMC-Sierra, Inc., and for its

    customers internal use. In any event, no part of this document may be reproduced or

    redistributed in any form without the express written consent of PMC-Sierra, Inc.

    PMC-2050704 (01)

    Disclaimer

    None of the information contained in this document constitutes an express or implied warranty

    by PMC-Sierra, Inc. as to the sufficiency, fitness or suitability for a particular purpose of any

    such information or the fitness, or suitability for a particular purpose, merchantability,

    performance, compatibility with other parts or systems, of any of the products of PMC-Sierra,

    Inc., or any portion thereof, referred to in this document. PMC-Sierra, Inc. expressly disclaims

    all representations and warranties of any kind regarding the contents or use of the information,

    including, but not limited to, express and implied warranties of accuracy, completeness,

    merchantability, fitness for a particular use, or non-infringement.

    In no event will PMC-Sierra, Inc. be liable for any direct, indirect, special, incidental or

    consequential damages, including, but not limited to, lost profits, lost business or lost data

    resulting from any use of or reliance upon the information, whether or not PMC-Sierra, Inc. hasbeen advised of the possibility of such damage.

    Trademarks

    For a complete list of PMC-Sierras trademarks and registered trademarks, visit:

    http://www.pmc-sierra.com/legal/

    Patents

    Relevant patent grants may exist.

    http://www.pmc-sierra.com/legal/http://www.pmc-sierra.com/legal/http://www.pmc-sierra.com/legal/http://www.pmc-sierra.com/legal/http://www.pmc-sierra.com/legal/
  • 8/8/2019 Ethernet Over Transport

    3/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 3

    Abstract

    A convergence of technology and applications has created an increased desire for Ethernet

    WAN connectivity. In many ways, Ethernet is an obvious technology choice. It simplifiesenterprise network administration, and provides a ubiquitous, inexpensive infrastructure of

    existing Ethernet interfaces on enterprise equipment. The development of VCAT and GFPallows efficient transport of Ethernet frames through SONET/SDH networks. In addition,

    Ethernet bridge and router technology requires less provisioning and administration than

    technologies such as ATM. As some carriers begin a migration to using packet switching

    technology such as MPLS in their core networks rather than traditional circuit switching,

    Ethernet again looks like an excellent fit as an access technology.

    This white paper describes how standards are developing to provide Ethernet WAN

    connectivity and primarily focuses on the set of new standards for Ethernet transport over public

    networks developed by ITU-T Study Group 15 (SG15). It discusses the Ethernet services that

    are being considered over public WANs; transport network models that can be used to supportEthernet services; the Ethernet-based user network interfaces (UNIs) and network-to-network

    interfaces (NNIs) required for transport network equipment to carry Ethernet services;

    Operations and Maintenance (OAM) capabilities; and protection and restoration technologies

    that can be used to guarantee carrier-grade service reliability for Ethernet WAN services.

    About PMC-Sierra

    PMC-Sierra is a leading provider of high-speed broadband communications and storage

    semiconductors and MIPS-Poweredprocessors for Enterprise, Access, Metro Optical

    Transport, Storage Area Networking and Wireless network equipment. The company offers

    worldwide technical and sales support, including a network of offices throughout NorthAmerica, Europe and Asia. The company is publicly traded on the NASDAQ Stock Market

    under the PMCS symbol and is included in the S&P 500 Index.

    About the Author

    Steve Gorshe, Ph.D. is a Principal Engineer in the Product Research Group and oversees ICs for

    SONET, optical transmission and access systems.

    Currently, Steve is a senior member of the IEEE and co-editor for the IEEE Communications

    magazines Broadband Access Series. He is the chief editor for the ATIS OPTXS Committee

    (formerly T1X1), which is responsible for SONET and optical network interface standards. Heis a recent recipient of the Committee T1 Alvin Lai Outstanding Achievement Award for his

    standards work and has been a technical editor for T1.105, T1.105.01, T1.105.02, and T1.105.07

    within the SONET standard series as well as the ITU-T G.7041, G.8011.1, G.7043, and G.8040

    recommendations. He has 26 patents issued or pending and several published papers.

  • 8/8/2019 Ethernet Over Transport

    4/36

  • 8/8/2019 Ethernet Over Transport

    5/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 5

    Table of Contents

    Legal Information...........................................................................................................................2

    Copyright................................................................................................................................. 2Disclaimer ...............................................................................................................................2

    Trademarks............................................................................................................................. 2

    Patents.................................................................................................................................... 2

    Abstract .........................................................................................................................................3

    About PMC-Sierra................................................................................................................... 3

    About the Author ..................................................................................................................... 3

    Revision History ......................................................................................................................4

    1 Preface .................................................................................................................................... 8

    2 Why Ethernet Over the Public WAN? ..................................................................................... 9

    3 Service Types and Characteristics........................................................................................ 12

    3.1 Ethernet Connection (EC) Defined ............................................................................. 12

    3.2 Ethernet Connection (EC) Attributes........................................................................... 15

    3.2.1 Network Connectivity ..................................................................................15

    3.2.2 Transfer Characteristics.............................................................................. 17

    3.2.3 Separation (Customer and Service Instance)............................................. 18

    3.2.4 Link Type.....................................................................................................18

    3.2.5 Connectivity Monitoring...............................................................................18

    3.2.6 Bandwidth Profile ........................................................................................ 18

    3.2.7 UNI List ....................................................................................................... 18

    3.2.8 Preservation................................................................................................ 19

    3.2.9 Survivability................................................................................................. 19

    3.3 Ethernet Private Line (EPL) Service ........................................................................... 19

    3.4 Ethernet Virtual Private Line (EVPL) Service.............................................................. 21

    3.5 Ethernet Private LAN (EPLAN) Service ...................................................................... 22

    3.6 Ethernet Virtual Private LAN (EVPLAN) Service ........................................................24

    4 Ethernet Client Interfaces .....................................................................................................254.1 Multiplexed Access......................................................................................................25

    4.2 VLAN Mapping ............................................................................................................ 26

    4.3 Bundling ...................................................................................................................... 27

    4.4 Bandwidth Profile ........................................................................................................ 27

    4.5 Layer 2 Control Protocol Processing ..........................................................................27

  • 8/8/2019 Ethernet Over Transport

    6/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 6

    4.6 Summary of UNI Service Attributes for Different Services.......................................... 27

    5 Protection and Restoration ................................................................................................... 29

    5.1 Service Protection or Restoration Provided by the Transport Network ...................... 29

    5.2 Service Restoration at Layer 2.................................................................................... 30

    6 Conclusion ............................................................................................................................ 31

    A Related Standards Activity .................................................................................................... 32

    B Definitions ............................................................................................................................. 34

    C References............................................................................................................................ 35

  • 8/8/2019 Ethernet Over Transport

    7/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 7

    List of Figures

    Figure 1 Illustration of Ethernet Service Areas (ITU-T Rec. G.8011)........................................ 13

    Figure 2 Illustration of Private and Virtual Private Connections................................................14Figure 3 Network Connectivity Examples..................................................................................16

    Figure 4 Network Portion of the Multipoint-to-Multipoint Topology (ITU-T Rec.G.8011) .................................................................................................................... 17

    Figure 5 EPL Service Illustration ............................................................................................... 19

    Figure 6 EPLAN Illustrations .....................................................................................................22

    Figure 7 Service Multiplexing Example ..................................................................................... 26

    List of Tables

    Table 1 Summary of the Types of Ethernet Services................................................................ 14

    Table 2 Ethernet Connection Service Attributes (Derived from ITU-T Rec.G.8011) .................................................................................................................... 15

    Table 3 EPL Connection Characteristics (Derived from ITU-T Rec. G.8011.1)........................ 20

    Table 4 EVPL Connection Characteristics ................................................................................ 21

    Table 5 EPLAN Expected Connection Characteristics ............................................................. 23

    Table 6 EVPLAN Expected Connection Characteristics........................................................... 24

    Table 7 UNI Service Attributes Common to All Services .......................................................... 25Table 8 Service-dependent UNI Service Attributes................................................................... 25

    Table 9 Summary of UNI Service Attributes for EPL and EVPL Services ................................ 28

    Table 10 Summary of Related Standards Activities.................................................................. 32

  • 8/8/2019 Ethernet Over Transport

    8/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 8

    1 Preface

    Due to a convergence of business requirements for higher data connectivity rates and flexible

    services, and the availability of new data transport technology, Ethernet has become thetechnology of choice for WAN connectivity both from the perspective of the enterprise and

    from the perspective of the carrier (network operator).

    This white paper examines the popularity of Ethernet over Transport and what factors

    enterprises and carriers need to consider. Included in the discussion is an overview of Ethernet

    services that are being considered for delivery over public WANs as well as the attributes and

    characteristics that must be defined in order to provide these services. This discussion is based

    on the Ethernet transport work done for the ITU-T Study Group 15 (SG15) and its series of new

    standards. (See Appendix A.)

    This paper also examines the Ethernet-based user network interfaces (UNIs) and network-to-

    network interfaces (NNIs) that are required for transport network equipment carrying Ethernetservices as well as Operations, Administration and Maintenance (OAM) capabilities. (Some

    OAM capability is already available in Ethernet services deployed over SONET/SDH (server

    networks), but additional capabilities are required at the Ethernet Client layer.) The paper

    concludes with a summary of some of the protection and restoration technologies that can be

    used to guarantee carrier-grade service reliability for Ethernet WAN services.

  • 8/8/2019 Ethernet Over Transport

    9/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 9

    2 Why Ethernet Over the Public WAN?

    On the enterprise side of the public WAN, Ethernet is the dominant technology. There are many

    reasons why there is growing interest in Ethernet WAN connection services:

    At the enterprise, where it is the dominant technology, network administrators are already

    familiar and comfortable with Ethernet.

    A number of applications, including voice and video, are already being encapsulated into

    Ethernet and run over enterprise network LANs.

    More and more companies have a number of remote offices that need to exchange data with

    the headquarters office or with each other. There can be significant advantages to placing

    data on common servers that can be accessed by all sites.

    The increased importance of the Internet for business applications has led to an increasing

    desire for incrementally higher WAN interface rates.

    It is desirable to have a single Internet firewall at one site that is accessed on the enterprise

    network by all sites.

    Most enterprise WAN data access today uses Frame Relay connections at rates of fractional

    DS1, full-rate DS1, fractional DS3, and full-rate DS3. In some cases, the data services use ATM

    instead of Frame Relay either for the customer connection or as a method of transporting the

    data through the core network. Frame Relay-based services suffer from a lack of scalability in

    terms of both high-speed connections and ubiquitous multipoint networks. ATM suffers from

    being somewhat provisioning-intensive in large networks and bandwidth inefficient.

    At the other end of the spectrum, some customers use WDM or dark fiber for their WAN or

    MAN connections with native Gigabit Ethernet, 10 Gbit Ethernet, or Fibre Channel. However,

    WDM equipment is expensive and not ubiquitously deployed, and dark fiber is not universally

    available, which limits the applications for these approaches. Another drawback of WDM and

    dark fiber applications is that unless they use the new G.709 Optical Transport Network

    standard [1], there is no embedded overhead for carriers to monitor the quality of the connection

    or provide protection switching in order to guarantee their service level agreements (SLAs).

    Some larger customers use high-speed SONET OC-N / SDH STM-N connections, which

    typically terminate the Ethernet frames and re-encapsulate the data into PPP, and use Packet

    over SONET/SDH (PoS) for transmission through a SONET/SDH pipe. On the carrier (network

    operator) side, the situation is somewhat complicated. In the wake of the telecom bubble burstin 2000, carriers face two clear drivers. First, they must deploy any new services on their

    existing SONET/SDH networks. Throughout the 1990s, carriers invested heavily in building

    SONET/SDH-based fiber optic networks, including developing the Operations, Administration,Maintenance and Provisioning (OAM&P) systems to run them. The enormous capital

    investment required to build an overlay network for data transport, be it native Ethernet orDWDM-based, makes overlay networks unrealistic. Fortunately, SONET/SDH has proven

    flexible enough to handle the challenge.

  • 8/8/2019 Ethernet Over Transport

    10/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 10

    Recent developments that extend the useful life of SONET/SDH include:

    Virtual concatenation (VCAT) [2][3] to create more flexible data channel sizes.

    Link Capacity Adjustment Scheme (LCAS) [4] to increase the flexibility and functionality

    of VCAT.

    Generic Framing Procedure (GFP) [5][6] to encapsulate data frames for robust and

    bandwidth-deterministic transport.

    Resilient Packet Ring (RPR) technology that provides a very flexible technology for data

    access onto SONET/SDH access and metro rings.

    A second driver for carriers is the desire to generate new, revenue-bearing services. With data

    traffic continuing to grow faster than voice, offering WAN connectivity with higher bandwidth

    and enhanced service capabilities appears to be a natural direction for new services. Equipment

    and device manufacturers, who are also interested in building new products, have also seen

    WAN connectivity as a natural direction.

    Ethernet is in many ways an obvious technology choice for WAN connectivity. The job of

    enterprise network administration is simplified if all their sites can be treated as part of the same

    Ethernet network. In addition, Ethernet interfaces are relatively inexpensive and are ubiquitous

    on enterprise equipment. The development of VCAT and GFP allows efficient transport of

    Ethernet frames through SONET/SDH networks. In addition, Ethernet bridge and router

    technology requires less provisioning and administration than technologies such as ATM. As

    some carriers begin a migration to using packet switching technology such as MPLS in their

    core networks rather than traditional circuit switching, Ethernet again looks like an excellent fit

    as an access technology.

    The IEEE, ITU-T Study Group 15 and 13, Metro Ethernet Forum (MEF), and Internet

    Engineering Task Force are working on developing Ethernet transport standards. (Refer toAppendix A.) This level of standards activity is a good indication of how many companies and

    organizations are interested in Ethernet WAN.

    OAM is critical once Ethernet is extended beyond the customer premise, especially when

    multiple transport service providers carry the traffic. In fact, although it is being actively

    addressed by multiple standards bodies, including the ITU-T SG13, IEEE, and the MEF,

    Ethernet currently lacks some of the OAM capabilities that will be required to monitor and

    guarantee service level agreements (SLAs). In a multiple carrier environment, for example,

    OAM is crucial for determining the locations of problems and degradations when they occur.

    From a transport network provider standpoint, this OAM requirement is an area where

    SONET/SDH really shines. The OAM capabilities inherent in the SONET/SDH backbone allow

    full monitoring and protection of the transmission facilities and transport path through theSONET/SDH network.

  • 8/8/2019 Ethernet Over Transport

    11/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 11

    Note that the relatively low cost of Ethernet enterprise equipment has created a customer

    expectation that Ethernet WAN interfaces should be less expensive than DS1/DS3 Frame Relay

    interfaces. For the carriers, even if the Ethernet connectivity is provided through their

    SONET/SDH backbone network, they still have to develop new OAM&P procedures and tools

    to provide the service. Therefore, even though customers and carriers agree that Ethernet (overSONET/SDH for the carriers) is the most attractive technology for high-speed WAN

    connectivity, there is an inherent conflict between carriers needing to invest in equipment and

    management system upgrades and customers demanding and expecting to pay less for Ethernet-

    based services.

  • 8/8/2019 Ethernet Over Transport

    12/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 12

    3 Service Types and Characteristics

    The description of Ethernet transport services varies depending on ones vantage point. This

    white paper approaches Ethernet transport from the perspective of a transport network orservice provider, rather than from a customer view of Ethernet transport services, and follows

    the approach of the ITU-T G.8011 recommendation [9] in its discussion of the service types andcharacteristics of Ethernet. (The G.8011.x series covers specific services within the framework

    of G.8011.)

    One can say that the goal of the network provider is to make Ethernet service look like an

    extension of the customers Ethernet LAN. In order to provide this transparency, the network

    provider must take into account a number of items that are not necessarily directly visible to the

    customer. These items are presented in this section in terms of Ethernet connection attributes

    and their associated parameters.

    3.1 Ethernet Connection (EC) Defined

    An Ethernet Virtual Connection (EVC), otherwise known as an Ethernet Connection (EC),

    provides a connection between two or more customer UNIs such that the Ethernet frames

    (service frames) associated with the EVC can only be transferred between its associated UNIs

    and not to any others. Note that to be consistent with ITU-T G.8011, this white paper uses the

    more generic term, EC, rather than EVC.

    Figure 1 illustrates an EC with its different reference points from the standpoint of both the

    Ethernet MAC layer network (ETH) and Ethernet physical layer network (ETY). Figure 1 also

    illustrates the three different Ethernet service areas in a multi-carrier Ethernet connection:

    Access (UNI-C to UNI-N)

    End-to-end/customer-to-customer (UNI-C to UNI-C)

    Edge-to-edge/intra-carrier (UNI-N to UNI-N)

  • 8/8/2019 Ethernet Over Transport

    13/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 13

    Figure 1 Illustration of Ethernet Service Areas (ITU-T Rec. G.8011)

    Operator FD Operator FD

    TransportTransportTransport

    OperatorOperatorOperatorBBBsss

    NetworkNetworkNetwork

    TransportTransportTransport

    OperatorOperatorOperatorAAAsss

    NetworkNetworkNetwork

    UNIAttributes Ethernet ConnectionAttributes

    demarc

    Customer Customer

    UNI-N UNI-NUNI-C UNI-C

    FD

    ETHETH

    demarc

    Access AccessUNI-N to UNI-N

    UNI-C to UNI-C

    NNIAttributes

    NNI

    FD

    ETYETY

    Operator FD Operator FD

    Notes

    1. FD = Flow Domain

    2. Diagram reprinted with permission from ITU-T Recommendation G.8011.

    From a customers perspective, the EC connectivity can be one of two types:

    Line connectivity (point-to-point)

    LAN connectivity (point-to-multipoint or multipoint-to-multipoint)

    From a transport network viewpoint, these line and LAN connections can either be provided:

    Through dedicated transport channels, including router bandwidth (private service)

    Through a shared medium (virtual private service)

    The difference between a private line connection and a virtual private line connection is

    illustrated in Figure 2. The service type variations are summarized in Table 1.

  • 8/8/2019 Ethernet Over Transport

    14/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 14

    Figure 2 Illustration of Private and Virtual Private Connections

    SONET/SDH/PDH/OTN

    Carrier Network

    Customer AEquipment

    Ethernet

    PHY

    CarrierEquipment

    CarrierEquipment

    Customer AEquipment

    Ethernet

    PHY

    Customer BEquipment

    Customer BEquipment

    SONET/SDH/PDH/OTH

    Carrier Network

    Customer AEquipment

    EthernetPHY

    CarrierEquipment

    CarrierEquipment

    Customer AEquipment

    EthernetPHY

    Customer BEquipment

    Customer BEquipment

    a) EPL for two customers, each with their own TDM channel

    b) EVPL for two customers where they share a TDM channel for increased

    efficiency

    Table 1 Summary of the Types of Ethernet Services

    Connectivity Resource Sharing Service Type

    Dedicated EPL (Ethernet Private Line)Point-to-point

    Shared EVPL (Ethernet Virtual Private Line)

    Dedicated EPLAN (Ethernet Private LAN)Multipoint

    Shared EVPLAN (Ethernet Virtual Private LAN)

    Note

    1. The Metro Ethernet Forum(MEF) refers to EPL and EVPL as E-Line services, and EPLAN andEVPLAN as E-LAN services.

  • 8/8/2019 Ethernet Over Transport

    15/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 15

    3.2 Ethernet Connection (EC) Attributes

    Various EC service attributes must be taken into account in a transport or service provider

    network. Table 2 lists these attributes.

    Table 2 Ethernet Connection Service Attributes (Derived from ITU-T Rec. G.8011)

    EC Service Attribute Service Attribute Parameters and Values

    Network connectivity Point-to-point, point-to-multipoint, multipoint-to-multipoint

    Address (deliver conditionally or unconditionally)

    Drop Precedence (drop randomly, drop conditionally, or not applicable)

    Transfer characteristics

    Class of Service

    CustomerSeparation

    Service instance

    Spatial or logical

    Link type Dedicated or shared

    Connectivity monitoring Sub-layer monitoring: On demand, proactive, none

    Inherent monitoring: Proactive

    Bandwidth profile Specified

    UNI list An arbitrary text string to uniquely identify the UNIs associated with the EC.

    VLAN ID (yes or no)Preservation

    Class of Service (yes or no)

    Survivability None, or server-specific

    3.2.1 Network Connectivity

    One way in which Ethernet services can be characterized is by the type of desired customer

    connectivity. The types of connectivity are:

    Point-to-point

    Point-to-multipoint

    Multipoint-to-multipoint

    Figure 3 shows examples of these different connectivity types. Figure 3 (a) shows a basic point-

    to-point connection between customers through a transport network. Figure 3 (b) shows a

    popular point-to-multipoint topology known as hub-and-spoke. Figure 3 (c) through (e) show

    examples of various multipoint-to-multipoint topologies.

    Care must be taken to avoid confusing the customer connectivity (logical topology) with the

    physical topology of the underlying network providing that connectivity. For example,

    Figure 3 (b) and Figure 3 (d) use the same physical topology. The difference between a hub-

    and-spoke and a star network is that a star network provides arbitrary multipoint-to-multipoint

    connectivity between all the customer nodes while a hub-and-spoke network connects the hub

    customer node to each of the spoke customer nodes (point-to-multipoint). Any connectivity

    between spoke nodes would have to be provided by a router at the customers hub node.

  • 8/8/2019 Ethernet Over Transport

    16/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 16

    A logical hub-and-spoke network could be provided over the physical topology of any of the

    networks in Figure 3 (b) through (e). Figure 3 (c) through (e) illustrate common transport

    network topologies. In reality, a transport network will often consist of a combination of star,

    ring, and more arbitrary mesh sub-networks.

    Figure 3 Network Connectivity Examples

    TRANSPORT

    NETWORKCE CE

    a) Point-to-pointTRANSPORT

    NETWORK

    CE

    CE

    CE

    CE

    CE

    b) Hub and spoke

    TRANSPORTNETWORK

    CE

    CE CE

    CE

    CE

    TRANSPORT

    NETWORK

    CE

    CE CE

    CE

    CE

    c) Ring

    d) StarTRANSPORT

    NETWORKCE

    CE CE

    CE

    CE

    e) Arbitrary

    CE = Customer Edge

    When discussing multipoint-to-multipoint connectivity, it is common to refer to the network

    topological componentthat provides the desired forwarding of information between source and

    destination UNIs as aflow domain (FD)1. In the example shown in Figure 4, if customers M and

    1 A topological component describes a portion of a layer network in terms of the topological relationship

    between the (flow) points.

  • 8/8/2019 Ethernet Over Transport

    17/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 17

    N are exchanging data, each is connected to a FD, with the flow domains being connected

    through an Ethernet link flow (LF) over the ABC network entity.

    Figure 4 Network Portion of the Multipoint-to-Multipoint Topology (ITU-T Rec. G.8011)

    ETH FD

    ETH FD

    ETH FD

    ETH LF

    ETH LF

    ETH LF

    ABCnetwork entity

    PQRnetwork entity

    XYZnetwork entity

    ETH UNI

    ETH UNI

    ETH UNI

    ABC, PQR, XYZ are server layer networks (can all be the same or different).They may be CO-CS, CO-PS, CLPS

    Customer M Customer N

    Note

    1. Diagram reprinted with permission from ITU-T Recommendation G.8011.

    A point-to-point connection can be characterized as either not having a flow domain or as

    having a flow domain with only twoflow points (that is, two end points of the networkconnection). The point-to-point connection is typically described as not having a flow domain

    since a flow domain implies a layer network with inherent switching/routing and other layer

    network capabilities. A flow domain with only two points is really the degenerate case of a

    multipoint network and leaves open the potential of adding additional flow points (UNIs) to the

    network.

    3.2.2 Transfer Characteristics

    The transfer characteristics of a network relate to what frames are to be delivered to the

    destination unconditionally, delivered conditionally, or are to be dropped. In the case of

    Ethernet, the three parameters that determine the disposition of a frame are:

    Address

    Drop Precedence (DP)

    Class of Service (CoS)

  • 8/8/2019 Ethernet Over Transport

    18/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 18

    For the address, a frame either can be delivered unconditionally, regardless of its destination

    address, or be delivered for only some destination addresses. The DP indicates the relative

    priority of the frame if it encounters a congestion situation in which frame dropping must occur.

    If dropping is based on the DP, frames are said to be dropped conditionally. Another option

    would be dropping randomly (that is, drop the overflow frames from a full queue). For someservices, frames cannot be dropped and hence DP is not applicable. The CoS parameter, which

    is based on the DP and indicates the frames class queuing, is not fully defined at this time.

    3.2.3 Separation (Customer and Service Instance)

    Separation refers to how the traffic of each customer or service instance is kept separate from

    the traffic of others. In the case of customer separation, the traffic from different customers is

    separated. In the case of service instance separation, the different service instances are

    separated, even for the same customer. A spatial separation implies a circuit switched network

    (for example, a TDM network in which each customer has its own TDM channel or facility, or a

    virtual circuit such as in an ATM network). Logical separation implies that customer or service

    instance traffic is separated at the packet level (that is, based on per-packet information such asaddress or tag values).

    3.2.4 Link Type

    A link can either be dedicated to a single customer service instance or shared among multiple

    service instances. For a dedicated link, a single customer service instance has a one-to-one

    mapping to a set of one or more Ethernet links and the associated server layer trail (that is, a

    spatial separation from other service instances). As such, the service instance does not compete

    for bandwidth/resources (transport and switch fabric bandwidth) with other service instances,

    and does not allow multiplexing on the access link (see Figure 2(a)). On the other hand, a

    shared link allows more than one service instance to share that link, (that is, logical separation)

    which means that the service instances can compete for the link resources (see Figure 2(b)).

    3.2.5 Connectivity Monitoring

    Connectivity monitoring is the mechanism by which network nodes determine their ongoing

    connectivity to their neighbor nodes in that layer network.

    3.2.6 Bandwidth Profile

    A bandwidth profile specifies that parameters that a traffic flow must meet at a UNI or NNI.

    Policing of the bandwidth profile is typically done at the edge of the transport network.

    3.2.7 UNI List

    For the purposes of management and control, a service provider assigns an arbitrary string to

    uniquely identify each UNI.

  • 8/8/2019 Ethernet Over Transport

    19/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 19

    3.2.8 Preservation

    Preservation refers to whether a customers Ethernet frame VLAN ID and/or CoS are preserved

    through the transport network. If the value is preserved, it will have the same value at the egress

    UNI that it had at the ingress UNI of the transport network. In some circumstances, however, itmay be necessary or desirable to change these values within the transport network. For example,

    the service provider may perform a translation between the VLAN ID values that a customer

    uses on the customer side of the network to a different set of VLAN IDs that are used within the

    service provider network. Another example is that if a frame fails to meet the specified

    bandwidth profile, the ingress node may choose to let it pass into the transport network but will

    set its DP value to a higher value so that it is prioritized for dropping if it encounters congestion.

    3.2.9 Survivability

    Survivability pertains to the ability of the network to continue to provide service under

    situations in which a fault(s) exists in the network.

    3.3 Ethernet Private Line (EPL) Service

    EPL, as illustrated in Figure 5 and Figure 2, consists of point-to-point Ethernet connections

    using reserved, dedicated bandwidth. With EPL, the transport network effectively looks like a

    piece of wire from the Ethernet client perspective. From the transport network provider

    standpoint, however, the transport network (server layer) provides the performance monitoring

    and protection capabilities required for guaranteeing the service level agreement (SLA) with thecustomer.

    Figure 5 EPL Service Illustration

    SONET/SDH/PDH/OTH(or ATM/MPLS CIR)

    Carrier NetworkCustomerEquipment

    EthernetPHY

    CarrierEquipment

    CarrierEquipment

    CustomerEquipment

    EthernetPHY

    The primary advantages of EPL are the simplicity of setting up a dedicated circuit, and the

    security for the traffic that is inherent when it is isolated in its own TDM channel2. Sharing

    bandwidth can lead to greater bandwidth efficiency in the transport network due to statistical

    multiplexing gains, however it is more difficult to administer since it requires additional effort(for example, traffic engineering and monitoring) in order to guarantee the customer SLA.

    2 It is also possible to provide EPL and EPLAN using constant rate ATM or MPLS virtual circuits instead

    of TDM circuits.

  • 8/8/2019 Ethernet Over Transport

    20/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 20

    EPL service is described in ITU-T G.8011.1 [10]. The EPL connection characteristics are

    summarized in Table 3.

    Table 3 EPL Connection Characteristics (Derived from ITU-T Rec. G.8011.1)

    EC Service Attribute Service Attribute Parameters and Values

    Network connectivity Point-to-point

    Address - deliver unconditionally

    Drop Precedence - not applicable

    Transfer characteristics

    Class of Service

    CustomerSeparation

    Service instance

    Spatial or logical (always connection oriented)

    Link type Dedicated

    Connectivity monitoring None, on-demand, or proactive

    Bandwidth profile Committed information rate (CIR) and committee burst size (CBS)

    UNI list An arbitrary text string to uniquely identify the UNIs associated with the EC.

    VLAN ID is preservedPreservation

    Class of Service is preserved

    Survivability None, or server-specific

    ITU-T G.8011.1 defines two types of EPL services:

    Type 1 EPL, where the characteristic information (CI) transferred between the UNIs is the

    Ethernet MAC frames. The Ethernet preamble, start of frame delimiters, and inter-frame

    characters are discarded, and the Ethernet MAC frames are then encapsulated (for example,

    into GFP-F) and mapped into the transport channel.

    Type 2 EPL, which is only defined for 1 Gbit/s Ethernet, treats the 8B/10B line code

    information as the CI to be transferred between the UNIs. The data and control code

    information from the Ethernet signals 8B/10B line code characters are translated into a

    more bandwidth-efficient 64B/65B block code, and multiple 64B/65B codes are then

    mapped into a GFP frame (GFP-T).

    The primary advantages of Type 2 EPL are the preservation of control codes (primitive

    sequences of special line code characters) and lower mapping latency. While is possible to

    also define Type 2 EPL for 4B/5B encoded 100 Mbit/s Ethernet, there has been no formal

    request for this service so far.

  • 8/8/2019 Ethernet Over Transport

    21/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 21

    3.4 Ethernet Virtual Private Line (EVPL) Service

    EVPL is also a line service; however, the line can be derived from a flow domain. It allows the

    sharing of network resources among multiple customers or service instances in order to achieve

    efficient use of those resources. EVPL is illustrated in Figure 2(b).

    In addition to increasing the efficient use of transport network resources, another potential

    advantage of EVPL is the reduction of the number of UNIs required at the customer edge. This

    is illustrated in Figure 7 in Section 4.1. For the customer edge node on the left to connect to four

    other nodes would require four different UNIs and their associated ports. Service multiplexing

    is the packet multiplexing of multiple ECs onto a single UNI.

    EVPL is specified in the ITU-T Recommendation G.8011.2. The EVPL connection

    characteristics are summarized in Table 4. For virtual connections, the separation is typically

    logical (that is, at the packet level). Due to the sharing of network resources, it is possible that

    frames will be dropped because of congestion. In addition, as discussed above, a service

    provider may wish to perform VLAN ID translation at the boundaries of the transport network.

    Table 4 EVPL Connection Characteristics

    EC Service Attribute Service Attribute Parameters and Values

    Network connectivity Point-to-point

    Address (deliver conditionally or unconditionally)

    Drop Precedence (drop randomly or drop conditionally, depending on theExcess Information Rate (EIR) and Committed Information Rate (CIR)parameters)

    Transfer characteristics

    Class of Service

    CustomerSeparation

    Service instance

    Logical or spatial

    Link type Shared or dedicated

    Connectivity monitoring Sub-layer monitoring: On demand and/or proactiveInherent monitoring: Proactive

    Bandwidth profile Specified Based on CIR and Committed Burst Size (CBS) (and in somecases also on EIR and Excess Burst Size (EBS))

    UNI list An arbitrary text string to uniquely identify the UNIs associated with the EC.

    VLAN ID (yes or no)Preservation

    Class of Service (yes or no)

    Survivability None, or server-specific

  • 8/8/2019 Ethernet Over Transport

    22/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 22

    3.5 Ethernet Private LAN (EPLAN) Service

    An EPLAN provides LAN-type connectivity between multiple customer sites through dedicated

    channels. Figure 6 illustrates some of the different basic transport network topologies that can

    support this service. From the customer viewpoint, these topologies are equivalent (that is, thecarrier network architecture is transparent to the customer).

    Figure 6 EPLAN Illustrations

    Carrier Network

    CustomerEquipment

    CustomerEquipment

    CustomerEquipment

    EthernetPHY

    EthernetPHY

    Ethernet

    PHY

    a) Mesh-type connectivity

    CarrierNetwork

    CustomerEquipment

    CustomerEquipment

    CustomerEquipment

    EthernetPHY

    EthernetPHY

    b) Traffic hauled to a centralized switch point(s)

    Carrier Network

    CustomerEquipment

    CustomerEquipment

    CustomerEquipment

    EthernetPHY

    EthernetPHY

    c) Edge node serves as a bridge or router

  • 8/8/2019 Ethernet Over Transport

    23/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 23

    In Options 1 and 3, the carrier does the switching at the edge of the network. Option 3 does the

    switching at one end of the network rather than at each end. In Option 2, the traffic is brought to

    a centralized switch (or a number of centralized switch points) in a star connection. Since the

    switching is performed at Layer 2 in these examples, an MSPP can be used to implement

    Options 1 and 3.

    Open issues to be resolved for an EPLAN standard include:

    How do the customer and carrier specify the bandwidth requirements? For example, if the

    traffic was evenly distributed among the different customer nodes, the bandwidth between

    nodes could be specified based on CIR. The more realistic scenario, however, is that

    multiple customer nodes will want to simultaneously communicate with a single node (for

    example, remote sites communicating with a headquarters office). A safe policy would be to

    reserve enough bandwidth for each node to simultaneously receive data at full rate fromeach other node; however, this would too inefficient to be practical.

    Closely related to the above issue, how much buffering must the carrier provide to handle

    congestion, and what will the discard policy be?

    Is protection handled at Layer 1 (for example, SONET APS) or Layer 2?

    Although EPLAN is still under study, its expected connection characteristics are summarized in

    Table 5.

    Table 5 EPLAN Expected Connection Characteristics

    EC Service Attribute Service Attribute Parameters and Values

    Network connectivity Multipoint-to-multipoint (and probably point-to-multipoint)

    Address - deliver unconditionally

    Drop Precedence (for further study)

    Transfer characteristics

    Class of Service

    CustomerSeparation

    Service instance

    Spatial or logical (always connection oriented)

    Link type Dedicated

    Connectivity monitoring None, on-demand, or proactive

    Bandwidth profile Committed information rate (CIR) and committed burst size (CBS)

    UNI list An arbitrary text string to uniquely identify the UNIs associated with the EC.

    VLAN ID is preservedPreservation

    Class of Service is preserved

    Survivability None, or server-specific

  • 8/8/2019 Ethernet Over Transport

    24/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 24

    3.6 Ethernet Virtual Private LAN (EVPLAN) Service

    EVPLAN is a combination of EVPL and EPLAN. The transport channel bandwidth is shared

    among different customers, as are the routers in the carrier network. Ultimately, the sharing of

    bandwidth in the transmission channels and switch fabrics gives EVPLAN the potential for verycost-effective carrier network resource utilization. Clearly, however, EVPLAN is the most

    complicated network architecture to administer.

    The open issues regarding EVPLAN transport architectures include all of those already

    discussed for EVPL and EPLAN; although, the magnitude of some of these issues is greatly

    increased for EVPLAN, which in turn restricts some of the potential solution space. For

    example, the tagging mechanism to differentiate the data from different customers, and the

    different data flows within each customer data stream, must have an adequately large address

    space. (For example, the 4K-address space of VLAN tags makes them impractical for large

    EVPLANs. Also, their applicability to only Ethernet frames further lessens their appeal for a

    generic data network.)

    Although EPLAN is still under study, its expected connection characteristics are summarized in

    Table 11-7.

    Table 6 EVPLAN Expected Connection Characteristics

    EC Service Attribute Service Attribute Parameters and Values

    Network connectivity Multipoint-to-multipoint (and probably point-to-multipoint)

    Address (deliver conditionally or unconditionally)

    Drop Precedence (drop randomly, drop conditionally, or not applicable)

    Transfer characteristics

    Class of Service

    CustomerSeparationService instance

    Logical

    Link type Shared

    Connectivity monitoring None, on-demand, or proactive

    Bandwidth profile Specified

    UNI list An arbitrary text string to uniquely identify the UNIs associated with the EC

    VLAN ID (yes or no)Preservation

    Class of Service (yes or no)

    Survivability None, or server-specific

  • 8/8/2019 Ethernet Over Transport

    25/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 25

    4 Ethernet Client Interfaces

    The client interface (that is, the UNI) for Ethernet services will typically be an Ethernet physical

    interface. In the past, the UNI for WAN data connections was typically based on telecom signals(for example, DS1, DS3, fractional-DS1, etc.), which required either special customer

    equipment (for example, a T1 multiplexer) or special WAN port cards on customer routers.

    Being able to use an Ethernet interface is advantageous for enterprise customers. Ethernet

    interfaces are typically less expensive than telecom WAN interfaces, especially for higher bit

    rates, and are supported by relatively inexpensive Ethernet LAN routers. Using an Ethernet

    interface also allows enterprise network administrators to keep their OAM in the Ethernet

    domain and allows carriers to handle the telecom signal domain.

    The Ethernet UNI service attributes that are common for all services are shown in Table 7 and

    the attributes that are service dependent are shown in Table 8.

    Table 7 UNI Service Attributes Common to All Services

    Layer UNI Service Attribute Service Attribute Parameters and Values

    MAC service IEEE 802.3 frame format

    UNI ID Arbitrary text string to identify each UNI instance

    ETH

    UNI EC ID Arbitrary text string to identify each EC instance

    PHY speed 10 Mbit/s, 100 Mbit/s, 1 Gbit/s, or 10 Gbit/sETY

    PHY medium IEEE 802.3 Physical interface

    Table 8 Service-dependent UNI Service Attributes

    Layer UNI Service Attribute Service Attribute Parameters and Values

    Multiplexed access Yes, no

    VLAN mapping Specify

    Bundling Yes, no, all-to-one

    Bandwidth profile For further study for most services

    ETH

    Layer 2 controlprotocol processing

    Block, process, or pass per protocol on ingressGenerate, or none per protocol on egress

    ETY PHY mode Full duplex, half duplex, or auto-negotiation

    The UNI service attributes common to all services are reasonably self-explanatory so are not

    discussed in this paper. The service-dependent attributes are explained as follows.

    4.1 Multiplexed Access

    This aspect of the UNI relates to whether a customer edge node has an individual UNI

    associated with each of the far-end UNIs to which it is connected, or whether a UNI is shared

    for the connections to multiple other customer UNIs. These cases are illustrated in Figure 7.

  • 8/8/2019 Ethernet Over Transport

    26/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 26

    Figure 7 Service Multiplexing Example

    TRANSPORT

    NETWORK

    CE

    CE

    CE

    CE

    CE

    4 x UNI

    TRANSPORT

    NETWORK

    CE

    CE

    CE

    CE

    CE

    UNI

    a) Multiple point-to-point EPL

    (one EVC per UNI)

    b) Service multiplexed UNI(e.g., for multiple EVPL)

    4.2 VLAN Mapping

    Ethernet (per IEEE 802.1Q) allows the insertion of tags into Ethernet MAC frames in order to

    create virtual LANs (VLANs). When a customer desires to preserve this VLAN segregation of

    traffic through the WAN, the carrier may simply preserve the entire MAC frame, including the

    VLAN tag.

    It is possible that both the customer and service provider desire to use VLAN technology. If the

    service provider also wishes to use VLAN technology, then it must either must insert a second(stacked) VLAN tag into the MAC frame, or perform a translation of the customer VLAN tag

    at the ingress in order to conform to the service providers VLAN tag assignments, and then

    restore the customer VLAN value at the egress.

  • 8/8/2019 Ethernet Over Transport

    27/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 27

    4.3 Bundling

    Bundling refers to whether multiple customer VLAN IDs can be mapped into a single EC at a

    UNI. The case where all VLAN IDs are mapped to a single EC is called all-to-one bundling. It

    should be noted that all-to-one bundling and multiplexed access are mutually exclusive.Multiplexed access refers to the multiplexing of multiple ECs into a UNI. Mapping all VLAN

    IDs into a single EC means that there is a single EC at that UNI rather than multiple ECs.

    4.4 Bandwidth Profile

    The bandwidth profile refers to the characteristics of the traffic that a customer presents to the

    network at the UNI. The parameters include such things as guaranteed committed information

    rate (CIR), peak information rate (PIR), burst sizes, and what is done with excessive traffic.

    4.5 Layer 2 Control Protocol Processing

    Layer 2 Control Protocols are used by enterprise network bridges for a variety of functions.

    Depending on the protocol application and the WAN service, these protocols can either be

    passed transparently through the WAN, processed (with the network provider equipment acting

    as a peer to the customer equipment), or discarded at the UNI.

    4.6 Summary of UNI Service Attributes for Different Services

    The UNI service attributes are currently only defined for EPL service (ITU-T G.8011.1). These

    attributes are summarized in Table 9. EVPL and EVPLAN (especially EVPLAN) will typically

    make use of bundling and multiplexed access and have more complex bandwidth profiles.

  • 8/8/2019 Ethernet Over Transport

    28/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 28

    Table 9 Summary of UNI Service Attributes for EPL and EVPL Services

    Layer UNI Service Attribute Service Attribute Parameters and Values

    EPL

    Multiplexed access NoVLAN mapping No

    Bundling All-to-one

    Bandwidth profile CIR, Committed Burst Size (CBS)

    ETH

    Layer 2 control protocolprocessing

    All are passed except PAUSE frames, which are discarded.(The network equipment providing the UNI-N may generatePAUSE frames.)

    ETY PHY mode Full duplex

    EVPL

    Multiplexed access Yes or No

    VLAN mapping Yes or No

    Bundling Yes or All-to-oneBandwidth profile For further study

    ETH

    Layer 2 control protocolprocessing

    Specified in G.8011.2

    ETY PHY mode Full duplex

  • 8/8/2019 Ethernet Over Transport

    29/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 29

    5 Protection and Restoration

    For WAN transport, service protection is typically provided by the underlying transport

    network. However, service restoration can also be provided only at the Ethernet layer usingEthernet protocols. Both approaches are summarized briefly here.

    5.1 Service Protection or Restoration Provided by the TransportNetwork

    Transport networks have traditionally been engineered for ultra-high service availability in

    order to ensure reliable communications for emergency voice traffic. The objective is for a fault

    to be detected and the service restored through a protection switching action within 60 ms

    (10 ms for the detection and 50 ms to complete the switch). A protection switch action is

    defined as the routing of the traffic from the failed facility or equipment to backup facilities or

    equipment that has been reserved in advance3. In most cases, the protection channels arepermanently reserved, with their bandwidth either being unused during fault-free conditions, or

    carrying low-priority traffic that can be dropped if the channel is required for protection. In the

    case of TDM transport systems such as SONET/SDH, the protection switch is between TDM

    channels or fibers [15]. In the case of ATM, the protection switch is between virtual circuits

    [16].

    The Link Capacity Adjustment Scheme (LCAS) provides an alternative approach for restoring

    data services. For constant bit rate services such as voice, it is necessary to have protection

    bandwidth that is at least equal to the service bandwidth. In the case of packetized data,

    however, it is possible in many cases to maintain the service connection at a lower data rate inthe event of a network problem. LCAS provides this capability through its control of virtually

    concatenated SONET/SDH, asynchronous/PDH, or OTH channels.

    Virtual concatenation is the construction of a larger channel through the combining of multiple

    smaller member channels. The data is then interleaved onto the constituent member channels in

    a byte-by-byte round-robin basis. The individual members can take different routes through the

    transport network with the network element (NE) that terminates the virtually concatenated

    channel providing the buffering to compensate for the differences in delay that result from the

    different routes. 4

    3 In mesh networks that are constructed with digital cross connect systems (DCS), it is possible to have

    the DCSs restore the traffic by communicating among each other to determine the protection path through

    the network in response to a fault. While this approach typically allows the network operator to reserve

    much less overall network bandwidth for protection, it can be difficult to meet the 50 ms switch time.

    4 Virtual concatenation is described in S. Gorshe, A Tutorial on SONET/SDH, PMC-Sierra white paper

    PMC-2030895.

  • 8/8/2019 Ethernet Over Transport

    30/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 30

    In the event that a subset of the members fail, the LCAS provides the signaling mechanism

    between the virtually concatenated channels source and sink to allow the source to place the

    traffic onto only those members that have not failed. This allows for a very powerful new

    paradigm for network operators since, for services that can tolerate the reduced data rate during

    faulted conditions, the operator can now provide very fast service restoration without having toreserve dedicated protection channels.

    5.2 Service Restoration at Layer 2

    When constructing a complex Ethernet network, it is important that only a single switching path

    exist between any pair of nodes. Otherwise, loops would exist that cause instability in

    forwarding tables and would cause broadcast data to be unnecessarily multiplied (that is,

    broadcast storms).

    The Ethernet solution to this issue is the spanning tree protocol (STP). The STP sends Layer 2

    control protocol messages between the Ethernet nodes in order to establish an overlaid logicaltree structure on the network. Messages are periodically sent between adjacent nodes to confirmthat no changes have occurred to the network connectivity. When a fault occurs, the node

    detecting the failure (that is, the change in network connectivity) will initiate a new spanning

    tree construction. The network is then unavailable for the 30 to 60 seconds that it can take for

    the STP to resolve. A new rapid STP (RSTP) protocol has been developed, but it still takes

    seconds to resolve.

    While this is a very powerful and flexible service restoration tool, there are several issues when

    it is applied to Ethernet WAN services. Of course, the larger the network in terms of number of

    nodes and geographical reach, the slower the STP resolution. Another potential issue occurs if

    the network operator uses Ethernet routing technology within its network. Here, conflicts

    between customer and network operator STPs must be avoided. Another issue is the interactionof the 50 ms protection switching (for example, SONET/SDH) with the STP. If the Ethernet

    node and the transport network node both detect the fault, it would cause unnecessary extended

    network unavailability if the STP were initiated for a fault that will be quickly protected by the

    SONET protection switch. It is desirable, then for the Ethernet node to wait at least 50 ms to see

    if the problem clears before it initiates the STP.

  • 8/8/2019 Ethernet Over Transport

    31/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 31

    6 Conclusion

    A convergence of technology and applications has increased the importance and value of wide-

    area Ethernet data transport services. This white paper has introduced some of the newtechnology and standards in development that will enable carriers to provide these services.

    This includes a discussion of Ethernets connection services, client interfaces, and protectionand restoration abilities.

    What makes the development of Ethernet transport service attractive is its evolutionary

    approach. It builds on the customers capital investment in Ethernet technology, and the

    transport providers existing SONET/SDH backbone networks and OAM procedures. All of the

    pieces are currently in place for offering EPL services. The tools and standards to provide the

    more complicated ELAN, EVPL, and EVPLAN services are currently being developed by the

    various standards bodies.

    While some proprietary data transport solutions exist, carriers require standards-based solutions,which Ethernet data transport provides, for any significant deployment of new services. The use

    of standards helps guarantee that the solution will be available from multiple vendors, will be

    stable and supported for many years, and ideally will be less expensive due to economies of

    scale. While each carrier will probably want to offer variations on the basic service types in

    order to differentiate themselves from their competitors, the services discussed in this paper

    provide the framework from which they can build.

  • 8/8/2019 Ethernet Over Transport

    32/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 32

    A Related Standards Activity

    The information in this white paper primarily focuses on the output of the Ethernet transport

    work in ITU-T Study Group 15 (SG15) and its series of new standards. SG15 is the ITU-T bodyresponsible, among other things, for standards related to public transport networks.

    The current level of standards activity is a good indication of how many companies and

    organizations see Ethernet WAN as the next key step both for Ethernet and for the public

    transport network providers (that is, carriers). The major standards activities and their status are

    summarized in Table 10.

    Table 10 Summary of Related Standards Activities

    Organization Activities Status

    IEEE

    802.3ae 10 Gbit Ethernet, which includes a WAN PHY interface to simplifyinterfacing to a SONET/SDH or G.709 OTN network

    Approved

    802.17 Resilient Packet Rings: Working on a ring-based network foraccess and metro applications

    Approved

    802.3ah(EFM)

    Ethernet in the First Mile, where work includes OAM aspects forEthernet Links (especially access links)

    Approved

    802.1 ad Provider Bridge specification This is the Q-in-Q standard. In Progress

    802.1ag Connectivity Fault Management, or Ethernet Service OAM In Progress

    802.1ae MAC Security (MacSec), including authentication, authorizationand encryption

    In progress

    ITU-T SG15(With input from ANSI T1X1)

    G.8011.1(Q11)

    Ethernet Private line service Approved

    G.8011.2(Q11)

    Ethernet Virtual Private line service Approved

    G.8012(Q11)

    Ethernet UNI and Ethernet Transport NNI Approved

    G.8010(Q12)

    Ethernet Layer Network Architecture, which is largely to translatethe IEEE 802 network material into ITU-T transport networkterminology and models

    Approved

    G.8011(Q11)

    Ethernet over Transport Ethernet Service Characteristics Approved

    G.esm

    (Q12)

    Ethernet over Transport Ethernet Service Multiplexing, which

    covers the multiplexing protocol(s) required to implement EVPLand EVPLAN

    In Progress

    G.8021[7](Q9)

    Characteristics of Ethernet transport network equipment functionalblocks

    Approved2004 (1stphase)

    Y.ethps

    (Q3)

    Ethernet protection switching In progress

    Q2 Studying Ethernet OAM aspects relating to access In progress

  • 8/8/2019 Ethernet Over Transport

    33/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 33

    Organization Activities Status

    ITU-T SG13

    Y.1730 Requirements for OAM functions in Ethernet-based networks andEthernet services

    Approved

    Y.ethoam

    (Q3)

    End-to-end and edge-to-edge aspects of Ethernet OAM includingPM

    In progress

    Metro Ethernet Forum (MEF)

    The MEF is studying various aspects of Ethernet MANs, including Ethernet architecture, service model,service definition, traffic management, UNI and NNI definition and OAM. The MEF work is covering allpossible OAM flows, such as end-to-end, edge-to-edge, access, inter-provider, intra-provider, etc.

    MEF1 Ethernet Services Model, Phase 1 Approved

    MEF2 Requirements and Framework for Ethernet Service Protection inMetro Ethernet Networks

    Approved

    MEF3 Circuit Emulation Service Definitions, Framework andRequirements in Metro Ethernet Networks

    Approved

    MEF4 Metro Ethernet Network Architecture Framework - Part 1: GenericFramework Approved

    MEF5 Traffic Management Specification: Phase I Approved

    UNI Type 1 Specification of UNI, data-plane aspects In progress

    UNI Type 2 Specification of UNI, control-plane aspects (ELMI) In progress

    EMS-NMS MIBS for Ethernet and network management In progress

    MEF Architecturepart 2

    Specifies functional elements of Ethernet trail, such as adaptation,conditioning, etc.

    In progress

    CES PDH Implementation agreement of PDH CES over Ethernet (includesboth AAL1 and Raw method)

    In progress

    Internet Engineering Task Force (IETF)

    PWE3 WG Working on defining an Ethernet transport over IP/MPLS usingMartini drafts (this is mainly EVPL service using UDP, L2TP orMPLS as multiplexing layer)

    In progress

    PPVPN WG Requirements for Virtual Private LAN Services (VPLS) In progress

    L2VPN WG Working on framework and service requirements of Ethernet-basedVPN, and defining EVPLAN service using IP/MPLS

    In progress

    Note that each standards organization has its own areas of expertise. The majority of the

    standards that will be required for the public transport network are being developed in the Q12

    and Q11 groups of ITU-T SG15. This work was partitioned not only logically by topic, but also

    in a manner that allowed for the earliest possible approval of useful standards/

    recommendations. The initial set of standards was approved in mid-2004 (see Table 10).

    ITU-T SG15 has established liaison contact with the other standards organizations and forums

    where their input is required or desired. For example, the G.ethsrv work is expected to use a

    considerable amount of input from the MEF regarding the definition of services.

  • 8/8/2019 Ethernet Over Transport

    34/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 34

    B Definitions

    The source of these definitions is the G.809 ITU-T recommendation.

    Term Definition

    Access group A group of co-located "flow termination" functions that are attached to thesame "flow domain" or flow point pool link".

    Characteristic Information(CI)

    A signal with a specific format, which is transferred on "flows". The specificformats are defined in technology-specific standards.

    Flow An aggregation of one or more traffic units with an element of commonrouting.

    Flow domain A topological component used to effect forwarding of specific characteristicinformation.

    Flow point A "reference point" that represents a point of transfer for traffic unitsbetween topological components.

    Flow point pool A group of co-located flow points that have a common routing.Flow point pool link A "topological component" which describes a fixed relationship between a

    "flow domain" or "access group" and another "flow domain" or "accessgroup".

    Flow termination A "transport processing function". There are two types of flow termination, aflow termination sink and a flow termination source.

    Link flow A "transport entity" that transfers information between "ports" across a flowpoint pool link.

    Topological component An architectural component, used to describe the transport network in termsof the topological relationships between sets of points within the same layernetwork.

  • 8/8/2019 Ethernet Over Transport

    35/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Document No.: PMC-2050704, Issue 1 35

    C References

    [1] ITU-T Recommendation G.709 (2001), Interfaces for the optical transport network

    (OTN).

    [2] ITU-T Recommendation G.707 (2003), Network node interface for the Synchronous

    Digital Hierarchy (SDH).

    [3] ITU-T Recommendation G.7043/Y.1343 (2004), Virtual concatenation of Plesiochronous

    Digital Hierarchy (PDH) signals.

    [4] ITU-T Recommendation G.7042 (2004), Link Capacity Adjustment Scheme (LCAS) for

    Virtual Concatenated signals.

    [5] ITU-T Recommendation G.7041/Y.1303 (2003) - The Generic Framing Procedure (GFP).

    [6] ITU-T Recommendation G.8040 (2004), GFP frame mapping into Plesiochronous Digital

    Hierarchy (PDH).

    [7] ITU-T Recommendation G.8021/Y.1341 (2004), Characteristics of Ethernet transport

    network equipment functional blocks.

    [8] ITU-T Recommendation G.809 (2003), Functional architecture of connectionless layer

    networks.

    [9] ITU-T Recommendation G.8011/Y.1307 (2004), Ethernet Services Framework.

    [10] ITU-T Recommendation G.8011.1/Y.1307.1 (2004) Ethernet Private Line Service.

    [11] ITU-T Recommendation G.8011.2/Y.1307.2 (2005) Ethernet Virtual Private Line Service

    [12] ITU-T Recommendation G.8012/Y.1308 (2004), Ethernet UNI and Ethernet NNI.

    [13] ITU-T Recommendation G.805 (2001), Generic functional architecture of transport

    networks.

    [14] ITU-T Recommendation G.8010/Y.1306 (2004), Ethernet Layer Network Architecture.

    [15] ITU-T Recommendation G.808.1 (2003), Generic Protection Switching Linear Trail and

    Subnetwork Protection.

    [16] ITU-T Recommendation I.630 (2000) ATM Protection Switching.

  • 8/8/2019 Ethernet Over Transport

    36/36

    Dow

    nloa

    ded[c

    ontr

    olle

    d]byAno

    uarL

    ofN

    oon

    Thur

    sday

    ,15

    April

    ,201

    004:4

    7:35

    PM

    Ethernet over TransportWhite Paper

    Contacting PMC-Sierra

    PMC-Sierra

    8555 Baxter PlaceBurnaby, BC

    Canada V5A 4V7

    Tel: +1 (604) 415-6000

    Fax: +1 (604) 415-6200

    Document Information: [email protected]

    Corporate Information: [email protected]

    Technical Support: [email protected]

    Web Site: http://www.pmc-sierra.com

    mailto:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]:[email protected]://www.pmc-sierra.com/http://www.pmc-sierra.com/http://www.pmc-sierra.com/mailto:[email protected]:[email protected]:[email protected]

Recommended