Carrier Ethernet Transport in Metro and Core Networks

Carrier Ethernet Transport in Metro and Core Networks Tutorial by Claus G. Gruber and Achim Autenrieth Nokia Siemens Networks 13th International Telecommunications Network Strategy and Planning Symposium - „Convergence in Progress” Networks 2008 September 28 – October 2, 2008 Budapest, Hungary

1

© Nokia Siemens Networks

About Us Dr.-Ing. Claus G. Gruber Claus Gruber is senior consultant and project manager at Nokia Siemens Networks, Munich, Germany. Division: Research Technology and Platforms, Network Technology, Network Control and Transport (RTP NT NCT). His main area of research focuses on next generation packet network architectures including Carrier Grade Ethernet and IP/MPLS over WDM. He is mainly interested in networking concepts, total cost of ownership, multilayer traffic engineering and resilience, control plane, and network management and configuration of ubiquitous communication technologies. Prior to his work at Nokia Siemens Networks he was a member of the research and teaching staff at Technische Universität München (TUM), Germany, where he received his Dr.-Ing. and Dipl.-Ing. degree in electrical engineering and information technology. Claus published about 30 articles in journals and conference proceedings and submitted about 20 invention reports in the area of routing, resilience, network planning, optimization and management that are currently under review at EU and US patent offices.

Dr.-Ing. Achim Autenrieth Achim Autenrieth is Head of IP Transport R&D Management Innovation (IPT RD Innovation) at Nokia Siemens Networks, Munich, Germany. Focus areas of his work are multilayer transport networks (OTN/DWDM, SDH/SONET, Ethernet/MPLS-TP, IP/MPLS), control plane protocols (ASON/GMPLS), network architecture evaluation, multilayer resilience and multilayer network design, routing and grooming. Prior to his current responsibility he was working as project manager and senior research scientist in internal innovation projects and funded research projects at Siemens AG, Corporate Technology and Siemens AG, Fixed Networks. Achim studied Electrical Engineering and Information Technology at the Technische Universität München (TUM) and received his Dipl.-Ing. and Dr.-Ing. degree in 1996 and 2003, respectively. From 1996 to 2003 he was member of the research and teaching staff at the Institute of Communication Networks at TUM.

2


Networks 2008 - Carrier Ethernet Transport in Metro and Core Networks Claus G. Gruber, Achim Autenrieth, {claus.gruber,achim.autenrieth}@nsn.com

2008/09/29

General Information • Schedule • Q&A • 9:00 – 10:30 Tutorial Part I • After each main section • 10:30 – 11:00 Coffee Break • 11:00 – 12:30 Tutorial Part II • To ensure proper knowledge transfer to the audience, some basic behavior rules should be strictly obeyed during the tutorial

3



2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 4



2008/09/29

Networks get run over by a huge traffic growth Technology innovation is a must on the way forward • The fastest and most cost efficient access technologies are not sufficient on their own 5 billion people connected

• Huge traffic volumes have to be transported throughout the network

• Data super highways and an optimized end-to-end transport are needed to connect 5bn people

5



2008/09/29

Challenges and Opportunities Reinventing the connected world

Add value beyond bit-pipe

100x traffic growth

User service experience

Internet for the next billion

5 Bn people connected Environmental Performance

6



2008/09/29

Tomorrow's communication world 5 Bn People connected Main growth in mobile subscriptions from new growth markets

Mobile Users Worldwide 5 Bn 4 Bn 3 Bn

Voice and high-speed Internet enabled (EDGE, HSPA, ... , LTE, WiMAX)

2 Bn 2 Bn

4 Bn mobile users Majority can be always online via mobile high-speed Internet access technologies

2 Bn fixed broadband users Wireline Broadband will facilitate usage of applications like TV and/or video streaming.

7


Voice and low-speed Internet enabled 2005

2010

2015

Fixed Broadband Subscriptions* Worldwide 0.8 Bn

0.6 Bn

xDSL

0.4 Bn

FTTx

0.2 Bn

cable 2005

2010

fixed WiMAX 2015

* Broadband subscriptions are typically shared by 2-3 people

Source: Nokia Siemens Networks estimations based external forecasts (Ovum, Strategy Analytics)


2008/09/29

Operators invest into the whole network CAGR 7,7%

• Optical Metro

Consumers Quality of life for citizens

• Rural connectivity • Photonic core

Enable next generation of connectivity

2007

Business Growth and efficiency

Government Productivity

2011

Transport investment worldwide

Broadband enabled network

Cost of data transport must go down Traffic

• Higher network efficiency One technology • Leased Line OPEX Profitable self built

Voice Dominant

8


Revenues

Data Dominant


2008/09/29

Source: Connectivity Scorecard

Broadband services drive transport network evolution

Demand for fixed broadband will increase over the next years Million subscriptions (world)

600

Total Broadband Access Market World [bn €]

500

total

400

6.4%

5.0%

300 200

5,0 1,0

5,3 1,5

5,6 1,8

5,9 2,1

6,2 2,3

2008

IPTV/VoD Subscriptions Cablemodem Subscriptions Total Broadband Subscriptions

2010

Fiber to the building/home subscription DSL Subscriptions

• In the year 2012, there will be more than 500 million Broadband subscribers worldwide

• Most subscribers will use a DSL connection • Fiber access subscription is expected to grow in

9


DSLAM

2,9

2,8

2,9

3

3,1

1,1

1

0,9

0,9

0,8

2006

2007

2008

2009

2010

2012

Source: internal research based on several analyst forecasts

line with IPTV subscription

Fiber access

Narrowband

100

2006

5 billion people connected

Source: internal research based on several analyst forecasts

• DSL is the dominant broadband market and will remain

• Driven by high bandwidth demand, fiber based access revenue will double in the next 10 years

• Narrowband revenue will decrease


2008/09/29

“100x traffic growth within 5 years” means a growing need for scalable networks Growing # of customers Consumer

Growing # of services Consumer

New services at lower cost Business

10



2008/09/29

Multimedia services drive bandwidth requirements Triple Play services require bandwidth from 25 to 100 Mbit/s per user!

IPTV: 20-30 Mbps (multiroom HDTV, VoD)

Internet: 5-10 Mbps VoIP: 0.1 Mbps

Source: Internet research, 2006 11



2008/09/29

Video and TV services as main driver „ Video Services will drive exponential growth in residential wireline traffic, ... with most growth from IPTV” 70 ExaByte 60 ExaByte

1600%

US residential Wireline Video related Traffic

50 ExaByte

1100%

40 ExaByte TV Services (unicast&broadcast)

30 ExaByte 530%

20 ExaByte

220% 230%

10 ExaByte 0

200%

100% 100%

2007

1500%

Streaming Video Clips

2011

* excluding P2P video and music exchange which dominate currently the Internet traffic

820%

2008

2009

2010

Source: Heavy Reading, June 2007, Internet TV, OTT Video & Future of IPTV

12


P2P Video*


2008/09/29

Increasing bandwidth demands require a simplified and more efficient infrastructure Operators go Ethernet

FT, Telefonica: “IP does not scale enough, Ethernet is an alternative”

13


Up to 100Gbit/s channels in the core Source: Conferences; Lightreading 2007

Level 3: “Ethernet is becoming a preferred enabler for leading applications, e.g. Internet, Content delivery, utility services, IP video, …”

Technology goes highest scalability and flexibility

Flexible Gigabit services & multi-Gigabit wavelength switching Ethernet switching @ all transport technologies Microwave Radio, NG SDH, DWDM, Carrier Ethernet


2008/09/29

The broadband telecommunication environment is enabled by next generation connectivity Carrier Ethernet Transport Megabit applications Gigabit services

14

Flexible bandwidths from access to core

Broadband access everywhere

Optimized connectivity in fixed and mobile environment

Reliable and secure traffic control

Solutions to balance networks and ensure Quality of Service



2008/09/29




2008/09/29

What is “Carrier Ethernet Transport” ?

In a sentence • Ethernet with Carrier Grade qualities for Transport Networks

But seriously… • Taking the simple, well known and widely deployed Ethernet service and extending it to the metro and core of public networks thus maintaining the simplicity, flexibility and cost effectiveness of the protocol and components on an end-to-end basis

16



2008/09/29

Carrier Ethernet Transport technology is defined by six key attributes Resiliency

End to End Ethernet • Seamless Ethernet across portfolio of IP Transport/Nokia Siemens Network • Differentiated service creation

Optimized Deployment • Scalable architecture with end to end portfolio • Technology agnostic multilayer optimization

17


• Connection Oriented Ethernet • 50ms protection • Resilient IP (ResIP) certification

Scalability • Standardized platforms • Prove worldwide deployment • Over 20,000 service and support personnel


Simple Management • Automation of network • Point and click provisioning • Standard Operation and Maintenance

Flexible Solutions • Integrated Solution for Mobile Backhaul, Business and residential services • Shared best practices

2008/09/29

Carrier Ethernet Transport – Defined Architecture Goals and Building Blocks Enable IP Services over a Converged Carrier Class Transport Architecture Add Scalability, Resiliency, and Manageability to Ethernet Packet Based

Connection Oriented

Service Transparent

Deterministic

Controlled

Multi-Service Convergence

Static Managed

Isolated Secure

Predictable Protected

Guaranteed SLA

Point-and-Click Provisioning Carrier Grade OAM

18


Stratum Quality Sync

High Reliability

High Scalability


Hard QoS Integrated TDM

2008/09/29

Carrier Ethernet Transport – Defined Fundamental Requirements Unified Architecture for Cost-Effective Transport of High-Speed Packet Services Ethernet Economics

• • • • •

19

Scalable Multi-Service Single UNI Synchronous Cost-Effective

Connection Oriented

• Provisioned • Deterministic • Predictable


L3 Service Transparency

• L2 Client Encapsulation • Secure Transport • L3 Proxy

Guaranteed SLA’s

Carrier Class Resiliency

Multi-Layer Service Management

• Provisioned • Strict QoS • Connection Admission Control

• NE Quality • SW Stability • Network Protection

• • • • •


2008/09/29

End-to-End Pt-and-Click Control Plane Robust OAM Reporting




2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics • Network Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 21



2008/09/29

Going Back to Where It Began • We have to go back to 1984

22



2008/09/29

Network Hierarchy Concept The OSI Reference Model The concept of layers • It is a simple and efficient way of communication n+1

Layer Provides services to higher layers with standardized interfaces Layer

n

Uses services of lower layers n-1

23

Layer


with standardized interfaces


2008/09/29

Network Hierarchy Concept The OSI Reference Model

7

Application

6 Presentation

24

The OSI reference model provides: • Standardized interfaces (compatibility, interoperability and competition) • Simplifies network technology development considerably (just trust and use the functionality of the lower layer)

5

Session

4

Transport

3

Network

Why seven layers?

2

Data Link

• Is an often discussed question (e.g. “Three layer approach of Future Internet projects)

1

Physical



2008/09/29

Network Hierarchy Concept The OSI Reference Model

7

Application

6 Presentation

25

5

Session

4

Transport

3

Network

2

Data Link

1

Physical


What we call application e.g. email client such as Thunderbird

9

The User

8

Real Application

Network service part of applications Provides network services to applications (e.g. protocols to applications such as snmp)

Data presentation Presents data in the right format to the application layer (includes encryption, reformating, restructuring of data)

Interapplication communication Maintains sessions between applications

End-to-end connection Ensures data transport reliability, information flow (includes maintaining of virtual circuits between hosts)

Data delivery Provides routes between two host systems (might be at different locations) (includes network discovery and routing decision)

Access to media Defines the data format and how the access to the media is controlled (includes bit-error correction)

Binary transmission on a physical link Electrical, mechanical, procedural, and functional specification


2008/09/29

Network Hierarchy Concept Data Encapsulation

7

Application

Header

Data

Header

Header

Data

Header

Header

Header

Data

Header

Header

Header

Header

Data

Header

Header

Header

Header

Header

Data

Header

Header

Header

Header

Header

Header

Data

Header

Header

Header

Header

Header

Header

Data

6 Presentation

26

5

Session

4

Transport

3

Network

2

Data Link

1

Physical


Header


2008/09/29

Network Hierarchy Concept Communication End System 2

End System 1

27

7

G

7

G

6

F

6

F

5

E

4

D

3

C

Only instances of the same layer can talk to each other!

5

E

4

D

3

C

3

2

B1

2

B1

B2

2

B2

1

A1

1

A1

A2

1

A2


Intermediate System (IS)

C


2008/09/29

Sample OSI Layer Protocols and Services

Specification

Protocols

Services

OSI Layer

28

7 Application 6 Presentation

End System

Transit System

Services e.g.: Services e.g.: FTP, HTTP, FTP, HTTP, Telnet Telnet Services e.g.: Services e.g.: MIDI, HTML, GIF MIDI, HTML, GIF JPG, ASCII JPG, ASCII

5 Session

e.g. Security (Firewall, Proxy)

4 Transport

TCP / UDP

3 Network

IP

2 Data Link

Ethernet (IEEE 802.1), LLC, MAC, ATM

1 Physical

SDH, OTH, optical frames

Services e.g.: Services e.g.: FTP, HTTP, FTP, HTTP, Telnet Telnet Services e.g.: Services e.g.: MIDI, HTML, GIF MIDI, HTML, GIF JPG, ASCII JPG, ASCII

Messages / Data Datagram

Datagram Packets Packets Frames

IP Ethernet & IEEE 802.3, LLC, MAC, ATM

Bits

Information unit


End System

SDH, OTH optical frames

Packets Frames Bits

Gateway

Gateway

e.g. Security (Firewall, Proxy)

Gateway

TCP / UDP

Gateway

IP

Router

Ethernet (IEEE 802.1), LLC, MAC, ATM

Bridge Switch

SDH, OTH optical frames

Information unit


Equipment

2008/09/29

Transceiver Repeater Hub, Cable

Network Hierarchy According to OSI Reference Model

NSN Location Espoo

Backbone Network A

NSN Location Munich

Backbone Network B

Routers are used to connect networks Switches are used to connect hosts

29



2008/09/29

Fixed Transport Network Structure Fixed Services

Access

Aggregation

IP Edge

Core Applications Server

IMS

Residential Voice, Video, HSI

CIS

Routing MSAN

Business L3 VPN

CIS

L2 switch

L2 Transport Carrier Ethernet / SDH/SONET

COS


Carrier Ethernet / SDH/SONET Optical Transport OTN/DW DM Core

OTN/DWDM Metro

HSI: High Speed Internet CIS: Customer IP service MSAN: Multiservice access node (PON, DSLAM) 30

IP/MPLS Core

BRAS

Layer 2 VPN, Ethernet /TDM Leased Line CES CLS Business Layer 1 Optical/ Wavelength Leased Line

VoIP, VoD, IPTV,…

CES: Customer Ethernet Service CLS: Customer Legacy Services

COS: Customer Optical Service


2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics • Ethernet Standards 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 31



2008/09/29

The original Ethernet by Bob Metcalf

Bob Metcalf, 1973 The original format for Ethernet was developed in Xerox Palo Alto Research Centre (PARC), California in 1972 and called Alto Aloha. Using Carrier Sense Multiple Access with Collision Detection (CSMA/CD) it had a transmission rate of 2.94Mb/s and could support 256 devices over cable stretching for 1km. The two inventors were Robert Metcalf and David Boggs 32



2008/09/29

Advantages of Packet and Ethernet Networks • Packet • Almost 100% of traffic generated by applications is packet based • Multiplex gain • Control plane often deployed in combination with packet services (restoration) • Advantages of Ethernet • Widely deployed • The standard for LAN equipment (10M, 100M, 1G, 10G, 100G)) available in almost every computing device – Chipsets are very cheap and high numbers • Plug and play – Very simple technology to operate • Combines data link layer and switching layer

• Drawbacks of Ethernet: • MAC addressing scheme • Different protocols (STP, RSTP, MSTP) • Limited traffic-engineering and slow failure recovery • Operation Administration and Maintenance 33



2008/09/29

IEEE 802 Standards • IEEE 802.1 – Architecture, management, switching • 802.1D • 802.1Q • 802.1p • 802.1d • 802.1s/w

MAC layer bridges Virtual LANs Quality-of-Service & Multicast support Spanning Tree Protocol (STP) Multiple STP / Rapid STP

• IEEE 802.3 – CSMA/CD (Ethernet) standards • 802.3u • 802.3x • 802.3z • 802.3ab • 802.3ad

34


Fast Ethernet (100Base-TX, 100Base-FX) Full-duplex Ethernet over LAN Gigabit Ethernet over fiber (1000Base-X) Gigabit Ethernet over copper (1000Base-T) Aggregation of multiple link segments (LAG)


2008/09/29

Ethernet Basics IEEE 802.3 Ethernet Interfaces Older Ethernet Implementations: 10 Base 5 “yellow cable” / 10 Base 2 “cheapernet” R

Application

Typical Implementation: Busses / Segments Disadvantage: Collisions multiply when data load Increases

Current Implementations with electrical Interfaces:

Presentation

10 Base T 100 Base T “Fast Ethernet” 1000 Base T “Gigabit Ethernet”

Session Transport Network

Current Implementations with optical Interfaces:

Data Link Physical

35


100 Base FX “Fast Ethernet” 1000 Base SX “Gigabit Ethernet” 1000 Base LX 10 Gigabit-Ethernet Networks 2008 - Carrier Ethernet Transport in Metro and Core Networks Claus G. Gruber, Achim Autenrieth, {claus.gruber,achim.autenrieth}@nsn.com

Typical Implementation: Point-to-Point Advantage: Collisions can be minimized with a switch In optical Ethernets, Collision detection is not possible 2008/09/29

Ethernet Basics IEEE 802.3 Ethernet Frames and MAC Addressing MAC-Address: (Media Access Control) Address on Layer 2 most commonly used on Ethernet, 6 Bytes long, linked to Hardware, worldwide unique Ethernet Frame Application Presentation

Source MAC

Type Field

6 Bytes

6 Bytes

2 By

Data of Layers 3 to 7 up to 1500 Bytes

Check sum 4 Bytes

The Type Field: specifies, which Layer 3 Protocol is contained The Checksum (CRC) secures both addresses, type field and data Minimum length 64 bytes, maximum length 1518 bytes

Session Transport

Destination MAC

MAC-Broadcast addresses all stations on a LAN (Address = ff:ff:ff:ff:ff:ff)

Network Data Link

MAC-Multicast addresses all stations with a particular property e.g. all switches supporting a particular protocol

Physical

36



2008/09/29

Ethernet’s timeline 10 000

Bit rate (Mbit/s)

1999: IEEE 802.3ab 1000Base-T 1998: IEEE 802.3z 1000Base-SX, -LX, -CX

1 000

1973: Ethernet is invented 3 Mbit/s, “thick coax”

1997: IEEE 802.3u 100Base-T2

1979: DIX is formed 10Mbit/s,”thick coax”

1995: IEEE 802.3u 100Base-T4,-Tx,-FX

1993: IEEE 802.3j 10Base-FL,-FB,-FP

100

1990: IEEE 802.3i 10Base-T 1985: IEEE 802.3 10 Base 2,

2002: IEEE 802.3ae 10GBase-SR,-LR, ER, LX4,-SW, LW, EW

1985: IEEE 802.3b 10 Broad 36

10 1987: IEEE 802.3d FOIRL 1987: IEEE 802.33 1Base5

1983: IEEE 802.3 10 Base 5

1 1970

37

1975


1980

1985

1990


1995

Year 2000

2008/09/29

2005

Ethernet Basics Ethernet Switching (1) A

B

C

D

F

E

A

C

MAC-Learning 1

2

3

4

6

5

C?

MACTable Address

38


Port 1

A

Port 2

Port 3

Port 4

Port 5

C


Port 6

F

2008/09/29

Ethernet Basics Ethernet Switching (2) A

B

C

D

F

E

A F

C D Flooding

1

2

3

4

6

5

D?

MACTable Address

39


Port 1

A

Port 2

Port 3

Port 4

Port 5

C


Port 6

F

2008/09/29

Ethernet Basics 802.1d – Spanning Tree (1) Path 1 (working)

active links blocked links In Ethernet networks loops are strictly forbidden because otherwise broadcast storms would bring down the network performance. With Spanning tree protocol loops are avoided in an Ethernet network: All links that would built up a loop are blocked by the Switches. So STP can be used for protection: If the working link fails, the protection link (i.e. a blocked link) is activated. 40



2008/09/29

Ethernet Basics 802.1d – Spanning Tree (2) Path 1 (broken)

Path 2 (unblocked) If the working link fails, the protection link (i.e. a blocked link) is activated. RSTP (Rapid spanning tree protocoll) improves the switching time from several seconds to approximately one second.

41



2008/09/29

802.1w – Rapid Spanning Tree Protocol (RSTP)

• Spanning Tree was designed for Enterprise. Recovery Time is not acceptable for Carrier Grade. • Rapid Spanning Tree Protocol is identical to STP, except: •STP – Learns the backup route after failure •RSTP – Learns the backup route before failure • The convergence time is significantly shortened:

42

Timing

STP

RSTP

Worst Case

~60s

1s



2008/09/29

802.1s – Multiple Spanning Tree Protocol (MSTP) • MSTP enables the use of different paths for different VLANs (or groups of VLANs)

SW 3

SW 1

VLAN 20

• Traffic can be organized to use all possible links, optimising traffic distribution

VLAN 10

• If a link fails, only the MSTIs (MSTP Instances – individual trees) using that link are affected

• MSTP only works together with RSTP

• Up to 32+1 instances per node

43


SW 4

SW 2

Advantages • Efficient VLAN Paths (e.g. SW 1 => SW 4) • Load-sharing


2008/09/29

Ethernet Basics 802.1Q – VLAN support IEEE 802.3 Frame without VLAN Tag Header Destination address

Source address

Type / Length

Data

CRC

IEEE 802.3 with 802.1Q 4-Byte VLAN Tag Header Destination address

Source address

Type/ Tag 8100

Data

CRC

4 bytes

TPID TAG Protocol Identifier

TCI Tag Control Identifier

2 bytes TAG Protocol Identifier TPID 0x8100

16 bit 44


2 bytes C Priority F I

3 bit 1 bit


VLAN ID 12 bit 2008/09/29

802.1Q Highlights Customer separation by VLAN VLAN Functionality Highlights

Physical view

Up to 4096 VLAN

S

R

Priority 802.1p associated with VLAN S

S

VLAN-based priority take precedence Allows Spanning Tree per VLAN Allows overlapping VLANs VLAN Advantages Better security Solve the broadcast problem

Logical view

Solve the physical location issue R

45



2008/09/29

Ethernet Basics Ethernet VLANs (1) A

1

B

2

C

3

D

4

F

E

A

D

B

D

6

5

? MACTable

Port 1

46


Port 3

B

VLAN 1

VLAN 2

Port 2

A

Port 4

D C


Port 5

Port 6

E E

F 2008/09/29

Ethernet Basics Ethernet VLANs (2) A

1

B

2

C

3

D

4

F

E

A

D

B

D

6

5

X

MACTable

Port 1

47


Port 3

B

VLAN 1

VLAN 2

Port 2

A

Port 4

D C


Port 5

Port 6

E E

F 2008/09/29




2008/09/29

Evolution of Ethernet Hierarchy

Standard Ethernet Frame acc. IEE 802.3

SA: DA: VID: C-VID: S-VID: VID: B-SA: B-DA: B-VID: B-TAG: I-SID: I-TAG: B-VID B-DA B-SA 49

802.1D

802.1Q

802.1ad

802.1ah

Pay load

Pay load

Pay load

Pay load

Contains IP packet

SA

VID

C-VID

C-VID

“Inner” VLAN ID

DA

SA

S-VID

S-VID

“Outer” VLAN ID

DA

SA

SA

DA

DA

Source MAC Address Destination MAC Address VLAN ID Customer VID Service VID VLAN ID Backbone SA Backbone DA Backbone VID a Provider Bridge S-TAG 24 bit Service ID allocated for 802.1Q service instance VLAN identifies per destination alternate path MAC identifies destination node MAC identifies source node


VLAN VLAN XC: based on VLAN ID

Q-in-Q

I-SID

Service ID

B-VID

Backbone VID

B-SA Provider Bridges


Customer MAC

B-DA

Backbone MAC

Mac-in-Mac Provider Backbone Bridges PBB 2008/09/29

PBB-TE

802.1ad Provider Bridge (Q-in-Q) The Concept ▪ Adding another layer of 802.1Q ▪ The purpose - expanding the VLAN space by tagging the tagged packets ▪ The expanded VLAN space allows the service provider to provide certain services, such as Internet access on specific VLANs for specific customers, and yet still allows the service provider to provide other types of services for their other customers on other VLANs. Frame without VLAN Tag Header Type / Destination Source Length address address

Data

Frame with single VLAN tag header 802.1Q Type / Destination Source C-VLAN Length address address Frame with double VLAN tag header 802.1ad Source Destination S-VLAN C-VLAN address address

CRC

Data

Type / Length

CRC

Data

CRC

Support of 4K S-VLAN x 4K C-VLAN = theoretical 16 Mill VLAN 50



2008/09/29

Transport of Ethernet Services Issues with Flat Ethernet Architecture C-DA

Transport TransportNetwork Network

S-TAG C-TAG 802.1ad Frame

C-DA

C-SA

6 octets

6 octets

TP S- TPI SL/T ID VID D VID

2

2

2

2

2

User Data

FCS

46 – 1500 octets

4 octets

• Full transparency ? • Use of client information as forwarding decision ? • Learning of all client MAC addresses in all transport nodes ? • Known issues with STP (resilience and traffic engineering)

51



2008/09/29

802.1ah – Provider Backbone Bridging (PBB) Adding a Transport Hierarchy C-DA

B-DA


Backbone Provider Bridge Frame

B-TAG B-DA

B-SA

6 octets

6 octets

TP BES-VID L/T ID VID

2

2

2

802.1ad Frame (/w or /wo FCS)

FCS

60 – 1526 octets

4 octets

User Data

FCS

46 – 1500 octets

4 octets


C-DA

C-SA

6 octets

6 octets

TP S- TPI SL/T ID VID D VID

2

2

2

2

2

Source: D. Allen, N.Bragg, A. McGuire, A. Reid, „Ethernet as Carrier Transport Infrastructure“, IEEE Communications Magazine, Feb. 2006

• Add a transport hierarchy “MAC in MAC” encapsulation • No learning of customer MAC addresses in the middle of the network • Transport spanning TREES instead • Use global meaning of tag (B-DA (48 bit) and B-VID (12 bit)) 52



2008/09/29

802.1ah – Provider Backbone Bridging PBB, MAC in MAC • Interconnect Provider Bridge networks

Q-in-Q Customer MAC

through a highly scalable Ethernet backbone • MAC in MAC encapsulation – Encapsulation at the backbone edge – Provider’s MAC and VLAN space, isolates provider from customer broadcast domains – Core is agnostic to customer MAC and customer services

Provider Backbone Bridging network

• MAC tables are learned automatically, xSTP prevents loops • Drawbacks – Lack of carrier grade protection

Provider Bridging nw

Payload C-VID S-VID SA DA 802.1ad Provider Bridging nw

(xSTP based) – Lack of effective traffic engineering Customer networks

53


Payload C-VID S-VID SA DA I-SID B-VID B-SA B-DA 802.1ah


2008/09/29

OAM for Carrier Grade Switches • OAM is the carrier tool kit for the network management functions such as fault indication, performance monitoring, security management, diagnostic functions and configuration

• An advanced management tool kit contains:

OAM management Network & Service Level Transport link level

54


Element Manager System

802.1ag

MPLS OAM

Connectivity Fault Management

VLAN OAM MEF recommendation

802.3ah – EFM (Ethernet at the first mile)


2008/09/29

Advanced Ethernet Features Quality of Service - QoS (1) • QoS allows to guarantee parameters like - Bandwidth - Packet loss rate - Maximum delay - Maximum jitter • Examples of typical service class definitions: - Gold: Guaranteed Bandwidth, very low packet loss rate, - Bronze: - Network

Minimum jitter suitable for VoIP and Video Boadcast No guarantees suitable for Data transmission (data packets can be re-transmitted in case of loss) Most important traffic, highest priority

Control:

55



2008/09/29

Advanced Ethernet Features Quality of Service - QoS (2) Link Capacity Bandwidth available to other services at time t

CIR

EIR

t

Service End-to-End

• CIR: Committed information rate • PIR: Peak information rate • CBS: Committed burst size • PBS: Peak burst size 56



2008/09/29

Advanced Ethernet Features SLA Guarantees for all Services Packet stream (one direction)

Packet classification

Packet scheduling

+ Egress buffers For one egress port

Ingress ports

Egress port

Waste The critical point in the packet flow is the summarization of several ingress ports to one egress port. Therefore one egress buffer per service class is required. In this buffers high priority packets can overtake low priority packets. 57



Many Low priority packets are dropped, Few medium priority, None high priority. High priority packet Medium priority packet Low priority packet 2008/09/29




2008/09/29

Connectionless and Connection Oriented Transport The OSI 7-layer model specifies two methods for packet transport: Connectionless

Connection oriented

• Every packet can be taken at any path as long as it gets to its final destination

• A predetermined path is used between two end nodes for packets of the same service

• Service BW can not be guaranteed

• Bandwidth reserved End-to-End to ensure quality

• In case of failure nodes are required to re-calculated path which may take long time

• Protection paths are preset and available for immediate usage

• No constant End-2-End monitoring

• Known path allows more E2E OAM capabilities

• BW for protection can be reserved in advance

Primary path Backup path 59



2008/09/29

Carrier Ethernet (cl) vs. Carrier Ethernet Transport (co) • Carrier Ethernet • Forwarding based on Spanning Tree • • • •

– Inefficient use of resources Limited traffic engineering possibilities Very complex optimization tasks when using multiple trees Slow restoration upon failures (seconds) Flat switching hierarchy (broadcast if unknown)

• Carrier Ethernet Transport • Forwarding based on transport label not on customer MAC address • Establishment of virtual tunnels (paths) • Packets are tagged and switched accordingly • Broadcast if unknown is disabled (hierarchy) hierarchy • Centralized management or distributed control plane (e.g. GMPLS)

60



2008/09/29

Carrier Ethernet Transport Traffic Engineering and Resilience

• Traffic Engineering can be done by applying tunnel characteristics • Route of tunnel can be optimized • Multiple tunnels and traffic distribution • Intermediate grooming and merging of tunnels • Multi-layer traffic engineering especially between Ethernet and WDM

• Resilience mechanisms can be based on tunnels • A large number of path-based resilience mechanisms can be applied for Carrier Ethernet • Protection and restoration • Multi-layer resilience optimization

61



2008/09/29

Carrier Ethernet switches with c/o Ethernet Benefits and features for packet transport Features and Benefits – Advanced connection oriented Ethernet mechanisms Well determined and predictable network operation Advanded resilience mechanisms possible – Traffic engineering (Traffic separation per VLAN, Classification per port and port+VLAN ++, Policing, QoS (basic- , Diffserv-, Enhanced-mode), horizontal split) efficient use of fibers, balancing of the traffic load on various links in the network

Challenges – Multicast: Interworking of IGMP and PBB-TE still to be verified – Synchronization and clock provisioning in mobile backhaul Interworking with DWDM – Increased scalability and cost-efficient long-distance transport (Grey interfaces up to 80 km)

62



2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies • Ethernet Label Switching 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 63



2008/09/29

Ethernet Label Switching (ELS) aka VLAN Cross-Connect “Q in Q Tunnelling” (IEEE 802.1Q, IEEE802.1ad)

• Idea: Use the existing Ethernet header

(802.1ad) but forward according to ingress port and VLAN-ID, not MAC address • Add tags if required (label stacking) • Forwarding decision based on single VLAN-ID (12 bit) or double VLAN-ID (24 bit) with local link scope (16M connections per port) • Replacing Flooding and MAC Learning with configuration of VLAN-Switching

Cross Connect

VID = 10

VID = 20

1

5

2

6

VID = 10

Single Tag

DA

SA

6 octets

6 octets

VID = 50

7

VID = 11 3

VID = 50 4

L/T 2 2

8

Bridge

User Data

FCS 4 octets

TAG1 TAG2 Double Tag

64

802.1ad Frame


DA

SA

6 octets

6 octets

TP TPI VID VID ID D

2

2

2

L/T 2 2


VID = 10 VID = 72

VID = 10

TP VID ID

2

VID = 17 VID = 50

TAG 802.1Q Frame

VID = 17

User Data

FCS 4 octets

2008/09/29

VID = 50

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies • PBB-TE 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 65



2008/09/29

802.1ah – Provider Backbone Bridging (PBB) Adding a Transport Hierarchy C-DA

B-DA



B-TAG B-DA

B-SA

6 octets

6 octets

TP BES-VID ID VID

2

2

L/T 2


FCS

60 – 1526 octets

4 octets

User Data

FCS

46 – 1500 octets

4 octets


C-DA

C-SA

6 octets

6 octets

TP S- TPI SID VID D VID

2

2

2

L/T 2 2


• Add a transport hierarchy “MAC in MAC” encapsulation • No learning of customer MAC addresses in the middle of the network • Transport spanning TREES instead • Use global meaning of tag (B-DA (48 bit) and B-VID (12 bit)) 66



2008/09/29

802.1Qay – Provider Backbone Bridging–Traffic Engineering (PBB-TE) Transparent Tunneling of Ethernet Services B-DA



B-TAG B-DA

B-SA

6 octets

6 octets

TP BES-VID ID VID

2

2

L/T 2


FCS

60 – 1526 octets

4 octets

User Data

FCS

46 – 1500 octets

4 octets


C-DA

C-SA

6 octets

6 octets


2

2

2

L/T 2 2


• Add a transport hierarchy “MAC in MAC” encapsulation • No learning of customer MAC addresses in the middle of the network • Transport PATHS instead GMPLS or NMS configured • Use global meaning of tag (B-DA (48 bit) and B-VID (12 bit)) 67



2008/09/29

C-DA

802.1Qay – Provider Backbone Bridging–Traffic Engineering (PBB-TE) B-DA

The Global Label Meaning

Same VLAN-ID

A

C

B-DA

Scalability: • 4096 different paths per B-DA • Reuse of B-VLAN-ID possible • Merging of paths possible

B

Incoming Port

Incoming B-DA / B-VLAN-ID

A

B

A Simplicity: • Inherent knowledge of the destination • Possibility to have PATHS and (multicast) TREES with the same forwarding technology 68



Outgoing Ports

2008/09/29

B&C

802.1Qay – Provider Backbone Bridging – Traffic Engineering

• A profile of PBB which allows

• •

• • •

69

engineering deterministic, protected, and secured connection oriented trunks and services MAC learning, xSTP, Broadcast unknown frames disabled Management plane is used to set up Ethernet Switched Paths (ESP) • Populates forwarding table • Calculates load for each ESP and allocates to physical link • Also sets up protecting path Alternative: control plane signaling Traffic Engineering 50ms protection switching (G.8031)


MAC in MAC PB nw

MAC in MAC

Provider Backbone Bridging network

Payload C-VID S-VID SA DA I-SID B-VID B-SA B-DA 802.1ah


Port 1, S-VID 1, C-VID1 Port 1, S-VID 1, C-VID2 I-SID 1 Port 2 I-SID 2 Trunk (B-DA, B-VID)

2008/09/29

PB nw

PBB / PBB-TE Summary Topic

Status

Availability/Resilience

☺

Predictability

☺

Currently shared end-to-end protection, Restoration mechanisms proposed Configured working and backup paths (SDH-like)

QoS

☺

Similar to IP packet forwarding

☺

Scalability Efficient use of resources

☺

Manageable and simple

☺

Remote configuration

☺

Client/Transport separation, ~4000 paths per destination in PBB-TE with implicit merging Not based on shortest paths, not based on global link weights OAM Mechanisms defined or standard proposed Currently static NMS solutions, GMPLS in discussion Via NMS or Control Plane (in definition)

Pt2Pt, Pt2Mpt, Mpt2Pt, Mpt2Mpt Others (e.g. mobility)

Pt2Pt ☺, Pt2MPt ☺ , other proposals in discussion in discussion (e.g. synchronization)

Standardization status

Many proposals not yet standard (most expected mid/end 2009) First products available

Product status 70



2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies • MPLS Basics 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 71



2008/09/29

MPLS Basics Two Aspects Multi Protocol MPLS is encapsulating Ethernet, Frame Relay, ATM and IP and transport them transparently through each network

Label Switching Each packet is marked with a short, fixed-length label Forwarding of packets is based on this label Reservation of resources is associated with this label

72



2008/09/29

MPLS Basics Acronyms LER

Label Edge Router

LSR

Label Switching Router

Non-MPLS access network Label LSP: Switched Path MPLS Backbone 73



2008/09/29

MPLS Basics Shim Header Structure Label (20 bits)

L2 Header MPLS Header

CoS S

TTL

IP Packet

32 bits

• • • • •

MPLS header consist of four fields Label—used to associate packet with an LSP Experimental bits—carry packet queuing priority (CoS) Stacking bit Time to live—limits packet lifetime within LSP – In most cases, the IP TTL is copied into the MPLS TTL • Some label values are reserved

74



2008/09/29

MPLS Principles • “Virtual connections in a connectionless network” • Add a label to an IP packet that encodes a predefined tunnel IP Packet

Label

IP Packet

Label

• Traffic towards different destinations can be separated or aggregated and forwarded along a pre-defined path using only small labels as “routing” decision • Label stacking is possible • Efficient Traffic Engineering due to source routing • Fast resilience mechanisms 1: The edge router classifies packets and adds an MLPS header, see table below to them

75


2: Small tables, fast “routing”

Label Edge Router (LER)

3: Routing according to IP header

Label Switch Router (LSR)


2008/09/29

MPLS Major Tasks • Information distribution (network topology and capacity) • Based on existing IP protocols (OSPF, IS-IS, EIGRP) - inband • Path calculation • What are the best paths? • Constraint Based Routing (CBR) • Path setup, label distribution and exchange • Label Distribution Protocol (LDP) • Reservation Protocol (RSVP-TE) • Forwarding of traffic along the MPLS path

76



2008/09/29

Forwarding Principle 3: Last but one router drops the MPLS header

1: The edge router classifies packets and adds an MLPS header to them (see table below)

Label Switch Label Switch Router (LSR) Router (LSR)

Address prefix 129.187.0.0/16 185.222.0.0/16

Label Edge Router (LER) Address prefix 129.187.0.0/16 185.222.0.0/16

Next Hop IP Address 10.152.4.2 10.152.4.2

Out Label A A


Next Hop IP Address 102.4.4.2 102.4.4.3

2: Forwarding using labels instead of IP addresses Swapping of labels at each intermediate router In Label A

77

4: Routing according to IP header

Out Label C


2008/09/29

Path Setup with RSVP • The Ingress LSR (I-LSR) sends a PATH message along the calculated route (source routing). • Each intermediate router checks if the required bandwidth is available and forwards the message to the tail of the path (last router). • The Egress LSR (E-LSR) sends a RESV message back along the same path. On the way back, the resources are reserved and labels are selected and signaled to the upstream LSR • Paths are updated / refreshed via a “soft-state mechanisms”

I-LSR

PATH message

E-LSR

RESV message

78



2008/09/29

Pros and Cons of MPLS Switching

Advantages: +Aggregation of traffic +Reduction of routing entries +Efficient traffic engineering possibilities (source routing) +Fast and efficient resilience mechanisms +VPN support +GMPLS support

79


Disadvantages: - Additional technology below the IP Layer - Handling of MPLS paths (number, soft state) - Complexity of Network Configuration - Tight interelation of IP and MPLS makes it quite complex to handle


2008/09/29

Challenge: Is MPLS ready to replace SDH/SONET? Carrier Ethernet will replace SDH/SONET infrastructure over time SONET/SDH infrastructure traditionally designed and managed by transport departments

Transport teams mentality

80

Transport teams view on IP/MPLS

Long term statically provisioned paths, pre-determined backup paths

Believe IP/MPLS is not suitable for transport applications

Highly automated operation environment

Consider it to be very complex (LDP, ISIS, OSPF, MPLS-TE, CSPF, FRR,..)

Strong reliance on automated OAM and fault management systems

Do not need dynamic routing protocols, and recovery times too slow

Simple static control plane scores well over complex dynamic control plane

IP/MPLS OAM tools not consistent with transport OAM requirements



2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies • T-MPLS 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 81



2008/09/29

T-MPLS (Transport Multi Protocol Label Switching) • Idea: Use the MPLS concept known from IP and adapt it for forwarding issues defined in ITU-T G8110.1 • Operate independently of its clients and its associated control networks (Management and Signaling Network). “IP/MPLS – IP + SDH” • MPLS with a few changes: • Use of Penultimate Hop Popping is prohibited • Uni-directional and bi-directional LSPs can be defined • Use of global or per interface label space • Three types of Signalling Communication Channels (in-band via native IP packets, in-band via dedicated LSP, out-of band) • OAM based on Y.1711 and Y.1731 • Protection switching (ITU-T Y.1720) • Merging and ECMP is prohibited • Multicasting in alignment to on-going work in IETF S-TAG C-TAG GFP or Ethernet

82


T-MPLS

DA

SA

6 octets

6 octets


2

2


2

L/T 2 2

User Data

FCS

46 – 1500 octets

4 octets

2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies • MPLS-TP 6. Carrier Ethernet Transport Network Architecture & Solutions 7. Outlook Towards Future Internet Architectures 8. Conclusion 83



2008/09/29

MPLS-TP standardization process and timeline

MPLS-TP MPLS-TP Timeline:

Joint Working Team

ITU-T - IETF Joint Working Team (JWT) was setup in March 2008 Agreement reached on recommendations: End of April 2008 First draft: July 2008 Expected final agreements: E 2009

84



2008/09/29

Technologies comparison IP/MPLS

ELS

T-MPLS

PBB-TE

MPLS-TP

Transport oriented Scalable Multipoint support Standardized or in process of standardization

~

~ /~ ~

/~

/~

Terminated by ITU-T

85



2008/09/29

Market shift from PBB-TE to MPLS-TP The options IP/MPLS

Mature but not transport oriented

T-MPLS

Transport oriented Terminated by ITU-T

NSN focus

T-MPLS

In standardization process

MPLS-TP ELS

PBB-TE

Standardized partly (single tagging) Double use of VID for user-traffic separation and routing

Based on MPLS maturity Enhanced for Transport

In standardization process Not mature No Control plane No multipoint support

MPLS-TP MPLS Transport Profile IP/MPLS & L2 MPLS can be categories as IETF MPLS 86



2008/09/29

Nokia Siemens Networks is a major player in MPLS-TP standardization MPLS

• Connection Oriented • Lacks some transport capabilities

JWT

T-MPLS

• A subset of MPLS plus additional capabilities providing packet transport • Uses the same Ether type as MPLS but some mechanisms are not compatible with MPLS

Nokia Siemens Networks has participants in the Joint Working Team Nokia Siemens Networks acts as author and co-author for requirements, framework and solution documents

87

2. Requirements

3. Frameworks

4. Solution Documents

2.1 MPLS-TP 2.2 OAM 2.3 Network management

3.1 MPLS-TP 3.2 3.3 OAM 3.4 Survivability 3.4.1 for LSPs 3.4.2 for PWs 3.6 Control Plane 3.7 Network Management

4.1 Generic ACH Alert Label Definition 4.2 ACH definition 4.3 OAM Procedure document 4.3.1 OAM Analysis document 4.3.2 OAM Tool documents 4.4 Survivability 4.4.1 Linear Protection 4.4.2 Ring Protection 4.5 Control Plane protocols



2008/09/29

What is MPLS-TP? A new standard focused on extending MPLS as a viable transport option to help building the next generation packet transport network Being developed by IETF as a result of a collaboration between IETF and ITU-T via joint working team (JWT)

Objective: To bring transport requirements into IETF MPLS and extend IETF MPLS forwarding, OAM survivability, network management and control plane protocols to meet those requirements through IETF standard process The JWT is divided into multiple sub-groups focused on:

88

Forwarding plane

Protection

OAM

Control plane



Management

2008/09/29

MPLS-TP defines a profile of MPLS targeted at Transport applications. This addresses specific MPLS characteristics and extensions required to meet transport requirements.

OAM extensions

MPLS-TP profile

Control Plane extensions

Survivability extensions

Management extensions

Alert Label Definition extensions

MPLS-TP foundation The architecture for a transport profile of MPLS (MPLS-TP) is based on IETF MPLS (RFC 3031) & IETF PWE3 (RFC 3985)

89



2008/09/29

Desire: To make MPLS more Transport Oriented Configuration of

NMS Point & Click

LSP’s & PWE’s via NMS and later dynamic control plane

LSP and PWE management via external

OAM and Data path

LSP, PWE, OAM

must be congruent (use the same path)

Protection and OAM

works independent of control plane

mechanism works within the MPLS architecture

LSP’s, PWE’s nesting similar to SONET/SDH environments

Management plane: Configuration of LSP, PWE with point & click MPLS-TP

90

Control plane:

Optional and separated from data plane

Data plane:

Data, OAM, protection congruent within architecture



2008/09/29

The Goal for the MPLS-TP technology MPLS-TP will enable the migration of SONET/SDH networks to a packet-based network that will easily scale to support packet services in a simple and cost effective way.

Efficient support of

packet based services on the transport network

Control and deterministic usage of network resources

Sonet/SDH

comparable Reliability and Operational Simplicity

MPLS-TP for Connection Oriented services, scalability and flexibility

Ethernet Economics Utilisation

Preserve the look-and-feel to which carriers have become accustomed to deploying SDH/Sonet networks

End-to-End monitoring

and control of customer services

91



2008/09/29

Main characteristics of MPLS-TP No modification of MPLS forwarding/data plane architecture Current Standards for LSP’s and PWE’s construct Configure LSP’s and PWE’s via Management plane Bidirectional and congruent point-to-point LSP’s Framework supporting transport OAM capabilities for PWE’s, LSP’s Complete Fault, Configuration, Accounting, Performance and Security (FCAPS)

Ability for LSP’s and PWE’s to be managed at different nested levels (path, segment, multiple segments) Interoperability with existing control and forwarding plane

92



2008/09/29

MPLS Transport Profile - Terminology

Emulated Service

Pseudo-wire (PWE)

Multi-node LSP network

CE1

Attachment Circuit

Attachment Circuit PE1

PW1

PE2

Definition of an MPLS Transport Profile (TP) within IETF MPLS standards Based on PWE3 and LSP forwarding architecture IETF MPLS architecture concepts

The major construct of the transport profile for MPLS are LSPs

93



2008/09/29

CE2

End to End LSP operations LFIB:AB-BC

Primary Path

PW-L, AB

LFIB:CD-DE

LSP OAM

DE, PW-L

E

LFIB:BC-CD

A

YZ, PW-L LFIB:XY-YZ

PW-L, AW

LFIB:AW-WX

Backup Path

LFIB:WX-XY

LSP OAM

Path diversity is not part of the OAM process. It is the responsibility of the Control Plane OAM function uses GAL with Generic Channel Association Pre-provisioned primary and backup paths LSP OAM running on primary and back-up paths OAM failure on backup path

Alert NMS

OAM failure on primary path A and E updating LFIB to send and receive PW-L traffic over backup path 94



2008/09/29

Overview: OAM hierarchy and mechanisms A

B L1/L2

C

D

L1/L2

L1/L2

E

F

L1/L2

L1/L2

Segment LSP Midpoint End to End LSP Pseudo-wire

L0/L1: Loss of Light; G.709, SONET/SDH LoS, LoF, ES, SES (NOT DISCUSSED)

Non MPLS L2 connectivity: Native L2 solution 802.1ag (Not Discussed) , Non IP BFD Failure propagation across layers is supported by this architecture

General LSPs : Generic Exception Label and Generic Associated Channel Includes End to End and segment LSPs Used to carry a variety of OAM, Mgmt, signalling protocols.

Pseudo-wires : PWE3 Associated Channel 95



2008/09/29

LSP example - end to end and per carrier monitoring

PE PE

Carrier 1

NNI

PE PE

PP

PP

PE PE

Carrier 2 NNI

PE PE

PE PE

PP

PE PE

NNI

end endto toend endLSP LSPOAM OAM MEP

MIP

MIP segment segmentLSP LSPOAM OAM (carrier 1) (carrier 1)

MEP

MIP

MIP

MIP segment segmentLSP LSP OAM OAM (inter (intercarrier) carrier)

MEP MEP

MIP

MEP

segment segmentLSP LSPOAM OAM (carrier 2) (carrier 2)

MEP MEP

MIP

MEP

A segment is between MEPs OAM is end to end or per segment In SDH/OTN and Ethernet segment OAM is implemented using Tandem Connection Monitoring (TCM) The OAM in each segment is independent of any other segment Recovery actions (Protection or restoration) are always between MEPs i.e. per segment or end to end

96


MEP: Maintenance End Point MIP: Maintenance Intermediate Point




2008/09/29

Market Trends in Infrastructure Growth of Internet, Video Services, hence IP traffic Revenue shifts from voice to data Effective technology for carrying IP

Ethernet Cost points drops Video accelerate the problem IP traffic doubles every year Drives infrastructure migration from SDH to packet – hence Ethernet Technology dominance

Packet All Packet

All TDM TDM Yesterday

98


Today

Tomorrow


2008/09/29

The migration to CET covers all IPT business lines and is key IPT strategy • Migration from NG-SDH installed base • Evolution to hybrid platform with CET connectivity

From TDM to Packet From current switches to NG portfolio From existing base to optimized networks

• Migration from existing non-Connection oriented L2 aggregation

• Multilayer optimization: L1 to L3 • Greenfield overlay with Carrier Ethernet Transport • Interworking with customer edge (L2, L3) and provider edge (L3) as well as with 3rd party L2 Carrier Ethernet Transport

Microwave Packet Radio Hybrid NG Metro

99


Multi-reach DWDM NMS

CET Switches


2008/09/29

Migration towards Carrier Ethernet Transport Evolution of Ethernet in all network technologies IP/MPLS

IP

Classical IP/MPLS T-MPLS MPLS-TP

Carrier Switches

PBB-TE PBB ELS

Classical Ethernet

Microwave TDM

Classical Ethernet

Integration of Ethernet functionality

Hybrid Packet/TDM

SDH/SONET

WDM/OTN

100

ROADM/PXC


Ethernet ADM with GFP-T/F

Integration of Ethernet/ODU-Switching Ethernet ADM with L2 Switch


Integration of Ethernet functionality

2008/09/29

CET Migration Scenarios A. Replacement of TDM by Packet Transport Platform based on c/o Ethernet – Deployment of a new Packet Transport platform to replace SDH/SONET TDM platform.

B. Hybrid (TDM/Packet) scenario – Hybrid platform deployment for all new traffic (packet and TDM) in a jointly network with an existing NG-SDH platform for TDM traffic

TDM + Packet

C. Integration of Ethernet in DWDM – GbE add/drop cards or L2 switch cards allows cost-efficient and scalable DSLAM aggregation / mobile backhaul in metro aggregation networks

101



2008/09/29

Rethinking the Role of the Layers

Services Layer 3 IP/MPLS IP/MPLS

Service-awareness High-touch networking

Fun ctio n sp lit

Layer 2

SONET/SDH SONET/SDH

Layer 1

Transition

Efficient end-to-end carrier-grade packet transport pt-pt, pt-mpt, mpt-pt, mpt-mpt ETHERNET ETHERNET

IncludingAggregation, Aggregation,Metro Metroand andCore Core Including

OTN/WDM OTN/WDM

Common OTN/WDM infrastructure

IP is the convergence platform for applications and services. services Ethernet and OTN/WDM will be the convergence platform for transport. transport 102



2008/09/29

Carrier Ethernet Transport in a Multilayer Network Optimized Transport based on Packet, TDM and Optics Subscriber/ Service

Access

Aggregation

Edge

Functionality only where needed Minimize intermediate routing - offload routers

Core Layer 3 / IP

Residential

LER

LSR LSR

IP, Voice, Video

DSLAM

Layer 2 / Carrier Ethernet Transport

Business Access Switch

L2 VPN Leased Line, E-Line

103

GPON FTTH

Metro DWDM

Core DWDM

Service Awareness Packet routing Traffic Engineering Robust network (Restoration) Packet switching Traffic Engineering Robust network Native layer 2 Predictable behavior Carrier Grade OAM Cost efficient 2.5Gbps, 10Gbps, 40Gbps, 100Gbps

Native transport Optimal mix of intermediate of services on grooming and routing and the lowest possible layer transparent bypass dependent on service (Ethernet + WDM) WDM Networks 2008 - Carrier Ethernet Transport in Metro and Core Networks requirements and cost © Nokia Siemens Networks Claus G. Gruber, Achim Autenrieth, {claus.gruber,achim.autenrieth}@nsn.com 2008/09/29

Contents 1. Introduction 2. Operator Requirements for Transport Networks 3. Ethernet Basics 4. Carrier Ethernet Evolution 5. Carrier Ethernet Transport Technologies 6. Carrier Ethernet Transport Network Architecture & Solutions • Applications 7. Outlook Towards Future Internet Architectures 8. Conclusion 104



2008/09/29

Carrier Ethernet enables service providers to deliver a wide range of mission-critical applications Inter-LAN or Inter-PABX Ethernet, Storage, Video (surveillance)

VPLS or VPLS-TE Service Video conference, Ethernet PMP, Intranet access Customer site

Customer site Customer HQ

Customer HQ

Carrier Ethernet

Customer site

Customer site

Customer HQ

High Speed Internet Access

Customer HQ

VPLS r Group

High Speed Access to IP-VPN Low cost access combined Customer with centralized router site to offer IP VPN

Carrier Ethernet

Carrier Ethernet www Customer site

Multi-Service Access Customer HQ

IP VPN OC-3/STM1 or NxT1/E1 Customer site

Carrier Ethernet

IP VPN

T1/E1

Network Customer site VPLS: Virtual Private LAN Services 105


VPN: Virtual Private Network


2008/09/29

CET Solution Focus addresses three main operator broadband challenges Residential and broadband • High Speed Internet (HSI) • IPTV, VoD • Voice

Mobile Backhauling • TDM PWE3 (CESoP) for 2G backhaul

2G

• Ethernet backhaul for 3G, I-HSPA, LTE & WiMAX

LTE

WiMax 3G

• Synchronous Ethernet

Business Services • E-Line, E-LAN (TE, non TE), E-Tree (Hub & Spoke) • SAN, legacy technologies (TDM, clock sync)

106



2008/09/29

Mobile Backhaul – QoS BSC E1 Eth

STM-1

RNC

E1 Eth

Mobile Operator QoS mapping and Traffic Policing on UNI ports

CET (MPLS-TP)

CET (Ethernet)

Eth

Fixed Operator

Traffic Engineered paths for different traffic types

Low Delay, Low Jitter and zero Packet Loss for realtime traffic types: ToP Synchronization, Signaling and Control, Voice

MPLS-TP isisQoS-enabled MPLS-TP QoS-enabledTransport Transportnetwork, network,providing providingTraffic TrafficEngineering,Queuing, Engineering,Queuing, NG-CET 1200 NG-CET 2200-1G Scheduling, Scheduling,Policing Policing\\and \\andMapping Mappingcapabilities capabilitiesthat thatmeet meetMobile Mobilesystem systemE2E E2Erequirements requirements NG-CET M

107


NG-CET Networks 2008 - Carrier Ethernet Transport in Metro and Core Networks Claus G. Gruber, Achim Autenrieth, {claus.gruber,achim.autenrieth}@nsn.com

L 2008/09/29

Mobile Backhaul – OAM Service level OAM: Delay, Jitter, Frame loss (Y.1731)

BSC E1 Eth

STM-1

RNC

E1 Eth

Mobile Operator Performance Monitoring: • Service level • LSP level • Statistics History • Traffic Crossing Alarms

CET (MPLS-TP)

CET (Ethernet)

Fixed Operator LSP OAM (Delay, Jitter, Frame loss ) Fault Management: • Link, segment and E2E Service continuity Check to identify failures

Network Maintenance: • Loopback • Ping and Traceroute

Protection Triggering: • Service & LSP OAM triggers Protection Switching events in