2010/9/13

1

Dynamic Cell Planning and Self-Organization

for GREEN Cellular Networks

UC4G Shanghai Workshop 2010, 12 Sep., 2010UC4G Shanghai Workshop 2010, 12 Sep., 2010

Zhisheng NIUNetwork Integration for Ubiquitous Linkage and Broadband (NiuLab)

Dept. of Electronic Engineering, Tsinghua University

Tsinghua National Laboratory of Information Science and Technology

12010/9/13

Content

Why and What’s GREEN?

Globally Resource-optimized and Energy-Efficient Networks

Dynamic Cell Planning for GREEN

Tradeoff between energy saving and coverage

Self-Organization for GREEN

Cell Zooming and Dynamic BS Sleeping Control

Conclusion

2010/9/13 2

2010/9/13

2

Why Green, Why Now?

Energy consumption of cellular networks is growing rapidly with increasing data rates and No. of users/BSs China is the largest ICT market in the world and still in fast growing

PC of China Mobile: 6 4b’06 8 1b’07 9 4b’0811 2b’09 PC of China Mobile: 6.4b’06 8.1b’07 9.4b’0811.2b’09

China Mobile has 460K+ BSs (‘09) and will go to 600K (‘10)

There is a push for “green” from industry & government from World-Wide-Web to World-Wide-Wireless

for World-Wide-Watch & World-Wide-Wisdom

but definitely should not World-Wide-Wait

and World Wide Waste!and World-Wide-Waste!

Where does PC majorly come from?

2500

3000

h)

500

1000

1500

2000

2500

onsu

mpt

ion

(108

kWh

Network Devices

Display Devices

x 13

x 5.2

4Source: Ministry of Economy, Technology and Industry “Promotion of Green IT”, Feb. 2008

0

500

2006 2025Pow

er C

o

Server, Storage Devices

PC

x 2.5

Power Consumption by IT Devices in Japan

2010/9/13

3

Where does PC majorly come from?BSs consume ~80% of energy in cellular networks useAir conditioners consume ~50% of energy in base stations

dSleep?

*1Mt CO2 = 2TWh

correspond to 25 millionhousehold average yearly

consumption

Sleep?

Energy ~2TWh ~60TWh ~3.5TWh ~10TWh

CO2 ~1Mt ~30Mt <2Mt ~5Mt

3 billion subscribers3 billion subscribers

4 million Radio Stations4 million Radio Stations

20,000 Radio Controllers20,000 Radio Controllers

Other elementsOther elements

Green as a fatal factor

Equipment Type TEEER Formula Min. TEEER Allowable

Transport -log (Ptotal / Throughput) 7.54

Verizon’s TEEER (Telecom Equipment Energy Efficiency Rating) since 2009

optical and Video 7.54

P2P Microwave 5.75

Switch/Router -log (Ptotal / Forwarding Capacity) 7.67

Media Gateway -log (Ptotal / Throughput) 6.54

Access (Access Lines / Ptotal ) +1 2.50

Power (POut Total / PIn Total ) X 10 9.20

2010/9/13 6

Power Amplifier (Wireless)

(Total RF Output Power / Total Input Power) X 10

1.05

BS?

2010/9/13

4

What’s Green, How to GREEN?

Green Environmental friendly

Low power

GREEN Globally Resource-optimized

BS/AP cooperation in homogeneous networks

Network cooperation in heterogeneous networks

Energy Efficient Networksfrom 3A to 3R: Right time Right place and Right service from 3A to 3R: Right time, Right place, and Right service

dynamic adaptation to traffic variation and QoS requirement

sleep when traffic is low

7

: Globally Resource-optimized and Energy-Efficient Networks

Scarce resources are under-utilized

resources are distributed in heterogeneous overlaid NWs WiFi, WiMAX, 2G, 3G, LTE, DTV, Operator A/B/C, …

And, they are quite under-utilized

Network Example ofUtilization

Carrier Network 33%

Internet Backbone 15%

LAN 1%

82010/9/13

LAN 1%

2010/9/13

5

CHORUS for GREEN

XX

X

X X 802.11b802.11b 802.11a

WiMax

X

DVB/DMB-TGSM/CDMA

2010/9/13 99

: Collaborative and Harmonized Open Radio Ubiquitous System

Dynamic Base Station Cooperation

2009-11-12 10

2010/9/13

6

Network Convergence and Collaboration

InternetMobile

InternetIPTV

Broadband

MobileTV

Broadcasting

Internet

Mobile TV

11: Network integration for ubiquitous linkage and broadband

Multi-mode Terminals

Future NWs should be energy-efficiency-oriented, while providing satisfactory performance (not over-provisioning)

• right-time, right-place, and right-service traffic variation can provide opportunities for energy saving!

GREEN by Traffic/QoS Adaptation

• traffic variation can provide opportunities for energy saving!

Pow

er Reduced Consumption

Usual Power Consumption

122009-11-12

0:00 12:00 24:00 t

Traffic

2010/9/13

7

In fact, traffic variation is very high

13

E. Oh, B. Krishnamachari, X. Liu, and Z. Niu, “Towards Dynamic Energy-Efficient Operation of Cellular Network Infrastructure”, IEEE Com. Mag., 2010 (to appear)

In fact, traffic variation is very high

Temporal traffic variation (public NW)Real Cellular Traffic Data (Korea Telecom)

Shift to half power mode?

2010/9/13 14

E. Oh, B. Krishnamachari, X. Liu, and Z. Niu, “Towards Dynamic Energy-Efficient Operation of Cellular Network Infrastructure”, submitted to IEEE Com. Mag., 2010.

2010/9/13

8

Temporal traffic variation (Enterprise NW)Typical traffic in an enterprise networkTypical traffic in an enterprise network

In fact, traffic variation is very high

Traffic decreasessharply at nighttime

Traffic almost disappearson the weekend

152010/9/13

Sun Mon Tue Wed Thu Fri Sat Sun

From Dr. Takehisa Hayashi (AlaXalA, Japan)

Switching off ?

Non-business hour of a typical company is more than half a year: (120X24h + 245X8H = 4840h)/8760h!

In fact, traffic variation is very high

Spatial traffic variation

2010/9/13 16

D. Willkomm, S. Machiraju and J. Bolot, Adam Wolisz, “Primary User Behavior in Cellular Networks and Implications for Dynamic Spectrum Access”, IEEE Com. Mag., March 2009

2010/9/13

9

Content

Why and What’s GREEN?

Globally Resource-optimized and Energy-Efficient Networks

Dynamic Cell Planning for GREEN

Tradeoff between energy saving and coverage

Self-Organization for GREEN

Cell Zooming and Dynamic BS Sleeping Control

Conclusion

2010/9/13 17

Concept of Cell Planning

Candidate sites: determined by buildings and geo info

Test point: abstracted user distribution

TPs

CSs

182010/9/13

2010/9/13

10

3 candidate sites, 12 userDeployment result without inter-cell Cooperation

A Deployment example

192010/9/13

A Deployment example

Deployment result with inter-cell Cooperation

202010/9/13

2010/9/13

11

Selected Published Papers

[1] Zhisheng Niu, Sheng Zhou, Yao Hua, Qian Zhang, and Dongxu Cao, “Energy-aware Network Planning for Wireless Cellular System with Inter-cell Cooperation”, Submitted to IEEE Trans. Wireless Commun., 2010

[2] S. Zhou, J. Gong, Z. Niu, Y. Jia, and P. Yang, "A decentralized framework for dynamic downlink base station cooperation," Proc. IEEE Globecom'09, Dec, 2009.

[3] S. Zhou, J. Gong, Z. Yang, Z. Niu, P. Yang, “Green Mobile Access Network with Dynamic Base Station Energy Saving”, ACM MobiCom2009 (Poster), Beijing, China, Sep. 2009

[4] F Zh Z Ni “D l A l i f Sl B d P S i M h i

212010/9/13

[4] F. Zhu, Z. Niu, “Delay Analysis for Sleep-Based Power Saving Mechanisms with Downlink and Uplink Traffic“, IEEE Commun. Letters, 2009

[10] E. Oh, B. Krishnamachari, X. Liu, and Z. Niu, "Towards Dynamic Energy-Efficient Operation of Cellular Network Infrastructure", IEEE Commun. Mag., 2010 (accepted)

Content

Why and What’s GREEN?

Globally Resource-optimized and Energy-Efficient Networks

Dynamic Cell Planning for GREEN

Tradeoff between energy saving and coverage

Self-Organization for GREEN

Cell Zooming and Dynamic BS Sleeping Control

Conclusion

2010/9/13 22

2010/9/13

12

Existing Cell Planning

For given traffic (mostly voice only) demand and resources (spectrum, power, etc), plan the coverage and resource allocation Coverage-oriented and performance optimized

Issues Traffic demand is getting highly uncertain due to smaller

and smaller cells and higher mobility

Traffic category is getting more and more diverse (Data and id t ffi i i t d i t )video traffic is going to dominate)

New technologies (MIMO, DAS, Relay, CoMP, etc) are emerging

Strong demand for energy consumption reduction (energy-efficiency oriented cell planning)

23

Cell Zooming by Dynamic Association Control

Cell zooming for load balancing

2010/9/13 24

Central cell zooms in as traffic load increases

Central cell zooms out as traffic load decreases

2010/9/13

13

Cell Zooming by Dynamic Association Control

Cell zooming for energy saving

2010/9/13 25

Central cell sleeps, and othercells zoom out

Central cell sleeps, and othercell cooperate to transmit

Cell Zooming for Energy Saving

A snapshot of BS sleeping pattern

2500

Low Load

y-ax

is (m

)

1000

1500

2000

2500

Medium

‐ Assuming BSs can extend coverage from 200m to 400m

‐ Users arrive according to Poisson Process and service

2010/9/13 26

x-axis (m)

500 1000 1500 2000 2500 3000

500

High Load

Active cells Sleeping cells

Poisson Process and service time is exponentiallydistributed

2010/9/13

14

Performance of Cell Zooming

70

75

nsum

ptio

n

10

12

nsity

(s-1

)

65

70

sum

ptio

n

0.008

0.01

roba

bilit

y

Average energy consumptionAverage blocking probability

0 5 10 15 20 25 30 35 40 45 5050

55

60

65

Syst

em E

nerg

y C

on

Time (h)

0 5 10 15 20 25 30 35 40 45 50

4

6

8

Syst

em T

raff

ic I

nten

System energy consumptionSystem traffic intensity

0 0.5 1 1.5 2 2.5 3 3.5 450

55

60

Ave

rage

Ene

rgy

Con

Ts(h)

0 0.5 1 1.5 2 2.5 3 3.5 4

0

0.002

0.004

0.006

Ave

rage

Blo

ckin

g Pr

2010/9/13 27

(1) Our proposed dynamic algorithm tracks the variation of the average traffic intensity well(2) As the minimum mode holding time increases, energy consumption changes a little, and the average blocking prob. increases, but still below the target

Tradeoff between Energy and Coverage

‐ Centralized and Decentralized algorithms‐ Outage Probability vs. Number of active BSs in low traffic conditions

28

Zhisheng Niu, Yiqun Wu, Jie Gong and Zexi Yang: “Cell Zooming for Cost-Efficient Green Cellular Networks”, IEEE Com. Mag., 2010 (to appear)

2010/9/13

15

Selected Published Papers

[1] J. Gong, S. Zhou, Z. Niu, and P. Yang, "Traffic-aware base station sleeping in dense cellular networks," IEEE Int'l Workshop on Quality of Service (IWQoS'10), Poster, Jun. 2010.

[2] D. Cao, S. Zhou, C. Zhang, Z. Niu, "Energy Saving Performance Comparison of Coordinated Multi-Point Transmission and Wireless Relaying", IEEE Globecom'10, Miami, USA, Dec.2010

[3] Z. Niu, Y. Wu, J. Gong, and Z. Yang, "Cell zooming for Green cellular networks," IEEE Communication Magazine, 2010 (accepted).

292010/9/13

Content

Why and What’s GREEN?

Globally Resource-optimized and Energy-Efficient Networks

Cell Zooming for GREEN

Energy saving by dynamic association control

Self-Organized Networks for GREEN

Energy saving by collaborative BS sleeping

Conclusion

2010/9/13 30

2010/9/13

16

Summary: Existing Wireless Technologies

It’s all about capacity! but do we have enough traffic?

It’s all about voice! It s all about voice! but data & video will dominate

It’s all about single cell! Micro cellular leads to high-density and overlapped cells

Multiple radio and multi-carriers coexist

It’s all about noise/interference It s all about noise/interference but power is also limited

31

Summary: Green Wireless Technologies

It’s all about traffic! Care more about carried traffic, not only channel capacity

Treat data/video differently (from 3A to 3R)

Adapt to traffic variation (Cell Zooming)

It’s all about cooperation/collaboration! Cross-node design, not only cross-layer design (CHORUS)

Inter-network collaboration (NiuLab)

It’s all about power/energy!p gy Power/energy is limited and prohibited (GREEN)

322010/9/13