Green Wireless Telecommunicationsleti.congres-scientifique.com/annualreview2009/AnnualReview/AR09...BTS ~90% = Operator OPEX ... Source: NSN. 13 ... activities for the integration

1Laurent Hérault, 20 June 2009 11th Leti Annual Review 2009 Copyright 2009 CEA-LETI

Green Wireless Telecommunications

Dr. Laurent Hérault

CEA-LETI

Business Development Manager


Outline

�Energy efficiency of ICT & mobile communications

�Sustainable mobile networks solutions�Holistic approach�Green metrics�Green radios�Green networks�Green communications for an EE society

�Conclusions

5 billion people

connected


“We simply must do everything we can in our power to slow down global warming before it is

too late. The science is clear. The global warming debate is over.”

Arnold Schwarzenegger, Governor of California

“The IPCC has unequivocally affirmed the warming of our climate system, and linked it directly to human activity (greenhouse gas

emissions).”Ban Ki-moon, UN Secretary-General


The Issue (1/2)

� Currently, 3 % of the world-wide energy is consumed by the ICT infrastructure � which causes about 2 % of the world-

wide CO2 emissions� comparable to the world-wide CO2

emissions by airplanes or ¼ of the world-wide CO2 emissions by cars

� The transmitted data volume increases approximately by a factor of 10 every 5 years


The Issue (2/2)

� ICT: 10% of electrical energy in industrialized nations� 900 Bill.. kWh / year = Central and

South Americas

� Power consumption of ICT is currently rising at 16-20% / year � Doubling every 4-5 years

� Wireless communications are used extensively to save energy in other industrial sectors.


ICT use: worldwide today

� Total = 156 GW ~ 8% of the global electricity consu mption� No dominating front

Sou

rce:

Ghe

nt U

nive

rsity

/IB

BT


ICT: Future estimations

� 1/7th of electricity goes to ICT use phase in 2020

� The GESI Smart 2020 report* predicts growth in ICTs p ower requirement of 70% over the period 2007-2020

* Global e-Sustainability Initiative / www.smart2 020.org

Sou

rce:

Ghe

nt U

nive

rsity

/IB

BT


EE & Mobile networks

� So far, mobile networks design rules have ignored EE� Cellular networks have been

optimized in terms of spectral efficiency, capacity or throughput, not really in terms of Energy Efficiency!

� EE consideration mainly for mobiles� EE considered only for high load

scenarios

� Long-term vision: sustainable energy powered micro/femto base stations

� Emerging IoT / RWI raises green challenges


Power consumption of cellular networks

Users

1kW 10kW0,1W

3 million x 1.5kW4,5GW

3 billion0,2-0,4GW

10.000 x 10kW0,1GW

~20Mt CO2/a

~1-2Mt CO2/a

~0,5Mt CO2/a

Mobile CS Core

Mobile PS Core

Transport

IP Core

BTS

~90% = Operator OPEX

Sou

rce:

NS

N


Power consumption of cellular networks & Internet of Things

Users

1kW0,1W

3 million x 1.5kW4,5GW

3 billion0,2-0,4GW

~20Mt CO2/a

~1-2Mt CO2/a

BTS

~90% = Operator OPEX

0,02W

IoT, M2M

300-10000 billion6 - 200 GW

LETI0,005W / ZigBee

Battery-free sensors


EE & Mobile networks

� Mobile operators are already today among the top en ergy consumers� Telecom Italia (fixed & mobile) is the 2nd largest energy consumer in

Italy

� Energy consumption of Mobile Networks is growing much faster than ICT on the whole� Rapid traffic growth and build-up of broadband coverage� Mobile replacing fixed in many areas� Enabling ICT services for energy saving in other sectors

(teleconferencing,...) further increases mobile networks growth

Annual growth rate of internet traffic is 85%.


Power consumption of a traditional cellular site

60W1500W

Backup system

Dieselgenerator Battery

Air conditioningCooling

Rectifier

Site support system

Mains(230V)

Only ~4% of the consumed power transmitted

50% energy lost(Feeder cable)

1000W 120W

35% energy lost(Site support)

Sou

rce:

NS

N


Power consumption of a modern cellular site

60W500W

Backupsystem

Rectifier

Site supportsystem

Mains(230V)

~12% of the consumed power transmitted

120W

430W

14% energy lost(Site support)

50% energy lost(Feeder cable)

Sou

rce:

NS

N


Outline

�Energy efficiency of ICT & mobile communications

�Sustainable mobile networks solutions�Holistic approach�Green metrics�Green radios�Green networks�Green communications for an EE society

�Conclusions

5 billion people

connected


LETI in the European Community

� « Green Communications »eMobility Working Group & white paper

� 1st International Workshop on Green Wireless Communications 2008, 2009

� ICC’Green Coms WS’09

� IST IP SENSEI -> WSNs (FP7 Call 1)

� EUREKA CATRENE PANAMA� 1 IP + 1 STREP in FP7 Call 4


EE mobile networks require an holistic approach

Reference scenarios

• Deployment strategies

• Cooperation schemes

• PHY layer parameters

• Transceiver architectures& components

• Traffic patterns

• User densities• Higher layer strategies

EE Metrics?

… from semiconductor technology to radios and networ ks


Mobile networks

� Large savings potential not only for quiet hours. Network load is not evenly distributed� Typically 10% of the sites carry 50% of all

traffic.

� 50% of sites are lightly loaded, carrying only 5% of the traffic.

� Energy savings should be considered primarily at the network level� NTT DOCOMO have calculated that, for their

52 million subscribers in 2006, the energy consumption of their network per mobile user per day was 120 times greater thanthe daily energy consumption of a typical user’s mobile phone .


EE Metrics

� Metrics should be defined to measure energy consumption (in e.g. Wh or Joule) on component, node and system/network levels, in relation to delivered QoS and system spectral efficiency

BS sub-

component

EE m

etric

BS EE

metric

Site EE

metric

Network

EE

metric

Global metric

PHY layerCircuitry MAC / higher layers …

Tx Rx SignallingSignalprocessing

Access scheme Cooperation

System level

Component level

Energy consumption Spectral efficiency QoS requirements


Towards Green Radios

Energy Efficiency EnablingRadio Interface Techniques

Integrated optimisation considering

component, radio andinterface to network-level

• Base station power adaptation

• Sleep mode and associated signalling• Transmission mode adaptation

• Dynamic load adaptation• Cross layer optimisation

EE Technologies and Components

Power scalable

transceivers

Power control on

component, front -end and system

level

Adaptable matching

networks

EE Application of Innovative Radio Transmission

Techniques

• MIMO

• Adaptive antennas• Coordinated multi -points

• Advanced retransmission


Innovations in RF front-end architectures

� Design flexible architectures with new components�MEMS to have better perf. (filtering)

with lower consumption

� Tune performances at run time to the required flexibility & reconfigurability

� Co-design of PAs and LNAs� Better performance (BW a gain) for a

given current� Lower current for given performance

� Avoid losses in the chain� Best matching between amplifiers and

the antenna� Use of PA in their non linear domain +

predistorsion techniques & digital compensation (Tx or Rx)


Innovations in digital BB architectures

� Multi-User MIMO

� Manage multiple standards in the terminal� Complex & dense digital

partitioned NoC (network on chip / multi-cores)

� Use of DFVS in GALS systems, i.e. Tuning of local power supply units & clock generators

� Run full frequency only when required

� Master leakages� Leakage currents represent up to

25% of the power consumption with submicron technos

• Spatial dimensions assigned to several users.

• Separation of users by TDMA, FDMA and SDMA.


Towards Green Networks (1/4)

� Deployment scenarios:� optimum cell sizes

� mix of cell sizes� hierarchical deployments

� multi-RAT deployments� relays & repeaters

Deployment

overlay macro cell

smallcells

relays

EE topology



� Management algorithms:� coverage adjustment� capacity management�multi-RAT coordination� base station sleep mode� protocol design

Management

multi-RAT

EE adaptive cov./cap.

Zzz



� RRM algorithms:� cooperative scheduling� interference coordination� joint power allocation and

resource allocation�EE vs. spectral efficiency

RRM

EE joint RRM



� Disruptive approaches:�multi-hop transmission� ad-hoc networks� terminal-terminal-

transmission� cooperative multipoint arch.�EE adaptive backhauling� cognitive/opportunistic

radios & networks

New Architecture

Adaptivebackhaul

multi-hop

Future EE architectures


Some disruptive power saving strategies

short range cooperative clusters in homogeneous networks

exploiting heterogeneous RATs

� Use of cognitive radios and networks


System level simulations

� Green metrics� Evaluation of green algorithms & protocols


Integration and testing

Integration and testingIntegration and testingDefinition of the validation procedures

Define the validation platform

Define the rules and the activities for the integration in the test bed

Block 1

divider

Device 2

a

c

Signal Generator

out in out in

b

1111

3333

2222

divider

divider

Ch 1 Ch 2

Block 1

divider

Device 2

a

c

Signal Generator

out in out in

b

1111

3333

2222

divider

divider

Ch 1 Ch 2

Test Plant


ICT can help society

� ICT has the potential to reduce society’s CO2 emissions (carbon footprint) by 5% - 20% in the short to medium term

� ICT is a cornerstone ina low-carbon economy

1800 1850 1900 1950 2000 2050

EU goal(-80%)

Currentdirection

ICT red.5% - 20%


Just how can ICT help society?

Source: GESI / Smart2020 report, 2008

Dematerialisation


Conclusion and outlook

� Significant energy savings can be expected� by defining and standardizing EE

metrics, � combining energy aware flexible

radios and networks, and� using wireless technologies in many

sectors.

� LETI is working on promising road blocks


Thank you …

Documents

Green Wireless Telecommunicationsleti.congres-scientifique.com/annualreview2009/AnnualReview/AR09...BTS ~90% = Operator OPEX ... Source: NSN. 13 ... activities for the integration