Shedding Light on Future Wireless Communications · C,B

November 9, 2011

Professor Harald Haas

with acknowledgements to:

Research Fellows and PhD Students I had the pleasure to work with in the past few years

Public funding bodies and industry who have been supporting my research

Shedding Light on Future Wireless Communications

School of EngineeringInstitute of Digital Communications

What is wireless communications?

November 9, 2011

© The University of Edinburgh2

http://stara-sofia.blogspot.com/2009/11/chusseau-flaviens_30.html

Wireless communications –An essential utility

November 9, 2011


> 5 billion(7 billion people)> 1.4 million

Cooper, et al., "Radio Telephone System", US Patent number 3,906,166; Filing date: Oct 17, 1973; Issue date: September 1975; Assignee Motorola

http://www.google.com/patents?id=nO8tAAAAEBAJ&dq=martin+cooper

http://en.wikipedia.org/wiki/Motorola

Mobile Data Usage per Month

November 9, 2011


Year

Exa

byte

spe

r Mon

th

6.3

3.8

2.2

1.20.6

0.24

Radio Frequency Spectrum

November 9, 2011


Network Capacity per Device vs. Forecast Traffic per Device

November 9, 2011


Forecast Traffic per Device

Network Capacity per Device

Year

Net

wor

k ca

paci

ty /

Fore

cast

traf

fic (M

B/M

onth

)

Shor

tfall

97.2

%

Basic questions

1. Are we using the existing radio frequency spectrum most efficiently for wireless communications?

2. Do we know what “most efficiently” is?

November 9, 2011


Do we know what “most efficiently” is?Claude E. Shannon helps, but ...

November 9, 2011


Capacity in bits/second Bandwidth Noise power

Path gain (<< 1)Transmit power

We need to reuse bandwidth...

November 9, 2011


When should be reuse...?

November 9, 2011


Do we better give each link half the bandwidth(each one having half the maximum capacity) andhave no interference, or should both links transmit using the full bandwidth, but allowinterference?

… it depends on a number of random factors

November 9, 2011


„Simple“ frequency reuse concept

November 9, 2011


B2

B3

B1

B2

B3

B1

B2

B3

B1

B2

B3

B1

B2

B3

B1

B2

B3

B1

B3

B1

B2

B3

B1

B2

B3

B1

„Simple“ frequency reuse concept

November 9, 2011


Full frequency reuse

November 9, 2011

BB

BB

B

B

BB

B

BBB

BB

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

B

BB

B

B

B

BB

B

B

B

B

BBB

BBB

BB

Self-organised interference coordination through busy bursts

November 9, 2011


B D

timeData – A Tx

B Tx

Mini Slot

Data – C Tx

Ib

IC,B<Ithr!

Data – A Tx

A C C

System Performance

November 9, 2011


Concept of reuse taken further…Single-input single-output (SISO)

November 9, 2011


Concept of reuse taken further…multiple-input multiple-output (MIMO)

November 9, 2011


Achievable capacity

November 9, 2011


Achievable capacity

November 9, 2011


Achievable capacity

November 9, 2011


Concept of reuse taken further…multiple-input multiple-output (MIMO)

November 9, 2011


Successive Interference Cancellation

November 9, 2011


ABCD

A+B+C+D

A+B+C+D

A+B+C+D

A+B+C+D

Detect the relatively strongest symbol and subtract from all other streams

A+B+C+D

November 9, 2011


ABCD

AB+C+D

B+C+D

B+C+D

B+C+D


November 9, 2011


ABCD

AB+D

CB+D

B+D


November 9, 2011


ABCD

AB

CD


Barriers / Disadvantages

November 9, 2011


Spatial multiplexing MIMO significantly improves spectral efficiency, but:

Suffers from inter-channel interferenceresulting in high computational complex algorithms (e.g., Vertical – Bell Labs Layered Space Time (V-BLAST) algorithm)Requires inter-antenna synchronisation (IAS)Requires multiple RF-chains (→ expensive)Suffers from error propagation

Spatial Modulation: How does it work?

11/9/201128

A(A)

B(A)

C(A)D(A)

A(D)

B(D)

C(D)D(D)

A (Tx1)

B (Tx2)

C (Tx3)

D (Tx4)

Signal Constellation

Spatial Constellation

Re

Im

Im

Im

Re

Re

© The University of Edinburgh

SM Principle

11/9/201129

A(C)

B(C)

C(C)D(C)

A (Tx1)

B (Tx2)

C (Tx3)

D (Tx4)


Re

Im

Im

Re

t1

Signal Constellation Diagram

C B D A B CSpatialsymbol

Signalsymbol

Spatialsymbol

Signalsymbol

Spatialsymbol

Signalsymbol

t1 t2 t3


SM Principle

11/9/201130

A(D)

B(D)

C(D)D(D)

A (Tx1)

B (Tx2)

C (Tx3)

D (Tx4)


Re

Im

Im

Re t2




Signalsymbol

Spatialsymbol

Signalsymbol

Spatialsymbol

Signalsymbol

t1 t2 t3

SM Principle

11/9/201131

A(B)

B(B)

C(B)D(B)

A (Tx1)

B (Tx2)

C (Tx3)

D (Tx4)


Re

Im

Im

Re

t3




Signalsymbol

Spatialsymbol

Signalsymbol

Spatialsymbol

Signalsymbol

t1 t2 t3

Computational Complexity

11/9/201132

Power efficiency

11/9/201133

SM Complexity vs. Performance

11/9/201134

Complexity

Data Rate

Key question

Does MIMO and smart interference management in cellular networks solve the future capacity problem?

November 9, 2011

35 © The University of Edinburgh

My answer is “No”

November 9, 2011


The electromagnetic spectrum

November 9, 2011


Radio-waves

Infra-red Visible Ultra-

violet X-rays Gammarays

Frequencyx 10 000 =

.

Vorführender

Präsentationsnotizen

But, we have 14 billion light bulbs These emit visible light which is part of the electromagnetic spectrum So, the intriguing question is, can we use them for high speed data communications… Look at this in the context of the entire electromagnetic spectrum…

Drivers for visible lightcommunication (VLC)

November 9, 2011


“Looming spectrum crises“: FCCEnergy efficiencySecurityAvailabilityAdvancements in Solid-State LightingCostThe Internet of Things

Forecast 7 trillion wireless devices in 2020 – 1000 per person on earth – source: World Wide Research Forum (WWRF)

November 9, 2011

40

What is different?

System

Radiofrequency

Incoherent light

Information

carried on electric field

carried on optical intensity

Signal

bipolar

unipolarnon-negative

complex valued

real valued


Traditional digital modulationtechniques

Pulsed modulation such as on-off keying

November 9, 2011


Time

Intensity

1 1 1 10 0 00 0

On

Off

Thres.

Finite slope limits achievable data rates

Im

ReOff On

Overcoming data rate limitation ofOOK - Transmitter

November 9, 2011© The University of Edinburgh

42

Orthogonal Frequency

Division Multiplexing

(OFDM)

AG

ETime

Frequency

Re

Im

G H

DCBA

E

LKJI

PONM

One symbol encodes log2(16) = 4 bits in this example

… 1 0 1 1 …

… 0 0 1 1 …

… 0 0 0 1 …

F

1 0 1 1

-fold increase in capacity

DC biased optical (DCO) OFDM

November 9, 2011


124 Mbps real-time andfrom off-the-shelf LED

Signal propagation in aircraftcabin

November 9, 2011


November 9, 2011

46

Data rates in aircraft cabin

wavelength reuse of 2




Vorführender


Finally, the hundreds of lights installed in an aircraft cabin could provide high speed internet access during flights...

Vorführender


Deep see exploration and communication between divers

Vorführender


Intrinsically safe environments – avoid antenna sparks

Vorführender


New cablesless medical equipment

Vorführender


Transportation – cars to communicate with each other to prevent collisions, Traffic lights to communicate with cars, etc.

Vorführender


Street lamps providing ubiquitous wireless access on streets

09/11/201153

Image credits (via Flickr)

“iPhone” by Vincent Huang “Ooievaarsnest in aanbouw” by Vincent Teeuwen“Luces Nubladas” by David Cabrera “Charing Cross” by Jenny Koske“expl1196” by NOAA Photo Library “ILM: Industrial Lighting and Magic” by Photogism Group “easyJet A321 Cabin” by WebDub‘s“Unloved Box, Bentley, Hampshire” by Mike Cattell

Shedding Light on Future Wireless Communications · C,B

Documents