1

Using Multi-Layer Routing to Provision Services across

MPLS/GMPLS Domain Boundaries

Using Multi-Layer Routing to Provision Services across

MPLS/GMPLS Domain Boundaries

Andrew G. MalisChief Technologist, TellabsChairman and President, MFA Forum

2

Discussion of Convergence is EverywhereDiscussion of Convergence is Everywhere

“…the access infrastructure telcos are building to support IPTV services will finally give them the ability to converge voice, video and data onto a single network…”– Americas Network

Pseudo-Wires are “the solution for convergence in future telecom networks, because it preserves profitable legacy services even as it enables the creation of a truly next-generation network” – Heavy Reading

“3GPP defines a new subsystem to enable the convergence of voice and data applications and the harmonization of various mobile network technologies over IP”– Deutsche Bank

Convergence is changingthe face of the network

3

The Benefits of ConvergenceThe Benefits of Convergence

Services reduced to applications on converged infrastructure

Reduced operations

Reduced core cap-ex

New services can be provided

Individual Networks per Service limited service interaction

The details in getting there are not small…

4

The Reality of ConvergenceThe Reality of Convergence Non-native service delivery

End User Services are always layered on something else

Which Technology for Convergence? WDM

Layer 1 (GMPLS)

Layer 2/3 (MPLS)

The whole network will not be converged overnight Too many existing services already deployed

The whole network will never be completely converged Fully depreciated equipment still generates revenue

How can this all be supportedoperationally?

5

Access A(WDM)

Access B(SONET/SDH)

Core(MPLS Pseudo Wire)

CPE-A

CPE-B

Islands of Technology

6

CPE-A

CPE-B

Different Organizations Manage These Technologies

Access A(WDM)

Access B(SONET/SDH)

Core(MPLS Pseudo Wire)

7

CPE-A

CPE-B

Provisioning Service is expensive and time consuming

We will fill your order

within 3 months

I forgot that we ordered this service!

Access A(WDM)

Access B(SONET/SDH)

Core(MPLS Pseudo Wire)

8

WDM NetworkGE/NNI

BPON

GE/NNI

GE/NNI

GE/UNI

T1, T3, OCNUNI

GE/NNI

10/100

BPON

10/100

T1, T3, OCNUNI

GE/NNI

STS1 Network*OCN/NNI

OCN/NNI

nxT1

OCN/NNI

OCN/NNI

GE/UNI

T1, T3, OCNUNI or Null

OCN/NNI

10/100

nxT1

GE/MPLS

10/100 T1, T3, OCNUNI or Null OCN/NNI

MPLS Network

OCNUNI

OCNUNI

BPON

CPE-A

CPE-B Access B(SONET/SDH)

What would be best?Access A

(WDM)

Core(MPLS Pseudo Wire)

9

WDM NetworkGE/NNI

BPON

GE/NNI

GE/NNI

GE/UNI

T1, T3, OCNUNI

GE/NNI

10/100

BPON

10/100

T1, T3, OCNUNI

GE/NNI

STS1 NetworkOCN/NNI

OCN/NNI

nxT1

OCN/NNI

OCN/NNI

GE/UNI

T1, T3, OCNUNI or Null

OCN/NNI

10/100

nxT1

GE/MPLS

10/100 T1, T3, OCNUNI or Null OCN/NNI

End-to-end Signaled Service Provisioning

Sure, we’re setting your service up

as we speak

Wow, that was

fast!

MPLS Network

OCNUNI

OCNUNI

BPON

10

How Can This Be Accomplished?How Can This Be Accomplished? Common flexible control method that understands

layering Path Computation that understand layered networks

Provides service routing given view of potential/available server layer resources

Signaling mechanisms that coordinate calls in different layers

OSSes that can handle integrated views of layer networks Relate services requests (client layer calls) to server resources

in use

Definitions for server layer resources other than SONET/SDH

An Added Benefit: Integrated operations Operations convergence possible due to Common control

methods

11

Why Use GMPLS?Why Use GMPLS? Establishes a common control plane for different networking

technologies

Converge Packet, Cell, TDM, and Optical administrative controls

Automates connection management for all traffic types

Path setup and management (for Packet, Cell, TDM, and Optics)

Handles topology changes automatically

Self-discovery and dynamic configuration of network resources

Provides static and dynamic path reroutes and restoration

Supports Peer-to-Peer and Overlay network models

Integration of optical switches, optical transport, and label switching routers

12

Making Path Computation Layer AwareMaking Path Computation Layer Aware

Routing today treats different technologies as separate topology graphs

MPLS-TE Network

SONET NetworkPON Access

WDM Network

13

Making Path Computation Layer AwareMaking Path Computation Layer Aware

Normal Path Computation cannot find paths between endpoints on different islands

MPLS-TE Network

SONET NetworkPON Access

WDM Network

14

Making Path Computation Layer AwareMaking Path Computation Layer Aware

As a result, end-to-end services are separately routed by each island, and interconnected by hand

MPLS-TE Network

SONET NetworkPON Access

WDM Network

15

Making Path Computation Layer AwareMaking Path Computation Layer Aware

By merging the graphs, adding client-layer matrixes and client/server adaptation, the graph becomes continuous

MPLS-TE Network

SONET NetworkPON Access

WDM Network

16

Making Path Computation Layer AwareMaking Path Computation Layer Aware GMPLS routing extensions

Link Attribute Announcements that:

remove ambiguity of adaptations supported

announce adaptations in a technology independent manner

necessary to allow for source routing to be done anywhere

include link costs that take into account:

different costs for each layer supported by a link

cost to utilize adaptation

Path Computation Algorithm that:

understands multiple matrices per node

updates “signal stack” when adaptations are pushed/popped

17

IP & Optical Layer Combination Practical ExampleIP & Optical Layer Combination Practical Example

GMPLS Layer

MPLS LayerPPE

OXC

Major Objectives:-Eliminate electrical packet processing in core-Minimal or no change to existing MPLS routers-Take advantage of TE in both networks – carry out multi-level TE

18

MPLSGMPLS Interworking:Overlay vs. Peer to PeerMPLSGMPLS Interworking:Overlay vs. Peer to Peer

Optical Domain is not visible to IP domain

MPLS domain can not perform efficient TE

Do not have to update MPLS nodes to GMPLS

Overlay Model

Optical Domain is visible to IP domain

Optimal route assured since the MPLS network can perform efficient TE by understanding GMPLS network resources

Must update MPLS node to GMPLS

Peer to Peer Model

Can we take advantage of both models?

19

Optimal Design of an MPLS Core NetworkOptimal Design of an MPLS Core Network

P routers

PE routers

Service Demands

Feasible and cost-effective with GMPLS

LSP mesh between PEs originally traverses P routers

Direct PE-PE tunnels preferable where sufficient commonality of traffic endpoints exists

Move LSPs to direct route reducing P-router load

20

Direct

Essence of Multi-Layer Core OptimizationEssence of Multi-Layer Core Optimization

Transport Layer

MPLS LayerP

PELSPs

Original

GMPLS affords dynamic, optimal size direct tunnels – without necessity to wait for transport service orders

Original tunnel

Six router interfaces

Forwarding on two P routers

Ten transport interfaces

2x bandwidth on links to P routers

Direct tunnel

Two router interfaces

Six transport interfaces

21

Standards ActivitiesStandards Activities GMPLS

IETF – Multi-region Network

draft-shiomoto-ccamp-gmpls-mrn-reqs-03.txt

draft-leroux-ccamp-gmpls-mrn-eval-02.txt

ASON

ITU – Multi-layer calls and Multi-layer Routing

G.8080 Amendment 2

Q12/15 March 2005 Interim WD53r1

OIF – UNI 2.0 Ethernet Services & E-NNI Routing

Multi-layer call support

Architecture supports multi-layer routing (oif2004.477)

Demonstrated at Supercomm 2005

22

ConclusionConclusion Convergence is occurring in the network

New Services

Lower CapEx and OpEx for network operators

MPLS has become the industry-standard mechanism for data network convergence (Layer 3 and Layer 2 services)

GMPLS is being deployed as optical network equipment replacement picks up

Flexible control methods that support layering are in progress

Standardization activities underway in ITU, IETF and OIF

Multi-vendor interoperability already demonstrated