Step-by-Step Migration to an All-IP Mobile Backhaul Network

1
Step-by-Step Migration to an All-IP Mobile
Backhaul Network
Customers Event Crete, May 2007

• Presented by
• Issar Krausz
• Business Development senior Manager

2
Outline

• Backhaul at the Spotlight
• The Challenge
• Transport cost reduction
• Packet-based Transport
• Pseudowires
• New Transport Alternatives for Mobile Backhaul
• Transport for Fixed and Mobile Convergence

3
Cellular Backhaul From the Backstage to the
Spotlight
4
Its All About Customer ExperienceHigher Speed,
Richer Content
Internet Surfing
HSPA
Interactive Gaming
Video Conferencing
UMTS
Video Telephony
Video Streaming
EDGE
Audio Streaming
Video Download
Audio Download
GPRS
MMS
Internet
Rich call services
E
-
mail
Content to person
SMS
Messaging
40 Kb/sec
75 Kb/sec
2 Mb/sec
14 Mb/sec

• 1 Minute of audio from MP3 music file equals
• 132 seconds GPRS
• 22.4 seconds UMTS
• 4.1 seconds - HSDPA

5
The Rapid Bandwidth IncreaseNo. of E1 Links per
Cell-site

• Worldwide mobile subscriber base keeps increasing
  
• According to Infonetics October 2006 report,
  Mobile Backhaul Equipment, Installed Base,
  Services Market Outlook, there were more than 2
  billion mobile subscribers worldwide in 2005, and
  that number will jump to over 3 billion by 2009
• Enhanced user-experience demands higher-speed
  data rates
• Video streaming, MP3/MP4 downloads, MMS,
  peer-to-peer applications, Mobile TV
• NG mobile standards and services are within reach
• LTE 100Mbps

6
Cell-site Transmission Evolution
Cell Site
2G TDM 1-2 ? E1/T1
2G BTS
Transport Network PDH, SDH, LL
3G Node B

• New services more transport bandwidth right out
  of the cell-site
• Access becomes the BW bottleneck
• PDH/TDM is not a scalable solution
• HSDPA the killer application for mobile
  backhaul

7
HSDPAThe backhaul killer application
Source Unstrung Insider

• Mobile data traffic to double within the next
  12-18 months
• Data volumes in backhaul to surge to 3-4 times
  current levels within 2 years
• Mobile data ARPU will not keep the pace
• According to Strategic Analysis, Nov. 2006

8
The ChallengeOptimizing the Cellular Backhaul

• Challenge - Handling capacity growth
• Backhaul represents 70 of the transport costs
• About 30 of the OPEX (incl. site rental,
  maintenance, technical sup., etc)
• Evolution to future ready transmission All-IP
  and FMC
• Solution Cost reductions thru
• Abis/Ater BW optimization and data compression
• Sharing a single transport network by 2G/2.5G and
  3G traffic
• Statistical Multiplexing with traffic
  prioritization
• Iub optimization adapted to HSDPA sales
  growth/strategy
• Backhauling over alternative technologies and
  media
• Packet-based transport
• Using low cost xDSL access

9
Optimization of A-bis/A-ter Backhaul Links
BSC
A-bis Optimization
TDM/IP Transport Network
E1s
Vmux-400

• Reducing backhaul expenses by more than 50
• Reduce no. of E1 trunks for BTS-BSC connectivity
  with no voice quality degradation
• Smallest footprint device in the market save on
  co-location costs
• A-bis/A-ter traffic optimization for efficient
  use of backhaul links leased lines, satellite
  links and microwave
• Data Compression

10
Cut Backhaul Costs thru Aggregation
GSM BTS
N x E1 gt n x E1
2G BSC
E1/T1s TDM
n ? E1/T1 TDM Ch-STM1/OC-3
Cell-site Gateway
Multi-Service Aggregation
SDH/SONET/ATM Transport Network
E1/T1s ATM IMA
n x E1/T1 (IMA) STM-1/OC-3
STM-1/ OC-3
ACE-3x00
ACE-3x00
3G RNC
IP-Node B
Ethernet

• Substantial cost savings by
• Aggregation, Grooming, Overbooking/Oversubscriptio
  n
• Aggregation/grooming of GSM and UMTS traffic onto
  a unified transport
• Statistical multiplexing of UMTS/HSDPA traffic
  based on ATM capabilities
• Dynamic bandwidth allocation based on operators
  predefined priorities
• Flexible port configuration to support GSM and
  UMTS

11
What have we achieved so far?
Cell Site
2G TDM 1-2 ? E1/T1
3G ATM Voice 2-4 ? E1/T1
2G BTS
The traditional approach 8-16 E1/T1 for a full
service offering
Transport network
HSDPA up to 8 E1/T1
3G Node B
Savings of over 80 Million Euros/year
12
What else can be improved?

• There are two main alternatives being considered
  for additional backhaul cost reduction
• Using packet-switched transport
• Complement by using DSL access

13
Why Packet-based Transport?

• Ethernet presents the most scalable and flexible
  transport solution to meet the BW challenge right
  off the cell site
• The cost for leased lines is about 2.5 times
  that of new wireline connections (Ethernet,
  DSL, cable, PON)

14
Requirements for a Packet-Based Radio Access
Network

• Transport network performance SLA

Service category Delay (msec) Jitter (msec) Packet-loss () Packet size (Bytes) Packet rate (PPS)
Clock
R99
HSDPA

• Transport elements cell-site and aggregation
  gateways
• Stringent clocking/synchronization requirements
• G.823/824, traffic and sync. interface, G.8261
• Compatible with all types/generations of mobile
  traffic
• CDMA GSM/UMTS, All-IP
• Using circuit emulation and service emulation
  (HDLC ATM)
• Traffic optimization (TDM grooming, ATM
  aggregation)
• Comprehensive network management, OAM and
  performance monitoring
• Right-sized access/backhaul equipment for radio
  site and BSC/RNC location

15
Lesson in HistoryWhat are Pseudowires?

• Pseudowire is a mechanism that emulates the
  essential attributes of a telecommunications
  service over a packet-switched network
• TDM leased line, ATM or Frame Relay
• It defines a way to packetize and encapsulate
  data and create a logical link between two
  network entities
• Pseudowire technology is an enabler for circuit
  emulation and service emulation of legacy
  services over packet networks

PWE3 1
PWE3 2
Tunnel
CLE/CPE
CLE/CPE
TDM, ATM FR, etc.
Legacy Service
Pseudowire Legacy Emulated Service
Legacy Service
16
Pseudowire Standards
Standard Application IETF ITU-T MFA MEF
TDMoIP TDM circuit emulation ietf-pwe3-tdmoip Y.1413, Y.1453, Y.1414, Y.1452 IA 4.0 IA 5.0 MEF 8
CESoPSN TDM circuit emulation ietf-pwe3-cesopsn Y.1413, Y.1453 IA 8.0 MEF 8
SAToP TDM circuit emulation RFC4553 Y.1413, Y.1453 None MEF 8
HDLCoPSN HDLC transport RFC4618 None None None
ATMoPSN ATM service transport RFC4717 Y.1411, Y.1412 None None
FRoPSN Frame Relay service emulation RFC4619 X.84 None None

• RAD is actively involved in most of the
  standardization bodies

17
Saving with Packet-based TransportAll-IP
Transport
GSM BTS
2G BSC
n ? E1/T1 TDM
E1/T1s TDM
Cell-site Gateway
Multi-Service Aggregation
Packet Switched Network
E1/T1s ATM IMA
FE
GbE
ACE-3x00
ACE-3x00
3G RNC
UMTS Node B
ETH

IP-Node B

• All-IP unified transport - Future-ready solution
• FMC/IMS
• Ready for UMTS Rev.6/7 and over
• All TDM and ATM legacy traffic is transported
  over pseudowires
• Clock distribution across the packet network
• Dedicated ACR (Adaptive Clock Recovery)
  mechanisms
• Meet the ITU-T G.823 sync. requirements and
  G.8261 mask
• Evolution to IEEE-1588

18
HSDPA Offload1st Migration Step The Hybrid
Approach
GSM BTS
ATM or SDH/SONET Access Network
2G BSC
E1/T1s TDM
n x E1/T1 STM-1/OC-3
n ? E1/T1 TDM
Cell-site Gateway
Multi-Service Aggregation
E1/T1s ATM IMA
STM-1/OC-3
ACE-3x00
UMTS Node B
GbE
ACE-3x00
3G RNC
ETH
Packet Switched Network
FE
IP-Node B

• First step of the migration to packet switched
  network
• Separation of transport to carry
• GSM/TDMA and UMTS real time voice and video over
  SDH/SONET or ATM
• UMTS HSDPA data traffic mapped to pseudo-wires
  over PSN
• Ensures QoS for voice and real-time traffic

19
The Hybrid Solution over a Wholesale Backhaul
Service
Voice and Real-time over existing infrastructure
GSM BTS
ATM or SDH/SONET Access Network
2G BSC
n x E1/T1 STM-1/OC-3
E1/T1s TDM
n ? E1/T1 TDM
Cell-site Gateway
Multi-Service Aggregation
E1/T1s ATM IMA
STM-1/OC-3
ETH
ACE-3x00
GbE
UMTS Node B
ACE-3x00
3G RNC
DSLModem
ETH
Packet Switched Network
IP-Node B
DSLAM
HSDPA over ADSL2/VDSL2 wholesale

• Reduce Opex by using E1/T1 links only for the
  voice traffic
• Scalable cost-effective broadband transport
  service over DSL (wholesale)
• HSDPA will serve mainly for Internet access and
  fast downloads mobile broadband
• Similar to residential DSL services
• Two options to be considered
• ADSL2 - applicable for HSDPA (HSUPA may require
  more UL BW)
• VDSL2 - applicable for HSxPA

20
Leverage Copper Using DSL
GSM BTS
2G BSC
E1/T1s TDM
n ? E1/T1 TDM
Cell-site DSL Gateway With Bonding
Multi-Service Aggregation
DSLAM
ATM or Packet-switched Transport Network
E1/T1s ATM IMA
STM-1 GbE
xDSL
LA-130
UMTS Node B
ACE-3x00
3G RNC
n x E1/T1 IMA STM-1/OC-3
ETH
Multiple SHDSL linksorADSL2, VDSL2
GbE
IP-Node B

• Leverage existing DSL and DSLAM infrastructure
  for access to cell sites
• Use multiple DSL links (with copper bonding) for
  higher bandwidth, longer reach and dynamic
  adaptivity to available DSL links
• TDM/ATM/Ethernet-over-PSN PW encapsulation over
  DSL

21
The Hybrid Approach in a DSL/PSN Environment
GSM BTS
2G BSC
E1/T1s TDM
n ? E1/T1 TDM
DSLAM
Cell-site DSL Gateway
Multi-Service Aggregation
ATM or Packet-switched Transport Network
E1/T1s ATM IMA
n x SHDSL
GbE
LA-130 NG
ADSL2 /VDSL2
UMTS Node B
ACE-3x00
3G RNC
n x E1/T1 IMA STM-1/OC-3
Ethernet
GbE
IP-Node B

• Mix of DSL technologies in a PSN environment
• Differentiate class of service
• SHDSL for high quality services
• ADSL2 and VDSL2 for lower class of service
• Solving reach and BW challenges
• SHDSL for symmetrical traffic with long reach and
  relative low bandwidth
• ADSL2 for asymmetrical traffic (HSDPA) with mid
  reach and high download bandwidth
• VDSL2 for symmetrical and asymmetrical traffic
  with low reach and high bandwidth

22
A Glance at the Transport for Fixed-Mobile
Convergence
23
Backhaul over the Fixed and Mobile Converged
Transport
Residential
DSLAM/ CMTS
Triple Play
ISP
GSM BTS
3G RNC
ATM
E1/T1s TDM
Multi-Service Aggregation
Cell-site Gateway
ETH
E1/T1s ATM IMA
Packet Switched Network
ETH
ETH
ACE-3100/3200
UMTS Node B
ACE-3x00
2G BSC
Business
ETH
TDM
IP-Node B

• Single packet-switched infrastructure for
  multiple mobile and wireline services
• Aggregation of 2G TDM traffic and 3G ATM traffic
  onto the Triple/Quadruple Play and business
  services infrastructure
• Tunneling of legacy traffic using standard
  pseudowire technologies
• TDM/CESoPSN and ATMoPSN based on IETF standards

24
Leverage the IP DSLAM Infrastructure
Residential
Triple Play
GSM BTS
ISP
E1/T1s TDM
Cell-site DSL Gateway With Bonding
2G BSC 3G RNC
TDM/ATMoPSN Aggregation Device
IP DSLAM
E1/T1s ATM IMA
Packet Switched Network
GbE
TDM
ATM Ethernet
xDSL
LA-130
UMTS Node B
ACE-3400
ETH
Business
IP-Node B

• Further integration for full FMC implementation
• Utilize the same infrastructure as the one used
  for IPTV

25
and one more to go
26
Voice TrunkingBW savings in the Inter-MSC Links
MSC Site
Voice Trunking
MSC
STM-1
E1s
MSC Site
Gmux-2000
Voice Trunking
MSC
STM-1
E1s
Remote MSC Site
Gmux-2000
MSC/BSC
Voice Trunking
E1s

• Significant cost savings
• Up to 161 voice compression ratio
• Excellent voice quality for customers
  satisfaction
• Fast ROI less than 6 months
• Future proof - supports TDM and IP networks for
  seamless migration to NG transport
• Supports all types of signaling
• Optimized HDLC/SS7 signaling transport

E1s
Vmux-2100
27
Migration to IP Backbones
MSC Site
Voice Trunking and Media Gateway
MSC
STM-1
MSC Site
Voice Trunking and Media Gateway
Gmux-2000
MSC
STM-1
Remote MSC Site
Gmux-2000
Voice Trunking and Media Gateway
MSC/BSC
E1s
Vmux-2100

• Leverage the existing backbone segments while
  introducing NG-MSC switches

28
to wrap up
29
Migration to All-IP
Cellular Network
Abis/Ater/A/E Optimization
n ? E1/T1
TDM
IP DSLAM
GSM/GPRS BTS
DSL
2G/3G Backhaul over xDSL
n ? E1/T1
CDMA BTS
ATM DSLAM
TDM/ATM
DSL
ETH
2G/3G Aggregation
n ? E1/T1
ATM
TDM/ATM
PSN Ethernet/IP/MPLS
Conversion to PSN
LTE/WiMAX
ETH
n ? E1/T1
CDMA 2000 Node B
n ? E1/T1
HSDPA Traffic Segregation
UMTS/HSDPA Node B
n ? E1/T1
TDM/ATM
Clock Distribution over PSN
E1/T1
BSC/RNC
STM-1/OC-3 TDM/ATM
30
thank you
for your attention
Issar Krausz Issar_k_at_rad.com
www.rad.com