brkagg-1000 c2 web

7/29/2019 Brkagg-1000 c2 Web

1/75

2009 Cisco Systems, Inc. All rights reserved. Cisco PublicPresentation_ID 1

Packet-Based RAN for Mobile Operators

BRKAGG-1000_c2

7/29/2019 Brkagg-1000 c2 Web

2/75

2009 Cisco Systems, Inc. All rights reserved. Cisco PublicBRKAGG-1000_c2 2

Agenda

Market Drivers forNext-Gen RAN

Technical Requirements

RAN Architecture Evolution

Packet Based RAN Concepts

Cisco Solution Components

Design for Packet-Based RAN Summary

7/29/2019 Brkagg-1000 c2 Web

3/75


Market Drivers for Next-Gen RAN

7/29/2019 Brkagg-1000 c2 Web

4/75


Evolution and DisruptionTransitioning to the Mobile Internet

B

usinessPerform

ance

Mobile Access Evolution and IP Infrastructure Impact

TDMInfrastructure

IP InsertionVoice andData

Mobile Internet

BroadbandMobile

Voice Traffic Dominates

Mobile Data Dominates

Users/Sessions

Traffic

7/29/2019 Brkagg-1000 c2 Web

5/75


Mobile Internet Is Changing the Industry

Broadband: High speednetworks based on HSPA and

EV-DO are now available in

many geographies

Billing Plans: None of this

would be possible without

aggressive flat rate all-you-can-

eat billing plans

Handsets: Powerful new deviceswith compelling UIs (iPhone,

Instinct, BB Bold, SE Xperia, etc.)

Apps: Lots of compelling apps

are moving over from the wired

world to join emerging LBS

services
http://maps.google.com/maps

7/29/2019 Brkagg-1000 c2 Web

6/75


Shifting the fundamental basis upon whichtelecom networks are designed so they are

optimized for carrying Packet based data ratherthan circuit based voice.

7/29/2019 Brkagg-1000 c2 Web

7/75 2009 Cisco Systems, Inc. All rights reserved. Cisco PublicBRKAGG-1000_c2 7

Opportunities with Packet Based RAN

Reduce Operational Cost Backhaul of cell tower traffic and

leasing T1s account for 20% ofmobile operator OpEx

Drive down per bit cost inexponentially

IP Based Converged Transport 2G networks use TDM circuits for RAN

transport

3G (UMTS) networks use ATM for RANtransport

4G is all IP

Service delivery over any access network

RAN Backhaul Scalability

Easier addition of new RAN bandwidth

Rollout new services faster

Meet capacity demand, expected to grow 4xto 10x as migration to 3G and 4G proceeds

LTE will drive 100Mbps 1Gbps per cell-site

7/29/2019 Brkagg-1000 c2 Web


Technical Requirements

7/29/2019 Brkagg-1000 c2 Web


Next-Gen Backhaul Requirements

Common and cheap transport

Generation and service independent

Traffic type awareness andprioritization (QoS)

Scalability

Service Resiliency

Clock distribution mechanism

Large scale provisioning and network visibility

Work with existing backhaul interfaces(T1/ATM/Sonet)

7/29/2019 Brkagg-1000 c2 Web


Mobile Operators Looking for Options

Convergence over ATM

RAN Optimization, with HSPA Offload

Microwave

Ethernet based BTS / Node-B

IP/MPLS based transport

Winner: IP/MPLS based transport

7/29/2019 Brkagg-1000 c2 Web


Convergence Over ATM

Aggregate 2G/3G traffic usingsingle ATM access link

Incremental deployment of 3Gwith existing 2G

Not flexible enough to deliverstatistical / bursty traffic

Cost per mobile phone increasessignificantly faster than ARPU

Multicast not easy

Not future proof

Aggregate traffic from 2G/2.5GBTS or 3G Node-B on a single

ATM trunk

Cell-Site

Mobile Core

ATMAggregation

E3/T3STM-1/OC-3E1/T1 TDM

E1/T1 ATM

NodeB

NodeB

BTS

Assemble TDMdata to ATM cellE1/T1 ATM

7/29/2019 Brkagg-1000 c2 Web


RAN-Optimization with HSPA Offload

Optimization by suppressingsilence/repetitive frames,compressing headers

Data offloading to DSL while2G and 3G voice still over

T1/E1

Temporary solution, Not futureproof

Reduction in voice quality

Not necessarily standardsbased

UMTS Voice andSignaling Path

GSM/GPRS/EDGE path

HSDPA andUMTS DataOffload

Cell Site

BTS

Node-B

RNC

BSC

Mobile

Core

Optimized GSM and UMTS RANBackhaul: Abis + Lub Over IP

T1/E1

BroadbandIP Backhaul

50% efficiency gain on GSM,15-90% on UMTS

HSPA offloaded to DSL, Eth etc

7/29/2019 Brkagg-1000 c2 Web


Microwave

Point to multipoint microwave radio links

On demand bandwidth allocation for Node-Bs

Nodal concept simplifies the end to endprovisioning

Geography based limitations (Line of sight)

Spectrum / license availability

Requires contract renegotiations / newpermits in buildings

Cheap until 16 E1 then cost goes upsignificantly

7/29/2019 Brkagg-1000 c2 Web


Ethernet Enabled NodeB

Makes data offloading easier

For voice traffic, NodeB must originate PWE

In most cases, basic ethernet connectivitynot sufficient for end-to-end reliable transport

Not necessarily standards based

RAN vendors have no MPLS legacy

Provisioning / troubleshooting MPLSadvanced features on NodeB is a challenge

Subject to inherent security risks of IP / MPLS

7/29/2019 Brkagg-1000 c2 Web


IP/MPLS Based Transport

High capacity packet network

Access Agnostic

Unified transport

Widely deployed

Ethernet to cell site results in even more costsavings

Operational experience with their existingIP/MPLS core

Proven QoS, high availability and security

Clock synchronization over packet network isrelatively new

7/29/2019 Brkagg-1000 c2 Web


RAN Architecture Evolution

7/29/2019 Brkagg-1000 c2 Web


BTS

SONETSDH

ADM

T1/E1

Cell site Aggregation site

BSC

MSC

PSTN

Air interface IP/MPLS and TDM core

G-MSC

RAN Core

Core site

RAN Edge

BTS ADM

T1/E1

BSC

RAN Architecture with 2G TDM Voice
http://images.google.co.uk/imgres?imgurl=http://www.theregister.co.uk/2005/06/10/nokia6680.jpg&imgrefurl=http://www.theregister.co.uk/2005/06/10/review_nokia_6680/&h=475&w=240&sz=45&tbnid=s890ZHqY2wsJ:&tbnh=126&tbnw=63&hl=en&start=3&prev=/images%3Fq%3D3g%2Bphone%26svnum%3D10%26hl%3Den%26lr%3Dlang_enhttp://images.google.co.uk/imgres?imgurl=http://www.theregister.co.uk/2005/06/10/nokia6680.jpg&imgrefurl=http://www.theregister.co.uk/2005/06/10/review_nokia_6680/&h=475&w=240&sz=45&tbnid=s890ZHqY2wsJ:&tbnh=126&tbnw=63&hl=en&start=3&prev=/images%3Fq%3D3g%2Bphone%26svnum%3D10%26hl%3Den%26lr%3Dlang_en

7/29/2019 Brkagg-1000 c2 Web


BTS

SONETSDH

ADM

T1/E1


BSC

MSC

PSTN


G-MSC

RAN Core

Core site

RAN Edge

BTS ADM

T1/E1

BSCIP/MPLS

SGSNGGSN

Internet

Frame Relay

2.5G Adds GPRS Data

7/29/2019 Brkagg-1000 c2 Web

19/75 2009 Cisco Systems, Inc. All rights reserved. Cisco PublicBRKAGG-1000_c219

IP/MPLS

BTS

SONETSDH

ADM

T1/E1


BSC

nxE1

MSC

SGSNGGSN

PSTN


G-MSC

Internet

Node B RNC

MGW

RAN Core

Core site

RAN Edge

ATM

BTS ADM

T1/E1

BSC

nxE1

Node B RNC

STM1/OC3

STM1/OC3

UMTS Adds ATM RAN

7/29/2019 Brkagg-1000 c2 Web


IP/MPLS

BTS

SONETSDH

ADM

T1/E1


BSC

nxE1

MSC

SGSNGGSN

PSTN


G-MSC

Internet

Node B RNC

MGW

RAN Core

Core site

RAN Edge

BTS ADM

T1/E1

BSC

nxE1

Node B RNC

ATMoMPLS

STM1/OC3

STM1/OC3

ATMoMPLS 3G voice and dataTDMoMPLS 2G voice and data

ATM Pseudowires in RAN Core

7/29/2019 Brkagg-1000 c2 Web


IP/MPLS

BTS

Pseudo wire

T1/E1


BSC

SGSN

GGSN

PSTN


G-MSC

Internet

Node B RNC

MGW

RAN Core

Core site

RAN Edge

BTS

T1/E1

BSC

Node B RNC

MGW

MSS

ATMoMPLS

ATMoMPLS 3G voice and dataTDMoMPLS 2G voice and data

Converged IP Backbone

7/29/2019 Brkagg-1000 c2 Web


Deployment Scenarios

IP/MPLS

Core Sites

MSC MSC

STM1/E1

STM1/E1

SGSN IP/MPLS

RAN CoreAggregation

IP/MPLS

Agg Sites

RNC

Cisco7600

BSCATM

Cisco7600

Pre-Agg Sites

E1 IMAClear channel STM1Channelized STM1

Iub-cs

ATM SwitchReplacement

E1Channelized STM1

Clear STM1

E1Chan. STM1

Iub-ps

TDM

Inter-MSC

Transport

BTS

T1/E1

Node B

Cellsite

7/29/2019 Brkagg-1000 c2 Web


Packet-Based RAN Concepts

7/29/2019 Brkagg-1000 c2 Web


Circuit Emulation Over Packet (CEoP)

Circuit Emulation = imitation of a physical communication link

CEoP imitates a physical communication link across Packet network

Allows the transport of any type of communication over Packet

Ideal for TDM or Leased Line replacement and legacy networkconsolidation

PacketSwitchedNetwork

TDM/ATM Circuits(ChSTM1/OC3,T1/E1 etc.)

TDM/ATM Circuits(ChSTM1/OC3,T1/E1 etc.)

Standards based CEoP

7/29/2019 Brkagg-1000 c2 Web

25/75


Pseudowire Types Used in RANTransport

ATM pseudowire

Used for 3G only

Inefficient for a single cell but only sends traffic when required

Use of cell packing can reduce overhead with minimal impact on latency

TDM pseudowire

Used for 2G; can be used for 3G

Just as a real TDM circuit, bandwidth is wasted when the circuit is not

being fully utilized.

For 3G networks an ATM pseudowire offers an advantage over aTDM pseudowire

7/29/2019 Brkagg-1000 c2 Web

26/75


Pseudowire Basics

Pseudowire (PW): A mechanism that carries the essential elements of an emulatedservice from one Device to one or more other Devices over a Packet SwitchedNetwork (PSN).

Within the context of PWE3, this uses IP or MPLS network as the mechanism forpacket forwarding.

Having a common PW layer provides the simplification of deployment,

management and provisioning.

Industry has GOOD experience deploying some of these PW types already, andthe concept now can be extended to TDM & ATM for RAN purpose.

BSC/ RNC

MPLS

Attachment CircuitsAttachment Circuits

Pseudo-Wire

ATMoMPLS

TDMoMPLS

Node B/ BTS

TDMoMPLS either SAToP or CESoPSN

SAToP : Structured Agnostic TDM over Packet : draft-ietf-pwe3-satop-05.txt , RFC-4553

CESoPSN : Circuit Emulation Services over Packet Switched Network : draft-ietf-pwe3-cesopsn-07.txt

7/29/2019 Brkagg-1000 c2 Web

27/75


SAToP Standards

RFC 4553: Structure-Agnostic Time Division Multiplexing (TDM) overPacket (SAToP)

This specification describes edge-to-edge emulation of the followingTDM services described in [G.702]:

E1 (2048 kbit/s)

T1 (1544 kbit/s)

E3 (34368 kbit/s)

T3 (44736 kbit/s)

The protocol used for emulation of these services does not dependon the method in which attachment circuits are delivered to the PEs.

7/29/2019 Brkagg-1000 c2 Web

28/75


CESoPSN Standard

CESoPSN protocol designed to meet the following constrains:

Fixed amount of TDM data per packet: All the packets belonging to a

given CESoPSN PW MUST carry the same amount of TDM data.

Fixed end-to-end delay: CESoPSN implementations SHOULD provide thesame end-to-end delay between a given pair of CEs regardless of the bit-rate of the emulated service.

Packetization latency range:

SHOULD support packetization latencies in the range 1 to 5 milliseconds

Configurable packetization latency MUST allow granularity of 125 microseconds

Common data path for services with and without CE application signaling.

Structure-aware TDM Circuit Emulation Service over Packet SwitchedNetwork (CESoPSN), draft-ietf-pwe3-cesopsn-06.txt

7/29/2019 Brkagg-1000 c2 Web

29/75


CEM Group Mapping to PW7

6

5

4

3

2

1

19

20

8

9

10

11

12

14

13

21

22

23

18

17

16

15

24

7

8

9

10

11

6

5

4

3

2

1

12

19

20

14

13

21

22

23

18

17

16

15

24

8

9

10

11

6

5

4

3

2

1

12

19

20

14

13

21

22

23

18

17

16

15

24

MPLS

controller t1 1/0/0

cem-group 2 timeslots 1-6cem-group 3 timeslots 15-18, 20-24

interface cem 1/0/0cem 2

xconnect 10.0.0.1 42 encapsulation mplscem 3

xconnect 11.0.0.1 33 encapsulation mpls

11.0.0.1

10.0.0.1

CEM = Circuit Emulation issimilar to Channel-Group.It identifies a group of DS0sor Clear-ChannelT1/E1/T3/E3The data in a CEM group canbe transported from one

router to another using aPseudo-wire.

T1 [NxDS0]

T1 [NxDS0]

T1 [NxDS0]

7600CEoP7

QoS

Ingress Classification :class-defaultIngress Marking :

set mpls exp imposition

7/29/2019 Brkagg-1000 c2 Web

30/75


Pseudowires in RAN Transport

MPLS

Attachment Circuit Attachment CircuitPseudo-Wire

Channelized T1/E1 to NxDS0

Channelized T3 to T1, NxDS0Channelized OC-3 to T1/E1, NxDS0

ClearChannel T1/E1/T3 ClearChannel T1/E1/T3

T1/E1 ATM IMA T1/E1 ATM IMA

ClearChannel T1/E1 ATMClearChannel T3 ATM

Channelized OC-3 to T1/E1 ATM

ATM PWE3Local Switching

Layer 3 IPv4

CESoPSNLocal Switching [Future]

SAToPLocal Switching [Future]

29417600

7600

ClearChannel T1/E1 ATMClearChannel T3 ATMChannelized OC-3 to T1/E1 ATM

Channelized T1/E1 to NxDS0

Channelized T3 to T1, NxDS0Channelized OC-3 to T1/E1, NxDS0

CEM Circuit CEM Circuit

T1 Data T1 Data

T1 DataControlMPLSMPLS

Targeted LDP Session

SAToP : Structured Agnostic TDM over Packet : draft-ietf-pwe3-satop-05.txt , RFC-4553

CESoP : Circuit Emulation Service over Packet : draft-ietf-pwe3-cesopsn-06.txt

IMA : Inverse Multiplex over ATM

7/29/2019 Brkagg-1000 c2 Web

31/75



Circuit Emulation for 2G

7/29/2019 Brkagg-1000 c2 Web

32/75


MPLS Core: Pseudo-Wire Signalling

C1

C2

xconnect

xconnect

Based on xconnect command, both PEs will create

directed LDP session if doesnt exist already

PE1

PE2

Directed LDP

7/29/2019 Brkagg-1000 c2 Web

33/75


MPLS Core: VC Label Distribution

VC1

VC2

xconnect

xconnect

PE1

PE2

NH: PE2

VC: VCID

Label: BCircuit type: CEM

NH: PE1

VC: VCID

Label: A

Circuit type: CEM

CEM = SAToP E1, T1, E3, T3,CESoPSN basic, CESoPSN TDM with CAS

VC Label distributed through directed LDP session

FEC TLV tells the circuit type

7/29/2019 Brkagg-1000 c2 Web

34/75


LDP: Pseudo-Wire id FEC TLV

VC TLV = 128 or 0x80

VC Type: 0x0011 E1 (SaToP)

0x0012 T1 (SaToP)

0x0013 E3 (SaToP)

0x0014 T3 (SaToP)

0x0015 CESoPSN basic mode0x0017 CESoPSN TDM with CAS

C: 1 control word present

Group ID: If for a group of VC, useful to withdraw many labels at once

VC ID : ID for the transported L2 vc

Int. Param: classical + IETF-PWE3-TDM-CP-Extension

VC TLV C VC Type VC info length

Group ID

VC ID

Interface Parameter

7/29/2019 Brkagg-1000 c2 Web

35/75


Generic CEoP Frame Format

IETF draft-ietf-pwe3-cesopsn-xx.txt: Structure-aware TDM Circuit Emulation Service overPacket Switched Network (CESoPSN)

RFC4553: Structure-Agnostic TDM over Packet (SAToP)

Encapsulation header

CE Control (4 Bytes)

RTP (optional 12B)

CEoPPayload

Frame#1

Timeslots 1-N

Frame#2

Timeslots 1-N

Frame#3

Timeslots 1-N

Frame#m

Timeslots 1-N

Encapsulation header


RTP (optional 12B)

CEoPPayload

Bytes 1-N

Unstructured mode (SATOP) sends bytes out as

they arrive on TDM line. Bytes do not have to be

aligned with any framing.

Structured mode (CESoPSN) identifies framing and

sends only payload. It can be T1s from DS3 as well

as DS0s from T1. DS0s can be bundled to the same

packet.

7/29/2019 Brkagg-1000 c2 Web

36/75


Generic CEoP Frame Format

IP (20 Bytes)

UDP (optional 8B)

RTP (optional 12B)


CEoPPayload

Frame#1

Timeslots 1-N

Frame#2

Timeslots 1-N

Frame#3

Timeslots 1-N

Frame#4

Timeslots 1-N

PSN Label (4 Bytes)

PW Label (4 Bytes)


RTP (optional 12B)

CEoPPayload

Frame#1

Timeslots 1-N

Frame#2

Timeslots 1-N

Frame#3

Timeslots 1-N

Frame#4

Timeslots 1-N

UDP/IPv4 MPLSIP (20 Bytes)

UDP (optional 8B)

Session ID (4 Bytes)

Cookie (optional, max 8B)


CEoPPayload

Frame#1

Timeslots 1-N

Frame#2

Timeslots 1-N

Frame#3

Timeslots 1-N

Frame#4

Timeslots 1-N

L2TPv3/IPv4

7/29/2019 Brkagg-1000 c2 Web

37/75


TDM Over L3 or L2 Core

MPLS

Attachment Circuit Attachment CircuitPseudo-Wire

2941-PE7600-PE

7600-PE

CEM Circuit CEM Circuit

T1 Data T1 Data


P

P

LDP LDP LDP

IGP IGP IGP

Ethernet / ATM

2941-PE7600-PE

7600-PET1 Data T1 Data

L2 SwitchL2 Switch

LDP

IGP


T1 DataControlMPLSMPLSL2

T1 DataRTPMPLSMPLS ControlL2

T1 DataControlMPLS802.1qL2

L3 Core

L2 Core

7/29/2019 Brkagg-1000 c2 Web

38/75



ATM Emulation for 3G

7/29/2019 Brkagg-1000 c2 Web

39/75


ATM / IMA Over Psuedowire

IMA terminated on Cell-siterouter.

ATM psuedowire between cell-site and aggregation router.

Aggregation router can map

VCs from psuedowire to ATMOC3 Clear Channel towardsRNC.

ATM VC mode allows VPI and

VCI rewrite.

ATM VP mode allows VPIrewrite.

39

MPLS / IPMPLS / IP

ATM / IMA

ATM / IMA

ATM / OC3c

MWR2941

MWR2941

Cisco 7600aggregation

Node-B

Node-B

RNC

MPLSL2 ATMMPLS ControlMPLSL2 ATMMPLS Control

7/29/2019 Brkagg-1000 c2 Web

40/75


High Level ATM Features

ATMoMPLS VC and VP

Single Cell Mode

Packed Cell Mode

ATM IMA Routed PVC and SVC

Local Switching

QoS UNI

AAL5

7/29/2019 Brkagg-1000 c2 Web

41/75


ATM and IMA QoS

Ingress Classification match atm clp

Ingress Marking set mpls exp imp, set prec/dscp

Ingress Policing 2r3c, set mpls exp imp, set prec/dscp

Egress Classification match prec/dscp, mpls exp

Egress Marking set prec/dscp, atm clp, mpls exp

Egress Policing 2r3c, set atm clp, set prec/dscp, set mpls exptop

Egress Traffic Management UBR, UBR+, VBR, VBR-nrt, CBR

Egress VP Shaping Supported

Egress VC Shaping Supported

7/29/2019 Brkagg-1000 c2 Web

42/75



Native IP for 4G

7/29/2019 Brkagg-1000 c2 Web

43/75


Native IP Over MPLS

43

MPLS / IPMPLS / IP

IP

IP

IP

MWR2941

MWR2941

Cisco 7600aggregation

eNodeB

eNodeB

MME

SGSN

IP

Pure IP routing from eNode-B toMME/SGSN in the mobile core.

Utilize MPLS/IP core

Leased Eth or Own-built

Efficient to operate, avoidsrouting in the entire core

7/29/2019 Brkagg-1000 c2 Web

44/75



Clocking

7/29/2019 Brkagg-1000 c2 Web

45/75


Why Is Clocking Important?

Data

Data

Reference A

Reference B

1 11

1 1

0 00

1 00 1

Reference ClocksOut of Sync

Interpretation A

Interpretation B

7/29/2019 Brkagg-1000 c2 Web

46/75


Clock Recovery

Clock recovery is required for TDM emulation as receiver issupposed to run with same clock as source.

Three principal methods to recover the TDM service clock:

Synchronous Reference clock at TDM systems or IWF

Cell-site and aggregation devices receive clock from externalsources e.g. BITS, Sonet/T1, GPS

Adaptive methods

Clock is derived based on packet arrival rates

Differential methods

Cell site and Aggregation routers have the same clock source. Inaddition, the TDM clocks are derived from differential information inRTP header of the packet with respect to the common clock

7/29/2019 Brkagg-1000 c2 Web

47/75


Clock Sync on Packet Network

TDM

SynchronizationNetwork

PRC/PRS-traceable



PRC/PRS-traceable


PRC/PRS-traceable



PRC/PRS-traceable


TDM

IWF

packet

TDM

IWF

packet

TDM

IWFpack

et TDM

IWFpack

et

The two PRS/PRCs mayalso originate from the

same source.

PRC/PRS

TDM TDM

IWFIWF IWFIWF


same source.Synchronization

Network

PRC/PRS



PRC/PRS


PRC/PRS



PRC/PRS



same source.

TDM TDM

IWFIWF IWFIWFPacket

SwitchedNetwork

TDMServiceClock

Differential Timing MessagesRecovered TDMtiming based onthe differentialtiming messages


PRC/PRS



PRC/PRS


PRC/PRS



PRC/PRS

TDM TDM

IWFIWF IWFIWFPacketSwitchedNetwork

TDMServiceClock

Recovered TDMtiming based onthe adaptiveclock recovery

Adaptive Clock Recovery

Synchronous, Refat End Systems

Synchronous, Refat IWF

Differential

Adaptive

Clock Recovery Using Clocking

7/29/2019 Brkagg-1000 c2 Web

48/75


E1/T1CEOP SPA

Clock Recovery Using ClockingPseudo-Wires

E1/T1

CEOP SPA

BITS

E1/T1

CEOP SPA

STM-1ATM SPA

E1/T1CEOP SPA

E1/T1CEOP SPA

STM-1ATM SPA

RNC

Aggregation 7600MWR2941

NodeB

NodeB

CEOP DataPW CEOP Primary SyncPW (adaptive or differential)

MPLS

PWs carrying Out-of-band Clock.

These PWs do not carry data.

This network contains

example for 3G Backhaul.

E1/T1CEOP SPA BITS

CEOP Backup SyncPW (adaptive or differential)

7/29/2019 Brkagg-1000 c2 Web

49/75


Synchronous Ethernet (PHY Layer)

Equivalent to SDH/SONET Synchronization Architecture

Enable to maintain ITU-T G.803 synchronization chain (clause 8.2.4)

As SDH/SONET SyncE is a Physical Layer Synchronization method.G.8261 defines Synchronous Ethernet clock performance limits.

Extend previous ITU-T (and Telcordia) node clock recommendations

G.8262 defines synchronous Ethernet Equipment Clock (EEC)

Ethernet Slow Protocol to extend the SSM traceability function

G.8264 defines ESMC (Ethernet Synchronization Messaging Channel) to support SSM

ITU-T G.8262

(EEC) Node

SONET/SDH PHY SyncE

BITS/SSUPRC/PRS BITS/SSU

PHY SyncE

ITU-T G.8262

(EEC) Node

ITU-T G.8262

(EEC) Node

ITU-T G.8262

(EEC) Node

7/29/2019 Brkagg-1000 c2 Web

50/75


IEEE 1588-2008 (PTPv2) In A Nutshell

IEEE Std 1588-2008 is actually a toolbox.

The protocol can use various encapsulations, transmission modes,messages, parameters and parameter values

Multiple Clocks are defined: OC (slave/master), BC, TC P2P, TC

E2E, with specific functions and possible implementations. IEEE 1588-2008 added the concept of PTP profile.

Every standard organization can define its own profile(s) using a subsetof the IEEE 1588-2008 protocol.

For telecom operators, saying IEEE1588 support is not sufficient

information.

Node characterization, interoperability, performance and metrics

7/29/2019 Brkagg-1000 c2 Web

51/75



Service Resiliency

7/29/2019 Brkagg-1000 c2 Web

52/75


Redundancy @ Box-Level

Cell-site router redundancy

Redundant Power Supply

Aggregation router redundancy

Redundant Power Supply

Redundant Supervisor

Non-Stop Forwarding(NSF/SSO)

Redundant line-cards

Redundant aggregation device(optional)

Node BNode B

(U-PE)

(N-PE)(P) RNC

(ATM

IMA)(Gig-E or POS)

CEOP

(Clear

STM-1)

ATM

(N-PE)

CEOP

ATM

BSC

(Channel

STM-1)

(P)

(U-PE)

(Gig-E or POS)

Node BNode B

(E1)

BTS

MPLS

7/29/2019 Brkagg-1000 c2 Web

53/75


Redundancy @ Link-Level

Cell-site router redundancy

Multiple links to BTS / Node-B

T1 (TDM or IMA)

Eth

Multiple links to MPLS Core

Load-balanced

Aggregation router redundancy

Multiple links to BSC / RNC

Sonet (APS)

Eth (STP / Routing)

Multiple links to MPLS Core

Load-balanced

Node BNode B

(U-PE)

(N-PE)(P) RNC

(ATM

IMA)(Gig-E or POS)

CEOP

(Clear

STM-1)

ATM

(N-PE)

CEOP

ATM

BSC

(Channel

STM-1)

(P)

(U-PE)

(Gig-E or POS)

Node BNode B

(E1)

BTS

MPLS

7/29/2019 Brkagg-1000 c2 Web

54/75


Redundancy @ PW-Level

Example setup: RNC and BSC are using MR-APS (traditional)

Primary PWE3 from NodeB (ATM) and BTS (TDM)

Backup PWE3 from NodeB (ATM) and BTS (TDM)

Force APS failover on RNC and BSC, MR-APS on Aggregation router

Node BNode B

(U-PE)

(N-PE)(P) RNC

(ATM

IMA)(Gig-E or POS)

CEOP

(Clear

STM-1)

ATM

(N-PE)

CEOP

ATM

BSC

(Channel

STM-1)

(P)

(U-PE)

(Gig-E or POS)

Node BNode B

(E1)

BTS

MPLS

PWE3 Redundancy: A Redundant L2 Connection both to the Active andBackup APS Interfaces on RNC and BSC

Can be used for redundancy of adaptive clocking.

Backup

Active

7/29/2019 Brkagg-1000 c2 Web

55/75


Redundancy in MPLS Core

MPLS Core:

TE Fast Re-Route (FRR) Link and Node

Tunnel selection

Well proven mechanisms

Leased or Built

Node BNode B

(U-PE)

(N-PE)(P) RNC

(ATM

IMA)(Gig-E or POS)

CEOP

(Clear

STM-1)

ATM

(N-PE)

CEOP

ATM

BSC

(Channel

STM-1)

(P)

(U-PE)

(Gig-E or POS)

Node BNode B

(E1)

BTS

MPLSLink Failure

Node Failure

7/29/2019 Brkagg-1000 c2 Web

56/75



QoS and Security

7/29/2019 Brkagg-1000 c2 Web

57/75


Why QoS?

Latency time taken for a packet to reach its destination

Jitter change in inter-packet latency within a stream over time i.e.variation of latency

Packet loss measure of packet loss between a source anddestination

QoS provides:

Congestion Avoidance

Congestion Management

Prioritize critical traffic over best-effort

Signaling and Clocking Voice Real-time Data

7/29/2019 Brkagg-1000 c2 Web

58/75


Factors Affecting End-to-End Latency

Packetization delay segment, sample, process data andconvert to packets

Serialization delay time taken to place bits of the packeton to the physical media

Processing delay time taken to accept packet, place iton the input queue, decide output interface, place it in theoutput queue

Propagation delay time taken to transmit the bits acrossthe physical media

Queuing delay how long the packet stays in the outputqueue before being sent out

FixedDelays

Variable

Delays

QoS addresses Queuing delayTE addresses propagation delay

Proactive Approach

7/29/2019 Brkagg-1000 c2 Web

59/75


Proactive ApproachMeasure Performance

Run IP SLA betweenthe cell-site and

Aggregation routers

Collect Latency, Jitterand Packet Loss

Source and Destination synced using NTP

T1 = origination timestamp

T2 = Arrival at destination timestamp

T3 = Departure (from destination) timestamp

P = (T3 T2), processing delay at destination

T4 = Arrival at source timestamp

RTT = (T4 T1 P), round trip time

T1

T2

T3

T4

Source DestinationSLA Responder

Use IP SLA

7/29/2019 Brkagg-1000 c2 Web

60/75


Security

Service Provider Best practices for box-level security:Lock-down VTYs, telnet

Disable unused services

Multiple bad password attempts

Protection from cell-site router hijack

ACLs on aggregation router

Control Plane Policing on aggregation router

Eavesdropping

3GPP has recommended using IPSEC security for signaling

7/29/2019 Brkagg-1000 c2 Web

61/75


Cisco Solution Components

7/29/2019 Brkagg-1000 c2 Web

62/75


Cisco MWR 2941-DC

MWR2941

Ciscos Latest MWR Series Product, 1RU

Six Built-In GE Ports (4 RJ-45, 2 SFP)

16 Built-In T1/E1 Ports

Expanded Capacity, 2 HWIC slots

Support for 2800/3800 HWICs

Multiple Industry Standard Clocking Options

IEEE 1588v2, Sync-E Master & Slave,Adaptive, Stratum 3, BITS Input, Sync T1

Operating Temp -10 to 55C

Key Applications

IP RAN: Flexible and efficient all-IP RANs,enable IP intelligence at cell-site

RAN Optimization: Optimize and reducebackhaul costs for 2G and 3G

Standards Based Pseudowire: Use IETFPWE3 to transport 2G, 3G and 4G wirelessnetworks over low-cost alternative networks

Most Com pact, Affordable

High Performance

Cell Site Rou terwi th

Features Enabled

7/29/2019 Brkagg-1000 c2 Web

63/75


Cisco 7600 Series

Already present in most mobile

operators RAN transport oraggregation location

Versatile and Feature Rich

Key Areas in RAN

Pre-Aggregation:Aggregate cell-sites, provide connectivity to mobilecore

Aggregation: Connectivity tomobile core, Inter-site connectivity

Transport Core: Used as PE or Pin MPLS RAN transport coreProven , Versati l e, High

Performance

Agg regat ion Routerwi th

Service Provider Features

Router Clock Synchronization Options

7/29/2019 Brkagg-1000 c2 Web

64/75


Router Clock Synchronization Options(7600)

Interface Clock Source Options [clock source internal | line]

A. Line : Use the clock input from the physical line.B. Internal Local : Use the clock input from the Oscillator on the Port Adapter or Line cardC. Internal Back-Plane : Use the clock from the back-plane.

Back-Plane Clock Source Options [network-clock-select interface | controller | slot | ...]

1. Controller : Map the clock from the controller to the Back-Plane. CEoPs SPA can input BITS clocking.2. Module : Map the clock from the Stratum-3 chip resident on SIP-200 or SIP-400 to the Back-Plane.

3. Interface : Map the clock from the interface, like, Sonet, Serial, to the Back-Plane

Back

Plane

Clock

Trace

Supervisor

SIP-400

SIP-200

SIP-400

T1 CEoPClock

Int POSClock

1

2

3SIP-400T1

CEoPInternal

ClockSource

LineClock

Source

CA

BITSClockInput

7600

SIP-400T1 CEoP

Clock

InternalLocalClock

Source

B

SIP-200OC3 POS

7/29/2019 Brkagg-1000 c2 Web

65/75


Circuit Emulation SPAs

High Level Features

Channelization of Low Speed Interfaces down to NxDS0

Circuit Emulation Function [CESoP and SAToP]

Configurable jitter buffer 1-500ms (+/- 250 ms )

Clock Synchronization

BITS Clocking

ATM and ATM IMA Function [ATMoMPLS, Layer 3]

SPA Name SPA Description SPA Height

SPA-24CHT1-CE-ATM

24 port Channelized T1/E1/J1 ATM & Circuit Emulation

SPA Single-Height

SPA-2CHT3-CE-ATM

2-port Channelized T3/E3 ATM & Circuit Emulation SPA

[E3 in Future] Single-Height

SPA-1CHOC3-CE-ATM

1-port Channelized OC3/STM-1 ATM & Circuit Emulation

SPA Single-Height

7/29/2019 Brkagg-1000 c2 Web

66/75


Channelization and Scalability for CEM

SPA Name

Channelization for

CEM Groups

DS0

Scale

Max Num

of CEM

groups*

SPA-24CHT1E1-CE-

ATM

Clear-Channel T1/E1

Channelize T1/E1 to

NxDS0 576 192

SPA-2CHT3E3-CE-

ATM

Clear-Channel T3/E3

Channelize to T1/E1

Channelize to NxDS0 1344 575

SPA-1CHOC3-CE-

ATM

Channelized to T1/E1

Channelized to

NxDS0 2016 575

STM-1STM-1 155.52 Mb/s

TUG-3 #2TUG-3 #2 TUG-3 #3TUG-3 #3

NxDS0

E-1

#1

E-1

#3

E-1

#19

E-1

#21

TUG-2 #1 TUG-2 #7

VC-12

VC-12

VC-12

VC-12

VC-12

VC-12

NxDS0 NxDS0

TUG-3 #1

AU-4AU-4

NxDS0

OC-3 155.52 Mb/s

STS-1 #2STS-1 #2

VTG #1VTG #1 VTG #7VTG #7

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

VT1.5

STS-1 #3STS-1 #3STS-1 #1STS-1 #1

DS-1

#1..DS-1

#4

NxDS0 NxDS0

DS-1

#25..DS-1

#28

NxDS0 NxDS0

DS-1

#1..DS-1

#4

NxDS0 NxDS0

DS-1

#25..DS-1

#28

NxDS0 NxDS0

DS-1

#25..DS-1

#28

NxDS0 NxDS0

* Per SPA

7/29/2019 Brkagg-1000 c2 Web

67/75


Scalability for ATM Psuedowires

Maximum Number of ATM PVCs per CEoP SPA 2,000

Maximum Number of ATM PVCs per Clear Channel SPA 4,000

Maximum Number of ATM PVCs per SIP-400 8,000

Maximum Number of ATM PVCs per 7600 16,000

Maximum Number of ATMoMPLS PWs per 7600 16,000

Maximum Number of Local Switched ATM VCs per 7600 16,000

Maximum Number of Packed Cell ATMoMPLS PWs per SPA 2,000

Maximum Number of Packed Cell ATMoMPLS PWs per 7600 8,000

Maximum Number of Links per IMA Group 16

7/29/2019 Brkagg-1000 c2 Web

68/75


Design for Packet-Based RAN

7/29/2019 Brkagg-1000 c2 Web

69/75


Packet RAN Scenario

Short-haul Backhaul

Node-B connected via T1/IMA ATM Psuedowires for voice and data VCs (Eth)

ATM Psuedowire for data VCs (Eth)

Voice VCs on IMA (T1)

Node-B connected via Eth Routed (Eth or MLPPP on T1)

BTS connected via T1 TDM Psuedowires for voice and data VCs (Eth)

TDM cross-connect (T1)

RAN Optimization (T1)

Cell-site router

7/29/2019 Brkagg-1000 c2 Web

70/75


Overall Design Procedure

Calculate bandwidth requirements for the cell-site and aggregationlocation

Choose the right packet based RAN option / design

MPLS Core Leased or Built, customer dependant

Choose appropriate redundancy and connectivity between:Cell-site router and Node-B / BTS

Aggregation router and RNC / BSC

Routing protocol between aggregation and cell-site routers

Ensure clocking / clock recovery at every node

Ensure resiliency for every failure type link and node

Apply QoS and Security

7/29/2019 Brkagg-1000 c2 Web

71/75


Summary

7/29/2019 Brkagg-1000 c2 Web

72/75


Summary

Operators looking for cost-effective and scalablenext-gen RAN solution.

IP/MPLS based RAN provides a converged solutionfor the operators 2G, 3G and 4G networks.

Packet Based RAN trend continues with LTE andFemto.

Packet Based RAN = Proven Seamless Integration

with Macro + Auto-Provisioning / Self Optimizing

7/29/2019 Brkagg-1000 c2 Web

73/75


References

IP RANhttp://cisco.com/en/US/netsol/ns675/networking_solutions_solution

_category.html

Mobile Transport over Psuedowirehttp://cisco.com/en/US/netsol/ns732/networking_solutions_solution.html

Cisco Live 2009: BRKAGG-3000 (Frequency and TimeSynchronization in Packet Based Networks):http://www.ciscolive2009.com/

Complete Your Online
http://cisco.com/en/US/netsol/ns675/networking_solutions_solution_category.htmlhttp://cisco.com/en/US/netsol/ns675/networking_solutions_solution_category.htmlhttp://cisco.com/en/US/netsol/ns732/networking_solutions_solution.htmlhttp://cisco.com/en/US/netsol/ns732/networking_solutions_solution.htmlhttp://www.ciscolive2009.com/http://www.ciscolive2009.com/http://cisco.com/en/US/netsol/ns732/networking_solutions_solution.htmlhttp://cisco.com/en/US/netsol/ns732/networking_solutions_solution.htmlhttp://cisco.com/en/US/netsol/ns675/networking_solutions_solution_category.htmlhttp://cisco.com/en/US/netsol/ns675/networking_solutions_solution_category.html

7/29/2019 Brkagg-1000 c2 Web

74/75


pSession Evaluation

Give us your feedback and youcould win fabulous prizes.Winners announced daily.

Receive 20 Passport points foreach session evaluation youcomplete.

Complete your session evaluationonline now (open a browserthrough our wireless network toaccess our portal) or visit one ofthe Internet stations throughoutthe Convention Center.

Dont forget to activate yourCisco Live Virtual account for access to

all session material, communities, and

on-demand and live activities throughout

the year. Activate your account at the

Cisco booth in the World of Solutions or visit

www.ciscolive.com.
http://www.ciscolive.com/http://www.ciscolive.com/

7/29/2019 Brkagg-1000 c2 Web

75/75

brkagg-1000 c2 web

Documents