Building Worlds First “5G” Network

Smartphones + Web services + Social networking + Mobile Handoff + Green Nodes

Bill St. ArnaudBill.st.arnaud@gmail.com

Unless otherwise noted all material in this slide deck may be reproduced, modified or distributed without prior permission of the author

Global Average Temperature

2009 second warmest year ever

Jan, Feb Mar 2010 warmest everThis is despite a solar sun spot minimum

Climate Forecasts

MIT

> MIT report predicts median temperature forecast of 5.2°C– 11°C increase in Northern

Canada & Europe– http://globalchange.mit.edu/pub

s/abstract.php?publication_id=990

> Last Ice age average global temperature was 5-6°C cooler than today– Most of Canada & Europe was

under 2-3 km ice– With BAU we are talking about

5-6°C change in temperature in the opposite direction in less than 80 Years

Climate Change is not reversible

• Climate Change is not like acid rain or ozone destruction where environment will quickly return to normal once source of pollution is removed

• GHG emissions will stay in the atmosphere for thousands of years and continue to accumulate

• Planet will continue to warm up even if we drastically reduce emissions All we hope to achieve is

to slow down the rapid rate of climate change

Weaver et al., GRL (2007)

Climate tipping points

• USGS report finds that future climate shifts have been underestimated and warns of debilitating abrupt shift in climate that would be devastating.

• Tipping elements in the Earth's climate - National Academies of Science– “Society may be lulled into a false sense

of security by smooth projections of global change. Our synthesis of present knowledge suggests that a variety of tipping elements could reach their critical point within this century under anthropogenic climate change. “

2009-10 Canada winter 4.2C warmer than average

Urgency of Action• “We’re uncertain about the magnitude of climate change, which is

inevitable, because we’re talking about reaching levels of carbon dioxide in the atmosphere not seen in millions of years.

• You might think that this uncertainty weakens the case for action, but it actually strengthens it.

• This risk of catastrophe, rather than the details of cost-benefit calculations, makes the most powerful case for strong climate policy.

• Current projections of global warming in the absence of action are just too close to the kinds of numbers associated with doomsday scenarios. It would be irresponsible — it’s tempting to say criminally irresponsible — not to step back from what could all too easily turn out to be the edge of a cliff.”

Nobel Laureate Paul Krugman

http://www.nytimes.com/2010/04/11/magazine/11Economy-t.html?pagewanted=1

The Global ICT Carbon Footprint isRoughly the Same as the Aviation Industry Today

www.smart2020.org

But ICT Emissions are Growing at 6% Annually!

ICT represent 8% of global electricity consumption

Projected to grow to as much as 20% of all electrical consumption in the US (http://uclue.com/index.php?xq=724)

Future Broadband- Internet alone is expected to consume 5% of all electricity http://www.ee.unimelb.edu.au/people/rst/talks/files/Tucker_Green_Plenary.pdf

The Global ICT Carbon Footprint by Subsector

www.smart2020.org

The Number of PCs (Desktops and Laptops) Globally is Expected to Increase

from 592 Million in 2002 to More Than Four Billion in 2020

PCs Are Biggest Problem

Data Centers & Clouds Are Low Hanging Fruit

Telecom & Internet fastest growing

IT biggest power draw

Heating,CoolingandVentilation40-50%

Heating,CoolingandVentilation40-50%

Lighting11%Lighting11%

IT Equipment 30-40%

IT Equipment 30-40%

Other6%Other6%

Sources: BOMA 2006, EIA 2006, AIA 2006

Energy Consumption World Wide

Transportation

25%

Transportation

25%

Manufacturing25%

Manufacturing25%

Buildings50%Buildings50%

Energy Consumption Typical Building

Digital vs Traditional appliances

• Greenpeace says that the energy consumption of cloud computing in 2007 was 622.6 billion kWh, which is 1.3 times larger than reported by the Smart 2020 report.

• 50% of today’s Data Centers and major science facilities in the US will have insufficient power and cooling;*

• By 2010, half of all Data Centers will have to relocate or outsource applications to another facility.*

• Data centers will consume 12% of electricity in the US by 2020 (TV Telecom)

Source: Gartner; Meeting the DC power and cooling challenge

Growth Projections Data Centers & Clouds

The Falsehood of Energy Efficiency• Most current approaches to reduce carbon footprint are focused on increased energy efficiency

of equipment and processes

• But growth in ICT deployment of equipment and services is outstripping any gains made in efficiency– Which is likely to accelerate as ICT is used to support abatement in other fields such as

smart homes, smart buildings, smart grids etc

• Also greater efficiency can paradoxically increase energy consumption by reducing overall cost service and therefore stimulates demand– Khazzoom-Brookes postulate (aka Jevons paradox - not to be confused with rebound effect)– In last Energy crisis in 1973 Congress passed first energy efficiency laws (CAFÉ) which

mandate minimum mileage for cars, home insulation and appliances– Net effect was to reduce cost of driving car, heating or cooling home, and electricity

required for appliances– Consumer response was to drive further, buy bigger homes and appliances

• The issue is not the amount of energy that we use, but the type of energy

Zero Carbon strategy essential

• Zero carbon strategy using renewable energy critically important if governments mandate carbon neutrality, or if there is a climate catastrophe

• With a zero carbon strategy growth in demand for ICT services will not effect GHG emissions– Anything times zero is always zero

• Wind and solar power are most likely candidates because of opportunity cost/benefit analysis especially time to deploy– Nuclear has high opportunity cost because of time to deploy– http://climateprogress.org/2008/12/14/stanford-study-part-1-wind-solar-bas

eload-easily-beat-nuclear-and-they-all-best-clean-coal/

• But renewable energy sites are usually located far from cities and electrical distribution systems are not designed to carry load– http://www.americanprogress.org/issues/2008/12/pdf/renewable_transmissi

on.pdf

• Purchasing green power locally is expensive with significant transmission line losses–Demand for green power within cities expected to grow dramatically

• ICT facilities DON’T NEED TO BE LOCATED IN CITIES–-Cooling also a major problem in cities

• But most renewable energy sites are very remote and impractical to connect to electrical grid.– Can be easily reached by an optical network– Provide independence from electrical utility and high costs in wheeling power– Savings in transmission line losses (up to 15%) alone, plus carbon offsets can pay for moving ICT facilities to renewable energy site

• ICT is only industry ideally suited to relocate to renewable energy sites– Also ideal for business continuity in event of climate catastrophe

Get off the Grid!

• Bill 44-2007 was introduced in 2007 and enacted into law in 2008. The law is known as the Greenhouse Gas Reductions Target Act.

• The Act establishes greenhouse gas emission target levels for the Province.– 2020 BC GHG will be 33% less than 2007.– 2050 BC GHG will be 80% less than 2007.

• Bill mandates that by 2010 each public sector organization must be carbon neutral.

• If a public sector organization can not achieve carbon neutrality then they are required to purchase offsets at $24/ton

SOURCE: “Greenhouse Gas Inventory Report 2007”, Ministry of Environment, Victoria, British Columbia, July 2009

Source: Jerry Sheehan UCSD

GHG Regulation in British Columbia

UBC Greenhouse Gas Liability 2010-2012

2010 2011 2012

Carbon Offset $1,602,750 $1,602,750 $1,602,750

Carbon Tax $1,179,940 $1,474,925 $1,769,910

Total $2,782,690 $3,077,675 $3,372,660

SOURCE: UBC Sustainability Office, August 2009

SOURCE: http://climateaction.ubc.ca/category/emission-sources

SOURCE: UBC Climate Action Plan, GHG 2006 Inventory

Carbon Costs for the University of British Columbia

Zero Carbon Data Center

source: Dan GillardBCnet 04/09

BC’s Green Data Centre MUST be in Proximity to a Clean Source of Power

MIT to build zero carbon data center in Holyoke MA

• The data center will be managed and funded by the four main partners in the facility: the Massachusetts Institute of Technology, Cisco Systems, the University of Massachusetts and EMC.

• It will be a high-performance computing environment that will help expand the research and development capabilities of the companies and schools in Holyoke– http://www.greenercomputing.com/news/2009/06/11/ci

sco-emc-team-mit-launch-100m-green-data-center–

Proposed new funding scenario for R&E networks and cost reduction for universities

• Many universities are proposing to go carbon neutral and/or mandated to reduce their energy and carbon footprint

• Purchase of high quality offsets difficult and costly– Better to find energy and carbon savings internally

• Computers, networks and data centers account for 30-50% of energy consumption on campus

• Video conferencing, eLearning, zero carbon data centers, clouds, grids collaborative cyber-infrastructure, etc should reduce energy consumption

• Most universities don’t have processes to allocate energy or CO2 costs to individual departments or researchers– Significant challenge as appropriating heating, cooling, electricity, computing and networks

costs can be very difficult

• There is no incentive for researchers or educators to adopt low energy or CO2 solutions

• Most R&E networks charge a membership fee or base fee based on size of institution, research dollars or number of students

• Instead propose to charge membership or base fee based on institution’s energy consumption– E.g. 1% of total Kwh for the past year

• R&E network agrees to provide a variety of services at no charge including• “X” miles of dedicated wavelengths• “Y” Mbps of Internet bandwidth• “Z” hours of video- conference• “W” time on a commercial compute cloud or central storage• etc

• Institution is encouraged to reduce energy consumption and there is penalty in services if they do so

New R&E network funding scenario“Cap and Reward”

GreenStar –Clouds and Virtualization

Distributed computing architectures, applications, grids, clouds, Web services, virtualization, dematerialization, remote instrumentation and sensors, etc.

Share infrastructure & maximize lower cost power by “following wind & sun” networks.

Develop benchmarking tools to earn CO2 offset dollars for university and ICT department

http://www.greenstarnetwork.com/

GreenStar Network• World’s first zero carbon network• Nodes in Ireland, USA Spain and

Belgium to be added shortly• http://www.greenstarnetwork.com/

Economic benefits of follow the wind/sun architectures

• Cost- and Energy-Aware Load Distribution Across Data Centers– http://www.cs.rutgers.edu/~ricardob/papers/hotpower09.pdf– Green data centers can decrease brown energy consumption by 35% by leveraging the green data

centers at only a 3% cost increase

• Cutting the Electric Bill for Internet-Scale Systems– Companies can shift computing power to a data center in a location where it’s an off-peak time of

the day and energy prices are low– Cassatt a product that dynamically shifts loads to find the cheapest energy prices– 45% maximum savings in energy costs– http://ccr.sigcomm.org/online/files/p123.pdf– http://earth2tech.com/2009/08/19/how-data-centers-can-follow-energy-prices-to-save-millions/

• Computing for the future of the planet– http://www.cl.cam.ac.uk/research/dtg/~ah12/– http://earth2tech.com/2008/07/25/data-centers-will-follow-the-sun-and-chase-the-wind

The NSF-Funded UCSD GreenLight Project: Instrumenting the Energy Cost of Cluster Computing

• Focus on 5 Communities with At-Scale Computing Needs:

– Metagenomics

– Ocean Observing

– Microscopy

– Bioinformatics

– Digital Media

• DC Generation Can Be Intermittent – Depends on Source

• Solar, Wind, Fuel Cell, Hydro– Can Use Sensors to Shut Down or Sleep Computers– Can Use Virtualization to Halt/Shift Jobs

• Develop Middleware that Automates Optimal Choice of Compute/RAM Power Strategies for Desired Greenness

Power Management in Mixed Use Buildings:The UCSD CSE Building is Energy Instrumented

• 500 Occupants, 750 Computers• Detailed Instrumentation to Measure Macro and Micro-Scale Power Use

– 39 Sensor Pods, 156 Radios, 70 Circuits– Subsystems: Air Conditioning & Lighting

• Conclusions:– Peak Load is Twice Base Load– 70% of Base Load is PCs

and Servers– 90% of That Could Be Avoided!

Source: Rajesh Gupta, CSE, Calit2

Evolving Internet & Impact on GHG• Van Jacobson predicted several years ago evolution away from “end-to-

end” Internet to information Internet– http://video.google.com/videoplay?docid=-6972678839686672840#

• Current Internet is based on old telco architecture of connecting users• New Internet is based on connecting users to information• Arbor study has vindicated Van Jacobson prediction

– C. Labovitz, et al "ATLAS Internet Observatory 2009 Annual Report" http://www.nanog.org/meetings/nanog47/presentations/Monday/Labovitz_ObserveReport_N47_Mon.pdf

– Over 50% of Internet traffic is from handful of information suppliers such as Google, Akamai, Content Distribution networks etc

• Will have major impact on wireless networks as we move from “end to end” cell phone network to cell phone Internet

30

Traditional Internet

Backbone ISP1

Backbone ISP 3

Backbone ISP 2

IXIX

RegionalISP

Large ContentProvider with multiple servers

Small ContentProvider

Last mile ISP

Last mile ISP

User’s Computer

User’s Computer

Thousands of miles

Fiber Networks

Information Internet – zero carbon

Backbone ISP1

Backbone ISP 3

Backbone ISP 2

IX IX

RegionalISP

Large ContentProvider

Small ContentProvider

Last mile ISP

Last mile ISP

User’s Computer

User’s Computer CloudCDN

Content hosted in ACI

Nearby

New Internet• Increasingly most connections will be local to nearest IX where user will connect to cloud,

CDN or social network– Connections to another user or computer at the edge of the cloud will be less

common• Network topologies and architectures will be driven by application and content rather

than connecting users “end-to-end”• Cyber-infrastructure and UCLP were forerunners of these development• Future wireless networks are likely to also evolve in this manner

– Don’t need an end to end network for accessing applications and content– User white space and wifi to provide data to cell phones– Next network neutrality battleground will be “last inch” – open devices and “last

tower” - • Much easier to deploy zero carbon Internet

– CDN, clouds and social network infrastructure can be built along GSN model– Users will increasingly use solar powered iPhone, iPad to access information and

applications

33

Grand Challenge – Building robust ICT services using renewable energy only

• 30% of electrical power will come from renewable sources

• How do you provide mission critical ICT services when energy source is unreliable?– Ebbing wind or setting sun

• Back up diesel and batteries are not an option because they are not zero carbon and power outages can last for days or weeks

• Need new network architectures and business models to ensure reliable service delivery by quickly moving compute jobs and data sets around the world to sites that have available power– Will require high bandwidth networks and routing architectures to

quickly move jobs and data sets from site to site

GENI Topology optimized by source destination

Sensor Network

Thin Client Edge Site

Source: Peter Freeman NSF

Mobile Wireless Network

Wind PowerSubstrateRouter

Solar PowerWirelessBase Station

GENI with router nodes at renewable energy sites

Sensor Network

Thin Client Edge Site

Source: Peter Freeman NSF

Mobile Wireless Network

Wind PowerSubstrateRouter

Solar PowerWirelessBase Station

Topology optimized by availability of energy

Power Consumption of IP network

Source: Rod Tucker

Challenge of efficiency

Source: Rod Tucker

Mobile data traffic growth

39

Some info• South Korea's average 'Net speed plunges 24%, iPhone blamed

– http://www.telecompaper.com/news/SendArticle.aspx?u=False– http://arstechnica.com/apple/news/2010/04/south-koreas-average-net-speed-plunges-

24-iphone-blamed.ars?utm_source=microblogging&utm_medium=arstch&utm_term=Main%20Account&utm_campaign=microblogging

• Wifi Mobile Data Offload– http://www.intellinet-tech.com/Media/PagePDF/Mobile%20Data%20Offload%20-

%20Can%20Wi-Fi%20Deliver.pdf

• See all www.stoke.com

40

Building a “5G” wireless network• Over 100,000 cell phone towers to be powered by renewable energy by 2012• Vertical axis turbines and solar• Ericsson (Montreal) world leader in these developments• Existing 3G and 4G networks cannot handle data load• Need to offload data at nearest node or tower• New Wifi standards 802.11u allow for data handoff from 3G networks• WiFi nodes can be powered by renewable sources such as roof top solar panel

over 400Hz power systems or ethernet power• Cell phones also become sensors

Possible Architectures

42

Impact of 5G networks

43

• The PC is out of the loop• The phone is a sensor platform– Hardware add-on innovation– Location based sensing– Touch screen UI

• Processing is done in real time in the cloud– Allowing processing that can’t be done on the device– Big data analysis

• Building new networks on the back of existing ones• Reinventing a major industry

Source: Tim OReilly

WiFi leadership• Finally, Wi-Fi has technology leadership. 4G leverages orthogonal frequency division multiplexing and multiple input multiple output, aka

MIMO. But Wi-Fi adopted OFDM in versions 802.11a (in 1999) and 802.11g (in 2003), allowing Wi-Fi to achieve 54mbps operation. And Wi-Fi adopted MIMO with 802.11n (draft in 2007). Today, 11n devices ship in high volumes, use 2.4gHz or 5gHz spectrum and provide 100-300mbps. New Wi-Fi silicon will deliver as much as 600mbps, and beamforming antennas will increase range and allow dramatically more wireless connections in the same area.

As consumer devices with access to more spectrum than either WiMAX or LTE, Wi-Fi can deliver more megabits per second per dollar. Expect to see both fixed and mobile carriers including free Wi-Fi access in their subscription bundles as Wi-Fi trumps femtocells. Conventional operators are not going away but, over the next decade, its Wi-Fi that will shake up business models and drive disruption

• 802.21 is an IEEE emerging standard. The standard supports algorithms enabling seamless handover between networks of the same type as well as handover between different network types also called Media independent handover (MIH) or vertical handover. The standard provides information to allow handing over to and from cellular, GSM, GPRS, WiFi, Bluetooth, 802.11 and 802.16 networks through different handover mechanisms.

802.21 Will Allow roaming between 802.11 networks and 3G cellularnetworks. A cellular phone user in the midst of a call should be able to enter an 802.11 network hotspot and be seamlessly handed off from a GSM network to the 802.11 network and back again when leaving the hotspot.

IEEE 802.11u is a proposed amendment to the IEEE 802.11-2007 standard to add features that improve interworking with external networks.

IEEE 802.11 currently makes an assumption that a user is pre-authorized to use the network. IEEE 802.11u covers the cases where user is not pre-authorized. A network will be able to allow access based on the user's relationship with an external network (e.g.hotspot roaming agreements), or indicate that online enrollment is possible, or allow access to a strictly limited set of services such as emergency services (client to authority and authority to client.)

44

Sensors using social tools

• Tweet-a-watt• http://www.ladyada.net/make/tweetawatt/

index.html• Neptune and many large sensor networks are

representing sensor as web services• Need new routing and discovery protocols

based on social networking tools• UI on the web• Phone as the sensor platform – processing in

the clouds

45

Open source sensors

46Source: Tim OReilly

Open Source Hardware

47Source: Tim OReilly

Sensors + Cloud

48Source: Tim OReilly

Source: European Commission Joint Research Centre, “The Future Impact of ICTs on Environmental Sustainability”, August 2004

Virtualization and De-materialization

Direct replacement of physical goods – 10% - 20% impact

Case Western pilot with Kindle DX• One pound of printer paper generates 4 pounds of

CO2

• One pound of newspaper produces 3 pounds of CO2

• One pound of textbooks produces 5 pounds of CO2

• Babcock school of Management textbooks for 160 students alone produces 45 Tons CO2– http://www.stewartmarion.com/carbon-footprint/html/

carbon-footprint-stuff.html

Free Wifi on Buses• There’s a school bus service called The Green Bus in Birmingham, UK which

operates double-decker, low-carbon emissions buses that carry over 1400 kids to school every day (saving over 2000 car journeys).

• In addition to encouraging kids to play peer-to-peer games, the access points allow the bus company to monitor where the buses are in the city in real time. Parents as well as staff can follow the progress of any bus via Google maps.

• Business bus service in San Francisco offers office on the move – free wifi, femto cell service etc

• • http://www.muniwireless.com/2009/01/14/school-kids-enjoy-wi-fi-on-green-bus/

32

Carbon Reward Strategy for last mile infrastructure

• Provide free high speed Internet and fiber to the home with resale of electrical and gas power (ESCOs)– http://www.newamerica.net/files/HomesWithTails_wu_slater.pdf– Pilots in Cleveland, Switzerland, Ottawa, etc

• Customer pays a premium on their gas and electric bill

• Customers encouraged to save money through reduced energy consumption and reduced carbon output

• Customer NOT penalized if they reduce energy consumption– May end up paying substantially less then they do now for gas + electricity +

broadband + telephone + cable

• Network operator gets guaranteed revenue based on energy consumption rather than fickle triple play

Final remarks

• The problem we face is NOT energy consumption, but carbon emissions

• Think carbon, not energy

• We must start addressing climate change now – not in 2050 or 2020

• 80% reduction in CO2 emissions will fundamentally change everything we do including universities and networks

• Huge potential for innovation for ICT sector because 30% of energy must come from renewable sources

Cyber-infrastructure in a Carbon Constrained World

http://net.educause.edu/ir/library/pdf/ERM0960.pdf

Let’s Keep The Conversation Going

Blogspot

Bill St. Arnaud

http://green-broadband.blogspot.com

Twitter

http://twitter.com/BillStArnaud

E-mail list

Bill.St.Arnaud@gmail.com

Background Slides

56

Other sectors (40%) (e.g. manufacturing, coal mining, export transport)

Emissions under direct consumer control (35%)

Consumer influenced sectors (25%)(e.g. retail, food and drink, wholesale, agriculture, public sector)

Heati

ng

Private cars

ElectricityOther transport

Consumers control or influence 60 per cent of emissions

http://www.cbi.org.uk/pdf/climatereport2007full.pdf

A More Accurate Term is ‘Global Climatic Disruption’

This Ongoing Disruption Is:• Real Without Doubt• Mainly Caused by Humans• Already Producing Significant Harm• Growing More Rapidly Than Expected”

Earth’s Climate is Rapidly Entering a Novel RealmNot Experienced for Millions of Years

“Global Warming” Implies: • Gradual, • Uniform, • Mainly About Temperature, • and Quite Possibly Benign.

What’s Happening is: • Rapid, • Non-Uniform, • Affecting Everything About Climate, • and is Almost Entirely Harmful.

John Holdren, Director Office of Science and Technology Policy June 25, 2008

Virtualization is key

• Movies and music delivered over Internet

• Virtual applications – Google docs, ESERI

• In many homes electronic devices consume more power than traditional appliances– http://www.iea.org/journalists/

headlines.asp

MIT’s Sixth sense

Zero Carbon Data Centers

Hydro-electric powered data centers

Data IslandiaDigital Data Archive

ASIO solar powered data centers

Wind powered data centersEcotricity in UK builds windmills at data center locations with no capital cost to user

Emerging “Follow the Sun” Technologies

• The ability to migrate entire virtual machines (routers and computers) to alternate data centres exists.

• Over HS networks the latency is tiny and transfer is invisible to the user. • Happens instantly without user knowledge, action or intervention

Nortel’s research labs developed and conceived the “Virtual Machine Turntable in 2006 and through collaboration with R&E networks in the US, Canada, Netherlands, and South Korea proved viability.

• Make – sensors and clouds • UI on the web• Phone as the sensor platform – processing in

the clodu• Amee – the worlds energy meter

62