About Bloobble

Random Presentation

.: Try it :.

NEW! CATEGORIES

I am just starting this, but I wanted to share before getting all categories done. Thanks to send me your feedback and comments.

Voice in an LTE Network

Enjoy this presentation?

You are not logged in. You will be automatically directed to the login screen if you want to download, bookmark or share this presentation.

These are available as part of the subscription package
May be the time to ckeck out a subscription!

These are the first of many nice features we plan
on adding to save your time ::

Get links

Just push the Select button corresponding to your choice, CTRL + C to copy and then use Paste in your target application.

Link to this presentation:

Embed in your web site:

Presentation Transcript:

Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Voice in an LTE Network
Iain Sharp
www.linkedin.com/in/iainsharp
Nortel
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
LTE Introduction
LTE driven by data
• “Long Term Evolution” (LTE)
• Latest iteration of the standard family
defined by 3GPP.
2G GSM => 3G UMTS => 4G LTE
• High-speed, IP only radio interface
• Multiple Megabit/s performance
• Can use “digital dividend” spectrum (800MHz, 2.6GHz)
• “Single Radio” operation with GSM and UMTS
• LTE packet performance is adequate for voice
• In-service dates starting around 2010
• Lead by operators migrating from CDMA
• Initial coverage will be in isolated hot-spots
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
LTE – voice problem
• Voice was planned to use IP and IMS…
• … but adoption of IMS has been slower than
expected creating a technology gap…
• …. so 3GPP can joke….
•… or we can find solutions.
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
IP Multimedia Subsystem
(IMS)
A bit of history
RAN Core
UMTS
Circuit + Packet
Channels
LTE
Only Packet
Channels
GSM
Circuit
Channels
Circuit Core
HLR/HSS
Circuit + Packet
Channels
LTE:
New packet core
No circuit (CS) mode
MSC
CSCF
AS
Packet Core
Mobile is
Single Radio
or
or
Enhanced Packet Core
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
IMS - IP Multimedia Subsystem
• The bright new future (well it was back in 2000)
• Main aims
• Builds on IETF SIP technology
Internet scaling and perceived low cost
• Split the access network from the service network
Common service network across fixed, mobile and internet
• Multimedia in addition to voice
• No need to standardised services only enablers
• What happened ?
• IMS got “standardised” - eventually
Complicated. Many, many options added to SIP
• More “services” became standardised
• Large legacy service load
• Service parity with CS is complicated
• Technology bubble burst
Is this parrot dead, or just resting?
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
IMS – today's reality
• Commercial systems available
• take up has been very slow
• IMS mass deployment dates undefined for most operators
• Issues relate to
• Cost – IMS requires a complete new core network
• Risk – Scaling up to support 10’s of millions to subscribers is still an
unknown. Performance is also seen as a risk
• Integration – Existing services have to be integrated with IMS based
service
• Operator business models – Services run as separate P&L centres
• The good news
• Gears are turning to bring IMS to mass market
• Convergence still seen as positive business model
• Demonstration infrastructure and mobile clients well down the
development lifecycle
IMS still seen as the target architecture
(the parrot is just resting)
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Five LTE Voice Options
1) Data Only
Voice Based on
Legacy MSC
2) CS Fallback
3) VoLGA
Voice Based on
New Infrastructure
4) IMS
5) Over-The-Top
Mobile networks are typically very standards driven
• Decentralized architecture with open multi-vendor interfaces
• International roaming
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
LTE as data only
• Devices with no native voice support - eg PC Dongles
• Likely the initial offering for many carriers
Data only common for initial roll-out
Cons
• Not useful for smartphones
• Single radio limitation
• May encourage over the top
voice
Pros
• Simple
• Device availability
• Debug/test LTE
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
“CS Fallback”
• Use legacy CS on GSM, UMTS or CDMA for voice
• 3GPP Release 8 standard
• Driven by NTT DoCoMo
LTE with voice on legacy
Interim solution. Will it be popular?
Cons
• Standard is ‘barely adequate’
• eg call setup delay
• Needs MSC upgrade
• No LTE during voice call
Pros
• Supports smartphones
• Supports migration from
CDMA as well as
GSM/UMTS
+ Others?+
KoreaJapan
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Tunnelled CS voice over LTE
“CS over PS” or “VoLGA”
Voice over LTE via Generic Access Network
• CS signalling and bearers are tunnelled over IP
• Similar to the GAN standard for WLANs
• Rejected by 3GPP – forced in to dedicated forum
Single operator solution
Cons
• Only T-Mobile strongly
enthusiastic right now
• Rejected by 3GPP
• Scaling and roaming
Pros
• Voice and data on LTE
• Supports Smartphones
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Standards-based VoLTE
• 3GPP defines IMS as the LTE voice solution
• Majority of operators support the IMS path in principle
• How/when will they migrate?
• What about interworking to non-IMS operators?
• Verizon plan to launch in 2010
Many operators add IMS after initial LTE roll out
Data Only
CS Fallback
IMS
LTE Launch
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Service support and migration
• Carriers pushing for “seamless” service migration
• Keep CS infrastructure
• Custom network services like pre-paid
• Billing and provisioning systems
• MSC takes key role in migration
• Direct platform for IMS
• MSC as “Telephony Application Server” (TAS)
• IMS vendor will often follow MSC footprint
Major vendors will define migration options
for their platforms
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Mobility & handover to non-LTE
• Solution needed to go back to GSM/UMTS
• Single Radio Voice Call Continuity (SR-VCC)
• Swap between IMS and CS without preserving
services
• Basic user experience
• IMS Centralized Services
• Service anchor in IMS to improve service consistency
• MSC based
Medium service availability.
Strongest buy-in from operator community
• User Equipment based IMS Centralized Services
Most feature rich
Requires simultaneous CS and PS access
Doesn’t require MSC support
Mobility scenarios are complicated!
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Proprietary Over the Top voice
• Gaps in carrier voice strategy
• OTT technically viable on LTE and UMTS
• Apps-stores make it easy for users to install clients
• Nokia/Skype leading way already
Cons
• Business case unproven
• QoS / DPI blocking?
• Relationship politics
• What happens outside LTE
footprint?
• Regulation
Pros
• Not limited by legacy
continuity
• No “IMS complexity tax”
• Integration with presence and
productivity apps
• Non-traditional voice apps
(Voice Twitter?)
Slow implementation of voice by carriers
may give toe-hold to over the top voice
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Failure in Standards Process?
• Why standards?
• Multi-vendor interoperability
• Economy of scale
• Roaming and consistent user experience
• 3GPP has traditionally aimed for ONE approach
• For VoLTE we have FIVE main options
• Is this a failure of the standards process?
• “No”: Many operations and business scenarios – one
size doesn’t fit all
• “Yes”: LTE dependency on IMS never matched reality
– should have reset assumptions earlier
Market reality moved faster than strategic
thinking in standards
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
What might happen?
Scenario 1 – Market Driven
• Early years of LTE: variety of voice options
• Many operators will dodge the problem by
going data only
• Confusion will lead to “experiments” including
over the top providers
• Eventual convergence driven by roaming –
probably to IMS
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
What might happen?
Scenario 2 – Industry Driven
• Industry renews efforts to simplify options
• Possibly via GSM-Association
• Still data-only LTE operators for operational
reasons
• Reduced uncertainly and more emphasis on
rapid migration to single scenario – probably
IMS
Despite its problems, IMS probably wins long-term
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Conclusions
Predictions
• Variety of options for early LTE networks
• Mostly no voice on LTE for 1-2 years
• Few early runners will debug voice for the community:
Verizon using IMS – drive to migrate from CDMA
T-Mobile – VoLGA
DoCoMo, KDDI – CS Fallback
• “CS fallback” may be further enhanced to improve its
performance
• IMS still positioned to win in the long-term
• Service migration will be an advantage for incumbent MSC
vendors in initial LTE voice core
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Backup
Nortel - Iain Sharp - www.linkedin.com/in/iainsharp
Biographical Notes
• Iain Sharp started working on GSM trial systems in
the late 1980s while he was still a student. In 1990
he joined Nortel and became heavily involved in the
standardisation of the GSM core network. He has
participated in 3GPP since its inception and served
as vice-chair of 3GPP Core Network and Terminals
(CT) plenary from 2003-2005 and 2007-2009.
• Iain’s technical contributions to GSM, UMTS and
LTE include the design of advanced voice services
for GSM and architectural design for GPRS and
LTE.
• Contact:
• www.linkedin.com/in/iainsharp