Transcript Presentation 1

Ofcom’s approach to ensuring Next
Generation Competitive Broadband
Chinyelu Onwurah, Head of Telecoms Technology, Ofcom
TRIS Copenhagen 9-10 February
Agenda
• Next generation access – where are we?
• Wholesale access to enable competition – Ofcom’s approach
• Is effective competition possible without physical unbundling of local
loop?
• Identifying the competitive drivers of Ethernet active line access.
• Ethernet ALA – the road to standardisation
• Appendix 1: ALA standardisation key issues
1
Consumer appetite for broadband continues to grow
Growing usage
Household penetration
Time spent using communication services
100
80
52
60
41
40
20 4
58
27
11
0
Q4
Q4
Q4
Q1
Q1
Q1
2002 2003 2004 2006 2007 2008
Minutes per person per day
Proportion of adults (%)
Growing take-up
250
-3%
224 218
200
-5%
295%
-7%
88%
24
15 14
5 10
173 164
150
100
50
6
0
Television
Radio
2002
Internet
2007
Fixed
Mobile
telephony telephony
Source: Ofcom, The UK Communications Market 2008, http://www.ofcom.org.uk/research/cm/cmr08/
2
… driven by significant infrastructure investment from
competitive providers
• 5,000,000 local loops unbundled
• UK DSL allows 2Mbps to up to 97% of
homes
• UK LLU operators delivering services
up to 24 Mbps
• Cable rolling out 50M services
• BT launched nationwide 8Mbps ADSL
Max service - 6Mbps to ~60% of
homes
3
Next Generation Access is beginning to be deployed
around the world…
Planned deployments (Jun. 08)
100%
100%
90%
90%
Target coverage
Achieved coverage
80%
80%
61%
40%
KPN
Neuf
NTT
FT
TDC
Illiad
Swisscom
Telefonica
9%
Belgacom
4%
Verizon
4%
17%
Telia
Sonera
2%
TI
0.4%
16%
AT&T
26%
20%
0%
38%
DT
40%
Telia
Sonera
60%
Country
Technology
FTTC/H FTTH/B FTTH/B FTTH/B FTTC
Target coverage 0.4%
2%
4%
4%
9%
Target year
n/a
2009
2008
2009
2010
Coverage so far
n/a
0.9%
0.6%
0.5%
n/a
FTTC
16%
2010
8%
FTTH
17%
2009
2%
FTTH
26%
2010
7%
FTTH
38%
2013
n/a
FTTC
40%
2009
21%
FTTC
61%
2008
59%
FFTC
80%
2010
75%
FTTC
90%
2010
n/a
FTTH
90%
2011
84%
FTTC
100%
2010
7%
Note: AT&T is present in
35 out of 50 states.
Verizon has presence in
28 out of 50 states.
Virgin coverage is 50%
of the UK. Illiad target
coverage extends to
70% of Paris
4
… and in the UK
Communications providers
New business models
Sewers and alternative wayleaves
NGA investment of £1.5bn for FTTC
overlay and FTTP in new build to
reach up to 40% by 2012
H2O will be deploy fibre through
sewers in Bournemouth,
Northampton and Dundee. Potential
build cost savings may be 70- 80%
versus traditional methods of
network build
Virgin Media has launched its new
50Mbps cable broadband service to
Geo sewers to deliver bespoke
70% of its customer base by end
fibre networks to carriers and
2008, rising to 95% in 2009
businesses. They have no plans
for a residential network, but their
approach could support NGA
backhaul
New build and municipal
We are seeing an increasing
number of fibre deployments
to new build developments
by new entrants and
municipalities
Examples include:
5
Wholesale access to enable competition
Duct access
• Digging is the major cost in fibre deployment – in France and other countries with food
infrastructure duct access can help enable competition
Dark fibre
• Where duct access is difficult or not available, dark fibre is often considered as the next
best passive access remedy
Sub-loop unbundling (SLU)
• In countries with FTTC deployments, there is focus on SLU.
Wavelength
• DWDM PON is being deployed in Korea but is thought to be to expensive in Europe –
for the moment
Bitstream
• Considered as a remedy where passive access is not available, rather than a preferred
remedy
6
In March we will publish our statement on the regulatory
environment for superfast broadband
Allowing pricing freedom for next
generation wholesale products
Securing timely
and efficient
investment
Creating room for passive access
products in response to demand
Promoting
competition
Supporting industry led development
of active access products
Developing a framework for
transition
http://www.ofcom.org.uk/consult/condocs/nga_future_broadband/
7
Passive and active wholesale access
Proposed regulatory products - FTTC
Copper
Fibre
Street
Cabinet
Customer
Metro Node
Local
Exchange
Passive line access - sub-loop
unbundling
Core Network
Active access - bitstream
Proposed regulatory products - FTTH
Fibre
Fibre
Customer
Splitter
Passive line access duct access, fibre or DWDM
unbundling
Metro Node
Local
Exchange
Core Network
Active access - bitrstream
8
Is effective competition possible with active access?
• In the past effective competition – including innovation and differentiation – has
only really been possible with ownership of the local loop
• We have seen the difference in quality between competition based on
bitstream and competition based on LLU
• For example, competition based on LLU has delivered competitive triple play
offerings whilst IP bitstream offers failed
• Ofcom undertook a qualitative comparison of examples of innovation in
passive and active access…
• … to identify ‘the levers of innovation’ : capabilities which help competitors
differentiate their offerings
9
The levers of innovation…
Lever
Example
Service Levers:
Features
By using ADSL2+ technology LLUOs can offer higher speeds
Infrastructure Costs
By managing contention ratios, LLUOs can control backhaul costs.
Service Levels
By having access to their own dark fibre, a CP can choose to offer
faster service delivery
Business Levers:
Investment choices
LLUOs can choose when to upgrade to ADSL2+
Customisation
LLUOs can choose to differentiate between business customers
and residential on QoS
Billing capability
Because they have access to the underlying call records, LLUOs
can choose how they bill for voice messaging.
Wholesale service
By having a differentiated wholesale offer an infrastructure owning
CP can acquire scale and share risk
Customer Information
By identifying the geographic location of their customers an
infrastructure owning CP can reduce fraud
10
‘Vertical’ Access Point
IPstream
Layer 3
(Network)
Datastream
Layer 2
(Link)
Layer 1
(Physical)
Active
Line Access
Active
Line Access
SLU
Datastream
Active
Line Access
LLU
‘Horizontal’ Access Point
11
‘Active line access’ is a form of Ethernet bitstream which
• Retains as much as possible of the level of innovation supported by passive
access
• Is neutral to higher layers:
– IP-VPN, VLAN, PBB, PBT…
• Is implementation neutral to the underlying media
– Ptp fibre, GPON, copper, bonded copper…
• Is service neutral to the applications:
– Video, HDTV, voice, data…
• Helps overcome technology isolation
– One wholesale access for all technologies
• Benefits from the economics of scale of Ethernet
• And the economics of distribution and management of bitstream
– Customer acquisition does not necessitate truck roll
– Interconnect at different points
12
It is for industry not regulators to define Ethernet ALA –
and there are significant challenges
• Wide range of potential standards
• Needs to be highly scaleable – thousands of CPs with millions of customers
and hundreds of services
• Maximum control of underlying network requires product flexibility
• Ability to support quality of service is key, however that is achieved
• Flexibility within consumer premises equipment required
• Product definition requires significant effort by all industry players
• Pan industry product specification
• Industry players need to understand requirements and trade off product
features to achieve an optimal definition…
13
Key competitive requirements of Ethernet Active Line
Access
Functionality
Justification
Technical requirements
Security
enablement
• Secure delivery of
services
• Authentication of users
• Separate traffic streams
• CPs implement own security
QoS enablement
• Satisfactory delivery of
voice and video
• ALA provider offers QoS information
• ALA user labels traffic
Multicast
enablement
• Bandwidth savings in
backhaul of one to many
services (IPTV)
• Optional provider or user solution
• Common interface
• Static and dynamic support
Flexible customer
• To allow CPs to innovate • Common Ethernet interface (initial)
premises equipment
in CPE functionality
• Wires- / Fibre-only interface (future)
Flexible
interconnection
• There is no universally
economical
interconnection point
• Local, regional, national interconnect
• Common interface
• Freedom to move
14
Why is Ofcom promoting Ethernet ALA?
•
•
•
•
•
•
•
Like most regulators, we prefer infrastructure access
We are also promoting sub-loop unbundling and looking at duct access
But these unlikely to be viable everywhere – like LLU
So some form of bitstream access is essential
And the better it is, the more innovation will follow
And the more consumers will benefit
Other regulators are also looking at active line access type products
What should this mean for communications providers?
• The availability of a standardised wholesale access product sooner rather than
later
• Giving easy access to fibre communities wherever they may be
• Supporting wholesale and retail products
• And allowing for differentiation in pricing, quality of service, security,
applications etc
15
Process
Regulators Regulatory
Objectives
Promote effective
competition
Competitive
Characteristics
1
2
3
4
5
Technical
Requirements
1
2
3
4
5
Standardisation
NICC
CP 1
IEEE
CP 2
ITU
ETSI
Specification
Broadband
forum
MEF
CP 3
Product
Specification
Industry
Product 1
Product 2
Product 3
16
Summary
• In many countries, consumers are reaping the benefits of a competitive market in
current generation broadband
• As next generation access networks roll-out, we want enable competitive providers to
compete
• Passive access products have many advantages but are unlikely to work everywhere
• Ethernet Active Line Access has the potential to be a generic wholesale access product
which enables providers to compete and differentiate themselves – attracting next
generation consumers
• And its standardisation would help overcome the potential for technology isolation
• We will actively work with industry and other regulators to bring about the best outcome
– Ethernet Active Line Access: Technical Requirements
http://www.ofcom.org.uk/telecoms/discussnga/eala/eal/
– Ethernet in the Access Study http://www.ofcom.org.uk/telecoms/discussnga/eala/eas/
– Delivering super-fast broadband in the UK
http://www.ofcom.org.uk/consult/condocs/nga_future_broadband/
• Updated Technical Requirements to be published early March.
17
Safeguarding investment in NGANs
• Not ‘investment vs competition’ but ‘investment and competition’.
Ensuring a competitive environment stimulates investment, it does not deter it.
• At EU level it’s important to avoid premature attempts to harmonise the detail
of regulation on the one hand, and attempts to hard-wire regulatory holidays or
other forms of anti-competitive discrimination on the other into the Framework,
therefore:
– It is important that the Commission avoids too much of a focus on passives and too
much prescription of detail in its NGAN recommendation and set out the key
principles instead;
– We remain wary of some of the previously proposed amendments on risk-sharing;
– But we believe that regulated prices should be adjusted to reflect uncertain future
demand for retail products and hence increased project investment risk;
– Consequently, we seek text to clarify what would be legitimate spreading of
investment risk as opposed to disguised erection of new barriers to entry (for instance,
where an incumbent proposes that altnets can only gain access if they commit to
upfront capital expenditure)
18
Functional separation and the impact on
investment:
•
The Undertakings have
provided BT with
regulatory certainty and
BT continues with a £10bn
(€11.3bn) investment in
NGN core network
19
Questions?
[email protected]
20
ALA Standardisation Key Issues
Appendix 1
Key Issues - Security
Area of Requirement
Key Issues
• The transport streams shall
identify both the ALA-user and the
customer
• Which scheme should be used to mark traffic?
• How should the customer and ALA-user
information be encoded?
• Ethernet ALA should be
transparent to security protocols
• Which existing standards could be used?
• Are there any security protocols that may cause
issues with transparency?
• Ethernet ALA should allow the
necessary intercept and tracking
functionality
• Is there any way that Ethernet ALA would preclude
this?
• Is it necessary to engage with the security
community?
22
Key Issues - QoS
Area of Requirement
Key Issues
• Traffic capacity should include a
committed information rate as a
minimum
• Is it necessary to define the size of these
committed rate ‘pipes’?
• Are there any network challenges in providing this
functionality?
• ALA-users must have sufficient
information to deliver their
services
• What (real-time) information is required, and how
does this vary by technology?
• Is this a technical requirement or product
definition?
• QoS should be managed using
existing standards
• Which standards should be used?
• How should available QoS labels be assigned to
applications and ALA-users?
• How far should we go in defining the QoS labels?
• Good network design practice
should ensure acceptable delay
and jitter performance, as well as
traffic policing
• What new challenges does Ethernet ALA bring to
designing a fit-for-purpose network?
• Is this a technical requirement or general best
practise?
23
Key Issues - Multicast
Area of Requirement
Key Issues
• Provider Multicast should be
standardised and optional as part
of Ethernet ALA, but ALA-users
should be able to implement their
own solution
• What challenges would be presented by an ALAuser implementing their own solution?
• Ethernet ALA should support
multiple ALA-user multicast
• How far has multiple-operator multicast been
standardised?
• What are the challenges of implementing multipleoperator multicast?
• Ethernet ALA should include static • Are both static and dynamic required?
and dynamic multicast
• How should this be implemented and managed?
24
Key Issues – Flexible CPE
Area of Requirement
Key Issues
• The interface at the customer
premises should be common
across all infrastructures
• Is this technically possible?
• There must be the option to move
back to wires- or fibre-only
• What will signify that the market is ready to move
from a common interface?
• What are the challenges for each technology?
• The CPE interface should be the
• Should the Ethernet scheme be the same at the
same Ethernet scheme as the rest
customer interface as at the interconnect
of the network
interface?
• The CPE should be easy to install
and report connectivity
• Are there any significant changes needed to
existing practices?
• Standards should be identified for • Is CPE management important?
the complete management of CPE • How can existing standards be used?
25
Key Issues – Flexible Interconnection
Area of Requirement
Key Issues
• ALA-users should be able to
interconnect to aggregated traffic
at any point
• Should interconnection be allowed and/or required
anywhere in the network (e.g. at the cabinet)
• How can backhaul products ensure the
requirement is met?
• Interconnection should be via a
standard sized interface, with a
common Ethernet scheme
• What are the appropriate standard interconnect
sizes?
• Should the interconnect scheme be common
across all infrastructures?
• ALA-users should have freedom
to change interconnection points
• Are there any additional challenges associated
with changing interconnection over time?
• Standards should be defined for
equipment compatibility
• What existing standards can be used and how
flexible should compatibility standards be?
26