Cisco IP RAN · PDF fileCisco IP RAN Architecture Peter Gaspar ... RNC Dift Layer 3 IP IP IP IP Drift RNC Node B Enhanced Node B Base Station Base Station Today Direct I

Cisco IP RAN Architecture

Peter Gaspar ([email protected])

© 2008 Cisco Systems, Inc. All rights reserved. Cisco ConfidentialPresentation_ID 1

Agenda

• RAN Requirementsq

• All-IP RAN Designs

• Legacy RAN over IPLegacy RAN over IP

• Cisco Carrier Ethernet Architecture

• Summary• Summary

2

© 2008 Cisco Systems, Inc. All rights reserved. Cisco ConfidentialPresentation_ID

RAN RequirementsRAN Requirements


Driving New Challenges for SPs

MobileInternet

man

ce BroadbandMobile

Voice Traffic Dominates Traffic

Perf

orm

TDM I f t t

IP InsertionVoice and

DataRevenue

usin

ess Infrastructure

Users/SessionsB

Mobile Internet Dominates

Users/Sessions

4


Mobile Access Evolution and IP Infrastructure Impact

Radio evolution pathRadio evolution path

LTE: 100 Mbit/s DL

HSUPA: 5.8 Mbit/s UL

50 Mbit/s UL

WCDMA R99: 384 kbit/s

HSDPA: 14.4 Mbit/s DL

GPRS: 160 kbit/s

EDGE: 384 kbit/s

2002 2003 2004 2005 2006 2007 2008

5


3xE1 5xE1 8xE1 ???xE1

Some calculations

• 1GB quota = 7.5kbps in busy hour per user (10% of traffic in busy hour)

• 10GB = 75kbps in busy hour per user

• 200 users per NodeB – 15Mbps sustain throughput200 users per NodeB 15Mbps sustain throughput

• 7 NodeBs – Gigabit Ethernet needed (more than 100 Mbps)

• 75 NodeBs 10GE needed (more than 1 Gbps)

• Busy eNodeBs can generate up to 40 Gbps

6


What is next – Release 6,7,8What is next Release 6,7,8

GGSN SAE GW

CSN GW

AccessGW

SGSN

GW

MME

GW

ASN GW

GW

ServingRNC

IP

IP

Serving RNC

D ift

Layer 3

IP IP IP IP

Drift RNC

Node B

EnhancedNode B Base

StationBaseStation

Today Direct I-HSPA+ SAE/LTE WIMAX EV-DO

7


Today Direct tunnel Direct tunnel

S / ORevC

3GPP/WCDMA Evolution

LTE/System Architecture EvolutionyS1-c Base Station to MME interfaceMulti homed to multiple MME poolsMulti-homed to multiple MME poolsSCTP/IP based

S11 MME to SAE GWGTP c Version 2GTP-c Version 2

X2 inter base station interfaceSCTP/IP Signalling

SAE GW to PDN GWGTP or PMIP based macro mobility

8


GTP tunneling following handover S1-u Base Station to SAE GW

GTP-u base micro mobility

Future RequirementsBackhaul Security

1. IPsec ESP using IKEv2 MMEgcertificate based authentication

2. Tunnel mode IPsec being mandatory and transport

Xu

MMES1-MME

mandatory and transport mode being optionalLikely that transport mode used to protect X2 *reduced overhead and low traffic) 1

Security Layer 1

SAE GW

X2

and low traffic)

3. SeGW used to offload EPC and allow IPSec scaling

4 Protection optional on S1

Security Layer 2

SAE GW

4. Protection optional on S1-MME and S1-U

5. Port based authentication on cell site demarcation

9


LTE RAN Requirements

• Any-to-any connectivity

• Low delay needed between eNodeBs (handover)

• Security concepts may varyy p y y

• MME may need to be distributed (messages count, delay etc.), depends on applications, not that much on

bilitmobility

• Multicast for MBMS

10


Edge DistributionProblem DefinitionProblem Definition

• Enormous increase in mobile data

• Need for more cost efficient networks

• Video Content Delivery Networksy

• Increasing peer-to-peer traffic•IMS

•File sharing

•Internet applications (Skype)

•Machine-to-machine

11


Edge Distribution (Peer-to-peer)

RNC SGSN

Eliminate long runs for peer-to-peer trafficAllow offload of Internet traffic to cheaper transports

Node B

IP RAN (GTP)

IPGGSNNode B

RNC SGSN

IP RAN (GTP)Node B

Node B Core IPGGSN

Some GGSN features

12


For HSPA needs Direct Tunnel supportLTE model is similar

Lowcost Internetcan even be distributed to cell-site router

Edge Distribution (Content Networks)

RNC SGSN

Eliminate repeating video traffic in the transport network

Node B

IP RAN (GTP)

IP

Content

GGSN

Content Engine

RNC SGSN

IP RAN (GTP)

Node B

IP

Content

GGSN

Root

13


Engine Content Engine

For HSPA needs Direct Tunnel supportLTE model is similar

Characteristics of the future RAN

• Increasing trafficE d t d IP h• End-to-end IP approach

• Components with Ethernet interfacesDi t ti b t N d B d th A• Direct connection between NodeB and the Access Gateway (Direct Tunnel)

• LTE specifics (any-to-any, multicast, security)LTE specifics (any to any, multicast, security)• Distribution of Edge

• IP/MPLS or Ethernet aggregation are the suitable technologies

14


All IP RAN DesignsAll-IP RAN Designs


Specifics of the mobile operators transport networktransport network

• Network provider is also the customer of the network, therefore he needs to take care of all services:

•IP Routing

O SLA (O&M Q S t )•Own SLAs (O&M, QoS etc.)

•Multicast

•SecuritySecurity

•ATM and TDM services for 2G and legacy 3G

• Layer 2 and Layer 3 services are necessaryLayer 2 and Layer 3 services are necessary

• This leads to optimized design where for example the CE and the PE functions may be combined

16


Contradicting Characteristics of RAN Aggregationgg g

• Scaling of services•Number of VPNs limited

•Number of VLANs and MACs limited

•Low number of Queues needed despite H-QoS•Low number of Queues needed despite H-QoS

• Scaling of ports•Low number of end devices on cellsite

•Limited number of rings in aggregation and pre-aggregation

• But, still needed•High bandwidth needed

•Carrier grade architecture (reliability, redundancy etc.)

•Hardened devices

17


RAN Transport HierarchyBTS

NxT1/E1

IMA

BSCRNC

MSC

Node B

IMA

SGSN

IP/MPLSL3VPN

GGSN

MSC

L3VPN

Eth tEthernetNodeB Ethernet

NodeBAccess Aggregation Core

Higher capacitiesRedundancies

Last-mileSmall aggregation sites

18


RedundanciesPartially meshed interconnectionsDifferent transport technologiesOften includes wireline services

Small aggregation sitesMostly Microwave transportLimited traffic volumesNo redundancies or ring

Aggregation TechnologiesMPLS PBB-TE

(802.1ah, 802.1Qay)MPLS-TP

Multiservice Yes (including Ethernet L2 only L2 onlyL3VPN, ATM, TDM)

Switching capacity High High High

Interoperability Yes Limited Limited

Transport Any Ethernet Only Any

Any-to-Any Yes No No

Multicast Yes No No

Core Interop Native L2 to L3 handover L2 to L3 handoverCore Interop Native L2 to L3 handover needed in Core

L2 to L3 handover needed in Core

Service distribution L3VPN,GGSNSAE/PDN

No No

19


Maturity Mature Early adoption Early adoption

All-IP RAN Aggregation - RSTP

Pros:+ Simple deployment

.1q/QinQRing

VLAN 100 802.1qtrunk

+ Simple deployment+ Compact IP addressing

VLAN 100 RNC

7600

802.1qtrunk

VLAN 100

7600802.1qtrunk

GGSN

VLAN 100

7600/MWRCons:- Suboptimal p-2-p- Slow convergence (STP)- Limited to Ethernet Aggregation

L b d t d i

20


- Large broadcast domains

Access Aggregation

All-IP RAN Aggregation – EVC/MPLS-TP

VLAN 100

Pros:+ Operational procedures as SDH+ Compact IP addressing

F t

.1q/QinQRing

VLAN 100

VLAN 101

VLAN 102

+ Fast convergence+ Flexible core transport+ Multiservice Aggr. (ATM, TDM etc.)

VLAN 1IP/MPLS RNC

7600

EoMPLS

7600VLAN 201

GGSN

7600/MWRVLAN 202 Cons:

- Suboptimal p-2-p- Large broadcast domains- Complex configuration

Challenging redundancy designs

21


- Challenging redundancy designs

Access Aggregation

All-IP RAN Aggregation - VPLSPros:+ Compact IP addressing+ Fast convergence

.1q/QinQRing

VLAN 100

+ Optimal p-2-p+ Flexible core transport+ Multiservice Aggr. (ATM, TDM etc.)

VLAN 100RAN VPLS RNC

7600

VLAN 100

IP/MPLS 7600 GGSN

VLAN 100

7600/MWR Cons:- Large broadcast domains- ES or Sip-400 needed

22


p

Access Aggregation

All-IP RAN Aggregation – L3VPNPros:+ Optimal p-2-p+ Fast convergence

S t d b d t d i

.1q/QinQRing

VLAN 100

10.1.1.0/24+ Separated broadcast domains+ Flexible core & access transport+ Multiservice Aggr. (ATM, TDM etc.)

RNC

7600

RAN VRF10.0.1.0/24

IP/MPLS 760010.2.1.0/30

GGSN10.0.2.0/24

7600/MWR10.2.1.4/30

Cons:- Complex IP Addressing

23


Access Aggregation

All-IP RAN Design Conclusions

• The Access is usually native Ethernet with REP if possiblepossible

• Aggregation is either native Ethernet or MPLS, whereby:y

•The question mostly is, where the line between MPLS and Ethernet should be made•MPLS is more flexible and multiservice•MPLS is more flexible and multiservice•Ethernet is more cost efficient

• Any-to-any connectivity between NodeBs will be moreAny to any connectivity between NodeBs will be more important in LTE than now

• Layer 2 or Layer 3 decision depends on operators preference and operational procedures

24


preference and operational procedures

Synchronization (Frequency)

Solutions:

1. GPS on each NodeBNot suitable for Europe

2. Synchronization over PseudowireCan be used directly between RNC and NodeB

3. IEEE1588v2Packet based synchronization distribution

4. ITU SyncELayer 1 based synchronous Ethernet

25


IEEE1588v2 in the RANCore Site

PDSN or SGS

Agg Site

7600with CEoP

ATM

SGSN

NxT1/E1

IMA

FRE1

with CEoP

MSC

FR

BTS BSC

Pseudowires

STM1

IMA C

BSCSTM-1Node B

RNC

Pseudowires

7600with CEoP

RNCEthernetNodeB

IP/MPLS

RNCPre-Agg Site

7600with CEoP

NodeBClock

SourceIEEE1588

MasterIEEE1588v2 Packets

Ethernet SwitchME3400

Ethernet

• Packet based

• One master, multiple slaves

26


EthernetNodeB Ethernet

NodeB

One master, multiple slaves

• Requires strict QoS in the network for IEEE1588 packets

• Supports frequency and time (phase)

ITU SyncE in the RANCore Site

PDSN or SGS

Agg Site

7600with CEoP

y

ATM

SGSN

NxT1/E1

IMA

FRE1

with CEoP

MSC

FR

BTS BSC

Pseudowires

STM1

IMA C

BSCSTM-1Node B

RNC

Pseudowires

7600with CEoP

RNCEthernetNodeB

SyncE SyncE

IP/MPLS

RNCPre-Agg Site

7600with CEoP

NodeB

ClockSource

SyncE

SyncE SyncE

SDH Sync

IP/MPLSEthernet Switch

With SyncE

Ethernet

• Layer1 based

• On point-to-point connections only

SyncE

y

27



NodeB

On point to point connections only

• Very precise but requires support of all ethernet interfaces involved in the path

• Does not support time synchronization

CombinationCore Site

PDSN or SGS

Agg Site

7600with CEoP

ATM

SGSN

NxT1/E1

IMA

FRE1

with CEoP

MSC

FR

BTS BSC

Pseudowires

STM1

IMA C

BSCSTM-1Node B

RNC

Pseudowires

7600with CEoP

RNCEthernetNodeB

SyncE SyncE

IP/MPLS

RNCPre-Agg Site

7600with CEoP

NodeB

ClockSource

SyncE

SDH Sync

IEEE1588v2 IP/MPLS

Ethernet

• Stable clock in the core

• Network element independent in access

Ethernet SwitchME3400 IEEE1588v2

28



NodeB

Network element independent in access

Legacy RANg yOver IP


Legacy and All-IP RANBTS

NxT1/E1

IMA

BSCRNC7600

MSC

Node B

IMA

SGSN

7600 IP/MPLSL3VPN

MWR GGSN

MSC

L3VPN

Ethernet

7600IP/MPLS


NodeB

30


Access Aggregation Core

Network ElementsAttachment Circuit Attachment CircuitPseudo-Wire

CEM Circuit CEM Circuit

Targeted LDP Session

MPLS

7600 7600T1 Data T1 Data

T1 DataControlMPLSMPLS

Channelized T1/E1 to NxDS0Channelized T3 to T1, NxDS0

Channelized OC-3 to T1/E1, NxDS0CESoPSN

Local Switching [Future]

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

ClearChannel T1/E1/T3 ClearChannel T1/E1/T3SAToPLocal Switching [Future]

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

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

31


SAToP : Structured Agnostic TDM over Packet : RFC-4553 CESoP : Circuit Emulation Service over Packet : RFC-4842IMA : Inverse Multiplex over ATM

Pseudowire in the RANRAN traffic groomed over MPLS Pseudowires for backhaul

MPLS needed in Aggregation

Reduced OPEX – Bandwidth Flexibility

Eliminates need for ADM and ATM switchesEliminates need for ADM and ATM switches“Flatten the network”, Simplify, Reduce costs

Pseudowire enables greater flexibility for traffic handlingg y gProactive scalability -- Self-Adjusting Backhaul Transport

Pre-provision new and future services

Provides Clock Recovery per 3GPP CESOPN Standards

LongevityCompliance with 3GPP/3GPP2 Reference Architectures up to R8

32


Compliance with 3GPP/3GPP2 Reference Architectures up to R8

Cisco Carrier Ethernet Architecture


Cisco’s Carrier Ethernet Approach

Carrier Ethernet AggregationAccess Edge Multiservice CoreCorporate

Business

BNG

Aggregation Node

Ethernet

VoD

Content Network

TV SIP

Distribution Node

STB

Residential

Aggregation Network

Business PEDSL

Core Core NetworkIP / MPLS

STB

Corporate

Business

Residential

2G/3G Node

Dark Fibre / CWDM / DWDM and ROADM

STB

Residential

Corporate

Business PON

Dark Fibre / CWDM / DWDM and ROADM

STB

IP/MPLS

34


IP/MPLSETHERNET

Cisco 7600Bridging the Solution

TDM to Packet MigrationT1 T3 OC Ch li d

Bridging the Solution

SAToPCES P

Mobile TransformationY 1731 CFM 8 1T1, T3, OCn, Channelized

ATM InterworkingMLPPP, IMAATM Cell Packing/Relay

CESoPAToM , MPLSoGRESyncE, 1588v2, AdaptiveBITS Clocking

Y.1731, CFM 8.1Multicast & HAMultisegment PseudowireIntegrated Routing + Bridging

Modular, Scalable40Gbps / Slot720Gbps Total Bandwidth

ManagementANAISC (Provisioning)

5 Chassis SizesPW3, BFD, Queues scaleVLAN, MAC scale

Customer Network Management (CNM)MIBs

Cisco 7600Cisco 7600

Carrier Class ResiliencyAPS, 802.3ad LACP, REP, BFDAccess Circuit Redundancy

Integrated ServicesL2 and L3 ServicesBusiness and Residential

76007600

35


yHot-standby PW RedundancyEnd-to-End OAMNSF/SSO, EFSU, ISSU

Business and ResidentialSession Border ControllerIntegrated SecurityMobile Services

Cisco ASR 9000 “At a Glance”

Optimized for Aggregation of Dense 10GE & 100GEDense 10GE & 100GE

Designed for Longevity & TCO: Scalable up to 400 Gbps of p pBandwidth per Slot

Based on IOS-XR for Nonstop Availability & Manageability

Enables Network Convergence of Business & Residential Services for Fixed & Mobile Access

36


Advanced Video DNA

Cisco MWR 2941-DCFCSFCSFebruary 16, 2009February 16, 2009Cisco MWR 2941 DC

1. Cisco’s Latest MWR Series Product2 Six Built In GE Ports (4 RJ 45 2 SFP)2. Six Built-In GE Ports (4 RJ-45, 2 SFP)3. 16 Built-In T1/E1 Ports, expandable to 244. Multiple Industry Standard Clocking

OptionsMWR2941MWR2941

IEEE 1588v2, Sync-E, Adaptive, Stratum 35. Expanded Capacity

Support for 2800/3800 HWICs6. Key Applications

IP RAN: Activate flexible and efficient all-IP RANs for new revenue-generating services with intelligent IP network features

RAN Optimization: Optimize and reduce backhaul costs for 2G (GSM) and 3G (UMTS/HSxPA) wireless networks

Standards Based Pseudowire: Use IETF

Most Compact, Affordable High Performance Cell Site Router with Features Enabled

Most Compact, Affordable High Performance Cell Site Router with Features Enabled

37


PWE3 to transport 2G, 3G and 4G wireless networks over low-cost alternative networks such as xDSL, Carrier Ethernet, MPLS, etc.

Features EnabledFeatures Enabled

Market Leading Mobile Network Solutions Positioning for Key Solution Segments

4G Ethernet Centric

New Ethernet/IP centric

Legacy 2G

T1/E1/ATM L2 Backhaul

Positioning for Key Solution SegmentsHybrid 2G/3G/4G

TDM/ATM hybrid with overlays/greenfieldHigh scale & densityCost Leadership Key Infrastructure

PWE3 Circuit EmulationNo Ethernet nodeBsACR, APS and TDM HA mechanisms

yEthernet/IPNew Ethernet / IP centric nodeBs/RNC deploymentInvestment protection Key Infrastructure

platform going forwardIP Gateways: 7600

P tf li

IP Gateways for Mobile EdgeUp to 10G Scale

Portfolio:

pIP Gateways for Mobile EdgeUp to 40G Scale

Portfolio: MWR2941 7600 Portfolio:

ASR9000, MWR2941

Portfolio: MWR2941, 7600

MWR2941, 7600

38


Network is the Platform

Cisco Carrier Ethernet - Mobile RAN ServicesMarket Transitions 1+2+3 or 1+2+4 or 3

UMTS ATM Node B GSM BTS

ATM VC, TDM (SATOP, CESoPSN)

E1 (w/ IMA)

ATM or TDM BSCATM RNC1

UMTS ATM Node B, GSM BTS

ATM or TDM S-PE, MS-PWBSCATM RNCThe BTS model may be overlaid on 3 fortactical sales reasons

Static or IGP Overlay on an MST/REP and not protected Ethernet Access Network (DSL, Ethernet P2P)

2

EthernetIP RNC, S-GW

MPLS/IP, MPLS VPNfor LTE IP RAN and UMTS IP RAN

Ethernet

REPThis model assumes GSM TDM infrastructure is used until GSM radio moved to UMTS or LTEMay coexist with 1 and integrate 2

3VPLS/HVPLS for UMTS

VPLS/HVPLS for UMTS

ME-3400E

Mobile RAN Edge Multiservice CoreEfficient Access Large Scale Aggregation

IP RNC, SAE

This model may integrate GSM TDM 4

REP MPLS/IP, MPLS VPNfor LTE IP RAN and UMTS IP RAN

29417600/ES+/CEoPS, SRD2

7600/ES+/CEoPSSRD2

Di t ib ti N d

MPLS enabled Cell Site ATM, TDM, Ethernet

NNIIATM or TDM or Ethernet NNII

2941CRS 3.0

SRD2(ASR09000 FCI)

SRD2(ASR9000 FCI)

39


Aggregation Node

MPLSMPLS / IPoDWDM

Distribution Node

2G/3G Cell Site

ATM, TDM, Ethernet Cell Site RNC or BSC

QOS Model - Downstream Distributed Business L2/L3 VPN ServicesDistributed Business L2/L3 VPN Services

Aggregation EdgeAccessPer subscriber service instance, hierarchical Default class basedShaping ,OQS with parent shaper and child queuing, policing and marking

AggregateDiffServ

Default class based queuing policy, to

minimize delay and jitter for Voice/Video

Queuing & scheduling

Marking

Policing

Shaped Rate = Access Line RateScheduling

Aggregation NodeAccess Interface

Business

Corporate

40


Aggregation Node Core NodeDistribution NodeCPE Access Node

QOS Model - Upstream Distributed Business L2/L3 VPN ServicesDistributed Business L2/L3 VPN Services

Aggregation CoreAccess

Shaping Ingress class-based policing

AggregateDiffServ

Per subscriber service instance

Ingress policing

Queuing & scheduling

Marking

Policing

based policing and marking per subscriber line

Scheduling

Aggregation NodeAccess Interface

Business

Corporate

41


Aggregation Node Core NodeAccess Node Distribution NodeCPE

Baseline Network Availability MechanismyIP Services:

• Fast IGP/BFD convergenceg

MPLS Services:• Pseudowire redundancy • MPLS TE FRR Link and Node protection with IP services PW/VPLS PW tunnel• MPLS TE-FRR Link and Node protection with IP services, PW/VPLS PW tunnel selection

Access:• Resilient Ethernet Protocol

Large ScaleAggregation

IntelligentEdge

MultiserviceCore

Efficient Access

Resilient Ethernet Protocol

Aggregation NodeAccess Node

BNG

Distribution Node

42


Aggregation NodeMSE

GTP, IP, MPLS MPLSMPLS / IPEthernet

Access Node Distribution Node

Carrier Ethernet Ring TopologiesCurrent Network Trends and ChallengesCurrent Network Trends and Challenges

• Large Spanning Tree domainsLarge Spanning Tree domainsIncreasing number of nodes in the ring

• Supporting higher number of subscribersIncreasing number of VLANs, MAC addresses per L2

domain

• Carrier Ethernet Trend• Carrier Ethernet TrendFast convergence requirement in the access and aggregation networksSpanning Tree not perceived as Carrier Class

• Complexity of management and troubleshooting as the network grows

43


the network grows

Resilient Ethernet ProtocolA segment ProtocolA segment Protocol

REP Segment

BlockedOpenAlternate Port

Link Failure

Edge Port Edge Port

• A REP segment is a chain of ports connected to each other and configured with a segment ID.

• One switch can have only two portsOne switch can have only two portsbelonging to the same segment.

• REP guarantees there is no connectivity between two edge ports

ton a segment. • When all interfaces in the segment

are UP, the alternate port is blockingWh li k it h f il

Blocked

44


• When a link or switch failure occurs on the segment, then blocked port goes forwarding REP Segment

REP Benefits

• Fast and predictable convergence – Convergence time: 50 to 250msConvergence time: 50 to 250ms– Fast failure notification even in large rings with high number of

node – Manual configuration for predictable failover behavior g p

•Co-existence with Spanning Tree– STP is deactivated on REP interfaces– Limit the scope of Spanning tree– Limit the scope of Spanning tree– Topology Changes Notification from REP to STP

• Optimal bandwidth utilizationVLAN L d b l i– VLAN Load balancing

• Easy to configure and troubleshoot– Topology archiving for easy troubleshooting

45


– Known fixed topology– Simple mechanism to setup the Alternate port (blocking)

Expansion on Demand IBTS

NxT1/E1

IMA

BSCRNC

MSC

Node B

IMA

IP/MPLSSGSN

IP/MPLSL3VPN

GGSN

MSC

L3VPN

Eth tEthernetNodeB Ethernet

NodeB

46



Expansion on Demand IIBTS

NxT1/E1

IMA

BSCRNC

MSC

Node B

IMA

SGSN

IP/MPLSL3VPN

GGSN

MSC

L3VPN

Eth t

IP/MPLS


NodeB

47



Expansion on Demand IIIBTS

NxT1/E1

IMA

BSCRNC

MSC

Node B

IMA

SGSN

IP/MPLSL3VPN

GGSN

MSC

L3VPN

Eth t

IP/MPLS

MultiserviceIndependent of the transportN ti IP (A t )Ethernet

NodeB EthernetNodeB

Native IP (Any to any)IPSec supportMulticast supportDistributed or centralizedSt d d b d

48


Access Aggregation CoreStandards based

From Ethernet to MPLS/IP

• Access based on simple Ethernet

• Aggregation based on MPLS/IP

• Larger access sites based L2 onlyg y• Migrate to L3 by license when needed

• Add TDM/ATM access ports when needed

• ML-PPP for TDM microwave to Ethernet NodeB

• Add MWR if needed on cell-site

• Expansion of Aggregation towards Access on as needed basis

49


50


SummarySummary


All IP RAN Summary

• Designs depend on operator’s needs

• MPLS in Aggregation is more flexible than L2 Aggregation

• Line between Access and Aggregation can move depending on traffic and operator’s structuretraffic and operator s structure

• Any-to-any connectivity may be required for LTE

• Circuit Emulation and ATMoMPLS for legacy RAN to reduce OPEX

Cisco Carrier Ethernet architecture supports various models and• Cisco Carrier Ethernet architecture supports various models and can evolve with the network

• Cisco certified for TDM transport by IP MPLS Forum

52


Cisco Networkers 25-28. januar 2010.BarselonaR i t jtRegistrujte se

53


54


Documents

Cisco IP RAN · PDF fileCisco IP RAN Architecture Peter Gaspar ... RNC Dift Layer 3 IP IP IP IP Drift RNC Node B Enhanced Node B Base Station Base Station Today Direct I