On access bandwidths: On access bandwidths Per O. Andersson
Outline: Outline Moore’s law and residential services
Symmetry
Operational expenses
Bandwidth
Television
Flexibility
New services
Future requirements
…in any Package…: …in any Package…
...by any Service...: ...by any Service... SIZE SECURITY REALTIME
…from the Home Operator…: …from the Home Operator… …to operate the home network
…to act as 24/7/365 all-line helpdesk
…to be legal part for the relevant SLAs …to specify, negotiate and deliver home network services to the end-users (family)
…to be responsible for multiple carrier agreements (broadband, mobile, fixed…)
…to take end-to-end responsibility for reliable service delivery to the end-user
…to track and follow the development of the relevant technologies (broadband, HDTV, convergence, VoIP, mobile, 3G, HSDPA, WiMax, fixed wireless, WiFi…
Will it continue…? - Does Moore hold up?: Will it continue…? - Does Moore hold up? 2000 2005 2010 x1 x10 x100 Internet expansion Image services? x1000 x10 000 10 Gb
155 Mb
56 kb 100 Gb
1+ Gb
560 kb 1 Tb
10 Gb
6 Mb 10 Tb
100 Gb
60 Mb 100 Tb
1 Tb
600 Mb
Service Breakthroughs During 130+ Years…: Service Breakthroughs During 130+ Years… 1900 2000 Time Sound Image Symbol Personal service Shared service Breakthrough â
Personal Control: Personal Control Increased personal control Is there another data type…? Is there another control…? VR ? Message Control
Capex/Opex - Two approaches: Capex/Opex - Two approaches Type A for high flexibility & optimised OpEx
Large Customer base with churn (incumbent)
Type B for lowest CapEx but higher OpEx
Build revenue streams with lower initial risk (competitive) Capex/Opex Type A Type B Future changes?
(A)symmetric Network Value: (A)symmetric Network Value Symmetry very valuable:
Any P2P network
Most human communication Symmetry irrelevant:
Centralized content distribution “Full disk” “Full disk” Symmetric Network Asymmetric Network Peering point Note that many peer2peer applications now go beyond “filesharing” to become a professional way of distributing content and programs, telephony, grid computing etc…
ADSL2 Reach & Performance Overview: ADSL2 Reach & Performance Overview 1 Km 2 Km 3 Km 4 Km 5 Km 6 Km Length, Km 8 24 ADSL ADSL2+ VDSL 52 Data Rate, Mbps ADSL2 12 NTT measurement
“centre of gravity” Shannon’s information theorem…
(S/N-limited, here: crosstalk) “Pretty deep fibre”…
Measured Distance and Speed for ADSL: Distance(km) Loss(dB) Downstream Bit rate Measured Distance and Speed for ADSL From ECOC’04 Y. Maeda, NTT 3 km @ 2 Mbit/s 1 km @ 10 Mbit/s Real upgrade path Performance spread
“But you code all this to x Mb/s with MPEG-y…”: “But you code all this to x Mb/s with MPEG-y…” 10 x 12 bits resolution
Total 120 pixels
21 bits of information
Information matched to display resolutio
E.g.5 Mb/s MPEG-2 on standard TV
...yes, but most of that means lower quality... : ...yes, but most of that means lower quality... 20 x 24 bits resolution
Total 480 pixels
21 bits of information
Information less than display resolution - for example due to compression
E.g.5 Mb/s MPEG-2 on HDTV
”High Definition” simply has more information!: ”High Definition” simply has more information! 20 x 24 bits resolution
Total 480 pixels
42 bits of information
Information matched to display resolution
E.g. 20 Mb/s MPEG-2 on HDTV
So what’s your Bandwidth...?: So what’s your Bandwidth...? Display Algorithm Technology Scene Quality (PQR) VHS DV HD1 HD2 D-Cinema RAW Lossless MP-1 MP-2 MP-4 Sports Mixed Face 3 5 10 Leading Medium Simple
The bandwidth cube: The bandwidth cube Better Codec
MPEG-2 to H.264
(x 0.5) Less cost & time
No quality review
Less coding (realtime!)
Processor, memory cost
(x 2 …or 3…) Image size
SDTV to HiDef
(x 4) Starting point
Hi performance MPEG-2 SDTV
(2 Mbit/s) 2 8 16 8 2 1 4 4
Revenue per Megabit & MegaByte…: Revenue per Megabit & MegaByte… ** “1$/day” = 30 Euro/mo * 0,2 Euro/min = ca 300 Euro/24h
The price of bandwidth: The price of bandwidth Expensive bandwidth is not an axiom… Price/bit Cost/bit Then Now Future
The Next x10 – DTV-based…?: The Next x10 – DTV-based…? DTV is not the same old thing as TV…
DTV requires > 95% penetration
DTV requires “telecom quality” – Five 9’s - 99.999% or ~5 min/yr
DTV requires 2-4 simultaneous channels/family, incl. HDTV
Video can get away with 2 Mb/s - DTV needs 20-50 Mb/s…
Which “Triple Play”?
A) 1ch VoD + VoIP + BE Internet (BE, 5 Mbs)
B) 3-4ch (HD)TV + IP Telephony + Future Internet (QoS, 20-30 Mbs)
Three delivery methods:
Air
Cu
Fibre Bandwidth, QoS
Competitive Landscape - Delivery: Competitive Landscape - Delivery 0,1 Mb/s 1 Mb/s 10 Mb/s 100 Mb/s Modem 512k DSL 5M DSL 20M xDSL++ DTB Cable Fibre Cu “Truckrolls” Modem STB, TV, PC… Servers, Transport Modem Delivery?
Mistakes in History - Lee de Forest 1926: Mistakes in History - Lee de Forest 1926 ”Although Television may be theoretically and technically possible, I consider it to be a financial and commercial impossibility” (Radio pioneer, inventor of the triode)
Help! I’m not used to this...!: Help! I’m not used to this...! Video services are “culturally uncharted territory”
Public acceptance may treat all services alike
Result has massive impact on telecom network
ECOC 2004 demo: ECOC 2004 demo The small room The large room Testing the “Window to the world” Small room “Telepresence” to KTH via Ericsson MINI-LINK A “2K” projector – a.k.a. 1080 HDTV A Community Hub model at ECOC 2004
The “Community Hub”: The “Community Hub” ”Broadband for all”
2nd wave/emerging markets ”Window to the world”
”Full-presence” video
– eye, size, distance
eye-commerce
e-government
e-healthcare ”Centre Stage”
e-Cinema & event
Personalized events
Politics & religion, sports
Connection to mobile & 3G
2D and 3D Holography: 2D and 3D Holography 2D Holography
2D images by diffraction from dynamic hologram
Essentially standard video-rates
Requires Hi-Def SLM
Requires RGB-lasers/LD(?)
Prototypes today
Gives pocket HiDef-projectors 3D Holography
Artificial real-time calculation and generation of a true 3D hologram
Requires x100 in processing power
Requires ca 1um pitch SLM (today 10um)
Requires RGB-lasers LD(?)
10 years with Moore?
Gives dynamic 3D imaging
Summary - bandwidths: Summary - bandwidths Home networking - will be complex
Hi-Def & triple-play emerging
Multi-service, multi-vendor, multi-technology…
Your bandwidth may vary…
Symmetry will be valuable
Low Opex will be necessary
100 Mb/s Ethernet natural choice
Challenges for 21st Century Society: Challenges for 21st Century Society Business
Original, self-sustained, creation of value and employment is necessary to create prosperity and security.
Culture
Unhindered access to local and world culture is necessary to be considered an attractive region to work and live in.
Society
High-quality societal services are necessary for a region to be able to care for its inhabitants.
With
Availability
To all people in all parts of a region - or of the world.
Sustainability
Economic, Social and Environmental
”Moore” is also a business and social factor...: Socio-economics Technology ”Moore” is also a business and social factor... Number of transistors per chip Lithography & vacuum equipment Service complexity &
Communication need Usefulness Commercial acceptance Business growth New investment
Global Market Need - Short term to 2010: Global Market Need - Short term to 2010 Short term growth markets – ”stock market analysis…”
Telecommunication
Convergence
Emerging markets (“second wave”)
Digitisation of TV and film (analogue switch-off)
Emerging bio-photonics
…?
Positioning for long-term?
Meanwhile – in the Middle East…: Peak oil, est. 2008 Demand
ca 1,5% /yr Supply Meanwhile – in the Middle East… Demand driven Supply driven “Regular Oil & Gas”
Total “Oil in Place” ca 2,4Tb (arrels)
Daily use ca 80 Mb (arrels)
Yearly use ca 30 Gb (arrels)
Peak Oil meets Fibre Optics…: Peak Oil meets Fibre Optics… Assume PO ca 2010
Mental change
Transport prices increasing
Political and market uncertainties
Government and market adjustments & steering
Financing redirected towards energy (and transport) use
Terabit Society ca 2010
Bubble-era systems 32- 40ch @ 10 Gb/s filled up and to be replaced
Multi-Mb @ home, 10G @ metro, Terabit in transport
Analogue TV and photo being replaced – all is going digital
Emerging home networking
Energy consumption focus (flat screens, fibre…)
“Second Opto-bubble”? – tempered by competition for funding…
Can “Moore” be maintained in the new paradigm?
Global Market Need - Long term to 2050: Global Market Need - Long term to 2050 Long term growth markets – “important problems”
Natural resource utilisation; oil, gas, water, forest, agriculture…
Sustainable manufacturing and recycling
Energy utilisation and distribution
Transport & Communication - main replacement for liquid fuels…
Medicine and care – demographic changes…
Success is required to obtain/maintain prosperity and security
Very large R&D resources will be directed towards energy “realignment”
Telecommunications will be truly key (physical transport replacement)
Global Market Need - Long term to 2050: Global Market Need - Long term to 2050 Long term growth markets – “important problems”
Natural resource utilisation; oil, gas, water, forest, agriculture…
Sustainable manufacturing and recycling
Energy utilisation and distribution
Transport & Communication - main replacement for liquid fuels…
Medicine and care – demographic changes…
Success is required to obtain/maintain prosperity and security
Very large R&D resources will be directed towards energy “realignment”
Telecommunications will be truly key (physical transport replacement)
Sweden has “pole position” in several core areas (at present…)
Technical and organizational
Public and private sector
Industry profile?
Role of Telecommunication after PO: Role of Telecommunication after PO Energy will be prime focus and driver of “everything”
Telecom should still be good priority - but amount of funding?
Development might have to be done by “old” technologies – IPR mining?
Developed areas (Sweden, Europe)
Telecommuting as substitution
Strong political demands from rural areas
Emerging markets (China, India)
Enhanced focus on Telecommunication “from scratch”