Backhaul challenges for Wireless Broadband Networks

Module 3G/4G Backhaul Challanges

Objectives
After this module the participant shall be able to understand: The user requirement for WCDMA/LTE Backhaul Challenges due to Network Architecture Evolution Required specifications for Wireless band Broadband Backhaul Backhaul Solutions and Strategies

Broadband is driving the growth

Wireless broadband constitutes most of broadband traffic and will supersede wireline traffic by 2013. Multimedia services plays key role in wireless broadband traffic

Backhaul
Movies, music, news, more music, text, web, more content ..

Wireless Broadband
1. Mobile traffic is seeing exponential growth in bandwidth Applications like high-speed internet, mobile TV broadcast, Video on demand become popular Most traffic over HSPA+/LTE (or EV-DO or WiMax)

2. Revenue not scaling proportionally!

3. Need cost-effective, scalable backhaul 4. Leverage packet-transport to provide: Statistical multiplexing gain for data traffic Native multicast support for video 5. Two alternatives: Enhance SDH/SONET with packet transport capability Overlay a native packet-transport network 6. Both need to match SDH performance Operational simplicity and Meeting stringent QoS requirements

Revenue

The combination of rising traffic requirements coupled with declining revenues is a key motivation for operators migrating RANs to a converged, packedbased architecture
Packet transport has been globally deployed in these types of networks and has been an important element in creating an environment for the delivery of new data services

Cost Analysis for Mobile Backhaul

Traditional Backhaul is Expensive Mobile operators are generating revenues from a range of new next generation data services that are designed to generate revenues in addition to those from legacy voice services.

However, these new 3G-based services require a substantial increase in bandwidth, which will in turn lead to greater mobile backhaul costs. It is estimated that backhaul can account for as much as 30% of a mobile operators operating costs (Opex) (source: Yankee Group, 2005)
If mobile operators were to expand the backhaul network to meet these new bandwidth requirements in the traditional manner, the move to 3G could represent a significant increase in required bandwidth and associated opex.

Backhaul as on today
Non packet backhaul for packet RAN
wave
BTS

Copper
Node B

BSC

TDM Backbone Network

Fibre

RNC

eNB

AGW

Legacy

Legacy Transport

Legacy RAN

Legacy RAN

Challenges for next generation Backhaul

Evolving backhaul requirements per cell site: 2 - 16 E1s in 2006; 2 - 8 E1s and 10 - 30 Mbps Ethernet in 2010, and goes beyond 100MBPS per site with new technologies

Mobile operators pay incremental charges for 2x to 10x bandwidth Legacy backhaul networks are prohibitively expensive, mobile operators transport costs are expected to skyrocket
Fierce competition is squeezing margins while new data and video applications such as Apples iPhone & iPAD, Samsungss Galaxy, and HTC One are driving demand for more bandwidth & Growing dependence on mobile connectivity Exponential traffic growth in parallel to flat or low growth of average revenue per user (ARPU) The bandwidth increase will primarily be on Best Effort data user services and driven by flat fee business models. More technologies need to be supported: 2G/GSM, 2G/CDMA, 3G/UMTS, 3G/EVDO, HSDPA, WiMAX and LTE

Evolution towards Ethernet and IP-based backhaul solutions

More and more capacity

Backhaul capacity growth for highdemand urban cell site is129% CAGR

25 Mbit/s today to up to 300 Mbit/s in 3 years

Key criteria for evaluating a backhaul solution

The new backhaul infrastructure must therefore meet three main criteria; it must be Flexible to support both legacy and IP services Scalable to support emerging future technologies Cost-effective to compensate for rising levels of traffic TDM natively supported Ethernet natively supported

Multiple synchronization streams supported

Guaranteed performance in terms of latency and jitter Extreme scalability decouples from N x cost structure Ability to complement and support or replace pseudowires Cost optimized options for cell sites over a range of distances Major TDM and packet synchronization standards supported Operations synergies within converged metro transport networks Applicability to both backhaul access and aggregation network

Business Challenges
Mobile Operators
Improve Customer experience: Fewer dropped calls Higher data rates Smoother transition between 2G, 3G, 4G (LTE), and WiFi Enhance performance: Add bandwidth Improve QoS mechanisms

Transport Providers
Generate incremental revenue: Wholesale backhaul services Business services

Expand service offering: TDM leased line services (i.e. pseudowire for 2G/3G backhaul over Ethernet n/w. Carrier Ethernet services (i.e. circuit bonding for 3G/4G backhaul services over TDM/SDH networks) Extend footprint: Revenue generating backhaul services help justify network expansion to more cell towers Two of every three towers already have more than one mobile operator Reach more customers and mob. operators

Expand coverage: Many markets underserved, especially for 3G/4G services

Business Challenges
Mobile Operators Keep backhaul costs in line with revenue generated: Transport Providers Capture share of rapidly growing backhaul market by offering mobile operators:

Implement more efficient network Less expensive backhaul options topologies than todays typical E1 connections Optimize network by moving best Scalable bandwidth effort data services to cheaper unpredictable needs facilities for

Take advantage of lowest cost, highest bandwidth wholesale transport networks

Technical Challenges
Mobile Operators Transport Providers Define performance criteria and ensure Build a transport network that delivers TDM their self-built or alternative transport and Ethernet services that exceed the mobile provider n/w meets the requirements operators performance criteria Research wholesale transport n/w available in each region and take advantage in each region of the lowest cost, highest bandwidth, and most reliable ones. Optimize usage of expensive access links: Idle flag suppression and dynamic bandwidth allocation (DBA) Statistical multiplexing of services Intelligent oversubscription Abis optimization per VC switching of voice and data Design the network to support multiple mobile operators per tower that share the same network facilities. Sell bandwidth pipes on a per E1 or per Mbps Ethernet basis: Deliver quality service (i.e. committed information rates (CIR) with low latency, jitter, and packet loss No optimisation Transparently support any mix of cellular protocols with TDM, ATM, and Ethernet interfaces

Technical Challenges
Mobile Operators Accurately regenerate timing Transport Providers Accurately regenerate timing per mobile operator to ensure they each get timing form their own clock source Support standards based Ethernet and pseudo wire OAM for connection fault management and performance monitoring Find hardened units for use in compact cabinets Migrate to Ethernet, IP and MPLS for more bandwidth at lower costs Provide services over available access infrastructure (fibre, microwave, bonded E1 circuits, SDH, DSL, etc.)

Monitor performance to ensure SLA guarantees are met

Backhaul network must evolve

Transform

Transition

Maintain

Evolved Backhaul Network

Primary Reference Clock , 1588v2 Master NxE1/DS1 TDM Ethernet/IP

~ ~
Hybrid TDM + Ethernet/IP

BSC

~
MME

IP/Ethernet
~
~

MGW

IP/Eth/MPLS
RNC

MPLS R4 LTE CORE

aTLA

2G/3G/ LTE/WiMAX

Sync Ethernet

~
BSC

S-GW

Cellsite Gateway

~
~

Aggregation Gateway
Distributed Sync Delivery Point 1588v2 Slave Metro/Core Gateway

ACCESS/LOW-RAN

AGGREGATION/Hi-RAN

METRO

CORE

Packet Transport Possibilities

Ethernet-LAN services over SDH Provides statistical multiplexing and multicast advantages over an SDH network Supports QoS mechanisms to effectively share Ethernet bandwidth among multiple services Reuses SDH OAM capabilities and adds additional OAM capabilites at the Ethernet layer Leverages SDH for synchronization Higher per-bit cost due to overhead of SDH Pure-Packet (Ethernet over Fiber) Backhaul Potential for least cost per bit but many aspects under standardization OAM and some protection switching standards available Synchronization supported

Key challenges for Packet Transport

Scalability aspects Scalable to large numbers of users/nodes/flows: Thousands of network nodes needed for wireless backhaul in each metro Scalable in bandwidth: To meet traffic growth requirements Resilience 50 ms restoration similar to present SDH/SONET network

Manageability Ability to troubleshoot and ease of provisioning

Synchronization Provide time and frequency synchronization Traffic Engineering Prioritization of traffic to meet stringent QoS requirements.

Scalability

Alternatives based on Ethernet or MPLS extension Traffic Engineering extension to PBB (PBB-TE) as per IEEE 802.1Q T-MPLS extensions as per ITU-T 8110.1
Provide tags to differentiate the customer domain from service provider domain. Ability to classify thousands of flows (based on I-SID in PBT, MPLS tag in T-MPLS) and provide a simplified and scalable provisioning model.

Resilience
Protection in linear, ring and mesh topology Linear protection is being dealt by ITU-T in G.8031 (PBT), G.8131 (TMPLS) Ring protection is being dealt by G.8032 (PBT), G.8132 (T-MPLS) Carrier class equipment reliability Effort in ITU-T to come out with G.8021(PBT), G.8121(T-MPLS)

Protection
Protection (point-to-point) 1:1 Path Protection provided on per unidirectional PBB-TE Trunk Uses one of the following concept

Source Based: Switching the traffic from work to protect in case of failure at the source
Destination Based: Send the traffic along both work and protect paths and select the appropriate traffic at the destination.

Multipoint-to-multipoint protection Based on heart beat messages to detect failure Drop and continue at each node for multicast Protection uses the fact that there are only two paths between any two nodes Need to reroute in the other direction upon failure

Synchronization
Synchronization type needed in mobile: Synchronization: Clock-offset from Reference clock be minimized Synchronization: Clock frequency in sync with Reference clock Others: Symbol-synchronization, Frame-synchronization, Packet synchronization, Network-synchronization

Synchronization is required to ensure seamless connection hand-over of handsets moving between coverage areas of different base stations TDD (time division duplex) requires both Frequency Synchronization and time synchronization to reference clock. FDD (frequency division duplex) requires only synchronization to reference clock.

Synchronization requirements
Synchronization requirements in various wireless standards:
Standard Frequency Accuracy Time Accuracy

CDMA 2000 GSM UMTS FDD UMTS TDD LTE FDD TD-LTE

3s NA NA 2.5s NA 3s

Packet Synchronization
Mechanisms used today include Rubidium Atomic clock GPS PDH/SDH transport IEEE 1588v2 PTP and IEEE 802.1as PTP uses hardware time stamping at PHY, and software runs at IP-Layer PTP is adapted to Layer-2 Ethernet bridged network as IEEE 802.1as; Time stamping is at Layer-1; software runs at Datalink-layer ITU-T G.8261 and G.8262 (draft) Specifies performance bounds for all the above techniques

Synchronization methodology

Traffic Engineering
Needs to maintain the same QoS for voice/synchronization frames as provided by dedicated TDM-based SDH service Ability to prioritize traffic based on service/Traffic type/SLAs. Ensuring Frame Delay, Frame Delay Variation, Frame Loss and Bandwidth requirements are met

Quality of Service (QoS) To ensure QoS we need: Traffic Classification: Identify traffic requiring preferential service based on port, C-VLAN, P-bit, ToS, DSCP, etc. Policing (rate-limiting): Discard traffic that exceeds CIR/EIR rates Mapping to queues: Congestion avoidance mechanisms such as WRED ensure that high priority traffic is not dropped Scheduling: Strict priority queues are handled first and WFQ thereafter Shaping: Conform egress traffic to a specific line rate

Comparision of Technologies
Technology Optimized For Service Activation Bandwidth Scalability Network OAM

PDH SDH/SONET

Voice Voice

Slow Slow

Very limited Limited

Low Medium

Complex Simplified, Carrier-class

Simplified, Carrier-class

NGSDH/ SONET

Voice & Data

Rapid

Fine

High

Carrier Ethernet

Voice & Data

Rapid

Fine

High

Simplified, Carrier-class

Backhaul options for Next Generation

MSPP Metro Ethernet Ethernet over SDH Ethernet over DWDM Ethernet over Fiber Ethernet over MPLS

Carrier Ethernet

Ethernet Solutions Mobile operators are looking towards Ethernet solutions for scaling their mobile backhauls as : Ethernet can provide high bandwidth scalability with low cost. As compared with the legacy TDM network, the Ethernet is more flexible in bandwidth options. simplifies the network. significantly reduces the operational expenses and is more reliable.

L2/L3 backhaul Challanges Wholesale backhaul providers typically prefer L2:

Simpler to provision Scalable BW pipes for unpredictable needs Strong Ethernet OAM mechanisms offer SLA Sub 50ms failover with 802.3ad and G.8032 Pseudowire helps support 2G/3G services, in addition to LTE Powerful diagnostic tools

Pure-Play wireless operators typically prefer L2:

Simple / automatic provisioning Ethernet circuit validation, PM, fault detection and analysis Traffic engineering oversubscribe link bandwidth

Integrated carriers may prefer L3 (skill sets)

Mesh, alternate routing, but less developed OAM

Why Carrier Ethernet ?

Connection Oriented Ethernet

Mobile Network Operators (MNOs) are evolving their wireless infrastructure to the Long Term Evolution (LTE) standard that offers low latency and high bandwidth to deliver rich content to smart mobile devices, challenges and complexities in LTE design with the recommendation to use Carrier Ethernet- based solutions for backhaul from cell sites to reduce Total Cost of Ownership (TCO)

Carrier Ethernet
Carrier Ethernet Economically meets exploding bandwidth requirements currently constrained by the prohibitive costs of legacy networks Leverages rapid move to Carrier Ethernet for wire-line traffic enabling a single integrated wire-line and mobile backhaul network Much easier for service providers to manage and maintain Most mobile traffic is broadband/IP centric Carrier Ethernet is optimized for packet data traffic
Overcomes TDM (T1/E1) services scalability This alone makes Carrier Ethernet the compelling choice Time/urgency Carrier Ethernet removes the barrier to timely progress

Pseudo wire aggregation

Primary reference source or PRC is provided to Pseudo wire aggregation equipment at the BSC/MSC and then distributed over packet networks to the Pseudo wire gateways at the BTS and Node B locations.

Evolving from current backhauling

Key comparison - Transport

Worldwide Mobile Backhaul New Connections by Technology

Worldwide Mobile Backhaul New Connections (by Technology)

900
Ethernet m icrow ave Ethernet copper and fiber SONET/SDH Other ATM over PDH

750 600

PDH/SDH m icrow ave PDH

Connections (K)

450 300 150 0 -150 -300 CY05 CY06 CY07 CY08 CY09 CY10 CY11

Conclusions and Recommendations

Manage your explosive packet traffic growth With industry leading transport capacity Extend your investment With seamless and low risk network migration, with continuity of existing TDM services Simplify your Network With a converged backhaul/transport solution Reduce capital and operational network costs

With future proof all IP Backhaul