Multiuser MIMO Indoor Visible Light Communications: University of Virginia

Multiuser MIMO Indoor Visible Light
Communications
A Dissertation
Presented to
the Faculty of the School of Engineering and Applied Science
University of Virginia
In Partial Fulfillment
of the Requirements for the Degree
Doctor of Philosophy in Electrical Engineering
by
Jie Lian
December 2017
APPROVAL SHEET
This dissertation is submitted in partial fulfillment of the
requirements for the degree of
PhD in Electrical Engineering
Jie Lian, Author
This dissertation has been read and approved by the examining committee:
Prof. Maite Brandt-Pearce, PhD Advisor
Prof. Stephen G. Wilson , Committee Chair
Prof. Daniel S. Weller
Prof. Malathi Veeraraghavan
Prof. Stephen D. Patek
Accepted for the School of Engineering and Applied Science:
Dean, School of Engineering and Applied Science
December 2017
Abstract
Visible light communications (VLC) is an energy efficient and cost-effective solu-

tion for indoor wireless multiple access and a candidate technique to provide high-
speed data transmissions. VLC systems are built as dual systems (illumination and
data transmission) and have potentially higher privacy than RF communication sys-
tems due to the natural character of light. Light emitting diodes (LEDs) that work
as transmitters in VLC systems have many advantages, such as ease of modulation,
high power efficiency and long life expectancy [1]. Since the radio frequency (RF)
spectrum is so congested, and the data transmission rate of RF communications can-
not satisfy the huge demand for a high data transmission, VLC has emerged as a
possible new technology for the next generation communications.
In this dissertation, we introduce a multi-LED transmitter model and a multi-
detector receiver model. Based on these models and the Lambertian law, we derive
the impulse response of the indoor channel and the optical power distribution in
space.
To support multiple access using VLC, we propose a centralized and four decen-
tralized power allocation algorithms. In the centralized power allocation algorithm,
all the LED lamps in the room are coordinated and controlled by a central controller;
each LED lamp supports all the users within the indoor area. For standard indoor
office illumination level (400 lx), about 40 users can be supported with bit error rates
less than 10−3 using on-off keying and 70 MHz bandwidth of receivers at 5 × 10−7
W/Hz noise spectral density. The decentralized power allocation algorithms proposed
have similar bit error rate performance and less computational burden compared to
the centralized algorithm. Compared with the centralized algorithm, the running time
of decentralized algorithms is less than 10% of the centralized algorithm. In addition,
3
some practical considerations, such as shadowing effects, illumination requirements,

dimming control and transmitted power quantization are taken into account. From
numerical results, the proposed adaptive power allocation algorithms can adjust the
transmitted power to reduce shadowing effects and provide an excellent communica-
tion performance.
High-speed data transmission is required by modern communication systems. For
VLC systems, the transmitted bit rate is also an essential consideration. An adap-
tive M-ary pulse amplitude modulation (M-PAM) scheme is proposed to provide
high bandwidth efficiency for different channel qualities. Given the bandwidth and
the power limit characteristics of LEDs, a waveform design algorithm with adaptive
M-PAM modulation can be applied for high-speed transmissions. When the 3 dB
bandwidth of the LEDs is 20 MHz, and the peak transmitted power is 3 Watts for 3
users, the system can achieve about 200 Mbps bit rate per user using the proposed
waveform design algorithm. Channel uncertainty is considered, which can be mod-
eled as a Gaussian random process. Together with the minimum mean squared error
filters at the receivers, the optimized waveforms can reduce intersymbol and multiple
access interferences together. We then propose an off-line waveform design algorithm
to diminish the computational time. For the off-line algorithm, a waveform lookup
table can be established in advance, and the proper waveforms can be selected from
the table based on the real channel gains in real time. The performance of the off-line
algorithm can be estimated by using the channel uncertainty model. Compared with
DC-biased optical orthogonal frequency division multiplexing, M-PAM with optimally
designed waveforms can provide an 80% higher data rate for single user.
Given the power distribution, we analyze the potential vulnerability of the system
from eavesdropping outside the room. By setting up a signal to noise ratio threshold,
we define a vulnerable area outside of the room through a window. We compute
4
the receiver aperture needed to capture the signal and what portion of the space
is most vulnerable to eavesdropping. Based on the analysis, we propose a solution
to improve the security by optimizing the modulation efficiency of each LED in the
indoor lamp. The simulation results show that the proposed solution can improve
the security considerably while maintaining the indoor communication performance.
5
Acknowledgements
First and foremost, I would like to thank my advisor Professor Maı̈té Brandt-
Pearce. Her determination and commitment have always been an inspiration to me.
Her deep knowledge of communication theory and her impeccable attitude towards
research have greatly helped me to form a rigorous, dedicated, and creative research
style for my future career life in the communication engineering field. Her strict
requirement towards the quality of publication and presentation also helps me to
hold myself to a higher standard.
I would like to thank my dissertation advisory committee members, Prof. Stephen
G. Wilson, Prof. Daniel S. Weller, Prof. Malathi Veeraraghavan and Prof. Stephen
D. Patek, for their invaluable time and helpful suggestions.
I am grateful to my parents for their patience, help and love. “A journey of one
thousand miles begins with one step,” my parents always encourage me to take the
first step. During the five years, my parents gave me the power to finish the PhD’s
program. Then, I would thank my lovely wife, who always tells me how good I am
when I feel stuck.
I would like to thank the China Scholarship Council for the funding support.
It is a blessing to be surrounded by many colleagues and friends at the University
of Virginia. Finally, I am grateful to all my friends in Charlottesville and for making
my life at U.Va. enjoyable.
Contents
1 Introduction 1
1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.2 Differences Between VLC and RF Communications . . . . . . . . . . 3
1.3 Modulation Schemes for VLC . . . . . . . . . . . . . . . . . . . . . . 5
1.3.1 On-Off Keying . . . . . . . . . . . . . . . . . . . . . . . . . . 5
1.3.2 M -ary Pulse Amplitude Modulation . . . . . . . . . . . . . . 5
1.3.3 Pulse Position Modulation . . . . . . . . . . . . . . . . . . . . 6
1.3.4 Orthogonal Frequency Division Multiplexing . . . . . . . . . . 6
1.4 Multiple Access for VLC . . . . . . . . . . . . . . . . . . . . . . . . . 8
1.4.1 Time Division Multiple Access . . . . . . . . . . . . . . . . . . 8
1.4.2 Optical Code Division Multiple Access . . . . . . . . . . . . . 9
1.4.3 Orthogonal Frequency Division Multiple Access . . . . . . . . 9
1.5 Literature Review and Dissertation Contributions . . . . . . . . . . . 10
1.6 Dissertation Organization . . . . . . . . . . . . . . . . . . . . . . . . 12
2 System Description 14
2.1 Transmitter and Receiver Models . . . . . . . . . . . . . . . . . . . . 14
2.2 Channel Model . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.3 Additive Noise . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 18
i
ii
2.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3 Power Allocation Algorithms for Multiuser VLC Systems 20

3.1 Transmitted and Received Signals . . . . . . . . . . . . . . . . . . . . 22
3.2 Centralized Power Allocation Algorithm . . . . . . . . . . . . . . . . 25
3.2.1 Single Detector Receiver . . . . . . . . . . . . . . . . . . . . . 27
3.2.2 Multiple Detector Receiver . . . . . . . . . . . . . . . . . . . . 29
3.3 Decentralized Power Allocation Algorithms . . . . . . . . . . . . . . . 31
3.3.1 Performance Comparison . . . . . . . . . . . . . . . . . . . . . 36
3.4 Computational Burden Comparison . . . . . . . . . . . . . . . . . . . 40
3.5 Practical Considerations . . . . . . . . . . . . . . . . . . . . . . . . . 43
3.5.1 Shadowing Effects . . . . . . . . . . . . . . . . . . . . . . . . . 44
3.5.2 Illumination Requirements and Dimming Control . . . . . . . 45
3.5.3 Transmitted Power Quantization . . . . . . . . . . . . . . . . 52
3.6 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53
4 Modulation Schemes for VLC Systems 56

4.1 Adaptive M-PAM for Multiuser MISO Indoor VLC Systems . . . . . 57
4.1.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57
4.1.2 Adaptive M-PAM . . . . . . . . . . . . . . . . . . . . . . . . . 59
4.1.3 OCDMA vs TDMA using M-PAM . . . . . . . . . . . . . . . 63
4.1.4 Numerical Results . . . . . . . . . . . . . . . . . . . . . . . . 66
4.2 M-PAM Joint Optimal Waveform Design for Multiuser VLC Systems
over ISI Channel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.2.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
4.2.2 Channel Model . . . . . . . . . . . . . . . . . . . . . . . . . . 72
4.2.3 M-PAM Joint Optimal Waveform Design . . . . . . . . . . . . 75
iii
4.2.4 Numerical Results and Discussions . . . . . . . . . . . . . . . 85

4.3 Comparison of DCO-OFDM and M-PAM . . . . . . . . . . . . . . . . 92
4.3.1 Background . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
4.3.2 Optimized DCO-OFDM . . . . . . . . . . . . . . . . . . . . . 92
4.3.3 Numerical Results and Comparison . . . . . . . . . . . . . . . 97
4.4 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 100
5 Interception Vulnerability Analysis 102

5.1 System Performance and Security . . . . . . . . . . . . . . . . . . . . 104
5.1.1 System Performance . . . . . . . . . . . . . . . . . . . . . . . 104
5.1.2 System Security . . . . . . . . . . . . . . . . . . . . . . . . . . 105
5.1.3 LED Modulation Efficiency Optimization . . . . . . . . . . . . 106
5.2 Simulation Results and Analysis . . . . . . . . . . . . . . . . . . . . . 107
5.2.1 No LED Modulation Efficiency Optimization . . . . . . . . . . 108
5.2.2 LED Modulation Efficiency Optimization to Diminish Vulner-
ability . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 111
5.3 Summary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
6 Conclusions and Future Work 114

6.1 Summary and Conclusions . . . . . . . . . . . . . . . . . . . . . . . . 114
6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 116
List of Figures
1.1 Applications for indoor VLC systems. . . . . . . . . . . . . . . . . . . 3

1.2 4-PAM modulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 6
1.3 Time waveforms for 16-PPM and OOK . . . . . . . . . . . . . . . . . 7
1.4 Orthogonal subcarriers in OFDM. . . . . . . . . . . . . . . . . . . . . 8
1.5 A TDMA stream divided into different time slots for different users. . 9
1.6 An OOC code with the length of 7. . . . . . . . . . . . . . . . . . . . 9
1.7 Subcarriers for multiple users in OFDMA. . . . . . . . . . . . . . . . 10
2.1 Multi-LED lamp model, (a) side view, (b) bottom view . . . . . . . . 15
2.2 Multi-detector model structure, (a) 4-detector model, top and side
view, (b) 7-detector model, top and side view (similar to [2]). . . . . . 15
2.3 Basic indoor VLC channel model . . . . . . . . . . . . . . . . . . . . 15
2.4 Light rays classification. . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.5 LOS light rays model. . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3.1 Top-down view of the two typical user position cases for the small
room. The small circles represent the lamps and the squares represent
the users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
iv
v
3.2 BER performance using “Fairness” and “Min-BER” for Case 1 and 2
for CM-PAJO with a single detector and 7-length OOC codes, in the
small room. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.3 Average BER performance of 4 users for CM-PAJO with different in-
clination angles and different FOV, the peak radiation power to noise
ratio is 48 dB. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 30
3.4 Top-down view of indoor environment. The small circles represent the
lamps and the squares represent the user 1 to user 6. . . . . . . . . . 30
3.5 Average BER performance of different users for CM-PAJO, with peak
radiation power to noise ratio is 48 dB, FOV= 80 degrees. . . . . . . 31
3.6 Histogram of BER performance for 4 randomly distributed users, with
peak radiation power to noise ratio of 51 dB, FOV= 80 degrees . . . 32
3.7 Illumination area and access area (the radius is R) . . . . . . . . . . . 33
3.8 Geometry structure of an example. . . . . . . . . . . . . . . . . . . . 33
3.9 Top-down view of the positions of lamps and users in a large indoor
environment. The small circles represent the lamps and the squares
represent the users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.10 Peak radiation power to noise ratio (PPNR) required for a BER of
10−3 in the large indoor environment using the minimum access area
needed cover the room with 25-length OOC codes. . . . . . . . . . . . 37
3.11 Average BER of 4 users for CM-PAJO and PDM-PAJO for different
radii of access area in the small indoor environment with single detector. 39
3.12 Average BER of 4 users for CM-PAJO and WDM-PAJO for different
radii of access area in the small indoor environment with single detector. 39
3.13 Histogram of BER performance for 4 randomly distributed users, with
peak radiation power to noise ratio of 61 dB . . . . . . . . . . . . . . 40
vi
3.14 Computational burden with the minimum access area in the large in-
door environment, 7 PDs per user, and length of OCDMA code of
7. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42
3.15 Normalized data rate of the user that is blocked under different shad-
owing conditions for a BER = 10−3 . 4 users are in the small indoor
environment, and a single detector is used with length-7 OOC codes. 44
3.16 Average illumination tolerance for different number of virtual users. . 47
3.17 BER comparison with different lighting tolerances, with 4 users and 16
virtual users for the 25-LED lamp case, 400 lx illumination requirements 48
3.18 Illumination distribution comparison of (a) data transmission case and
(b) no data transmission case. The red dots identify the real users,
and the blue dots represent the virtual users, with 10% tolerance. . . 49
3.19 Minimum illumination tolerance under different illumination require-
ments for different LED’s semiangle in the small indoor environment;
16 virtual users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 50
3.20 BER performance for CM-PAJO under different background light il-
lumination conditions in the small indoor environment with 400 lx re-
quired illumination and 7-detector model, 4 users with length-7 OOC
codes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.21 LED grouping scheme for 8 quantization levels . . . . . . . . . . . . . 53
3.22 The average BER performance for different quantization levels with 2,
3, 4 and 5 users using length-7 OOC codes in the small environment,
semiangle is 30 degrees, no dimming control. . . . . . . . . . . . . . . 54
4.1 Block diagram of the proposed adaptive M-PAM algorithm using our
MISO technique . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
vii
4.2 OCDMA and TDMA comparison using M-PAM, K < L case (L = 7

and K = 5). . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
4.3 Top-down view of indoor environment. The small circles represent the
lamps and the squares represent the users. . . . . . . . . . . . . . . . 67
4.4 Average modulation constellation size for adaptive M-PAM for 30 and
40 user cases with shadowing effects. . . . . . . . . . . . . . . . . . . 68
4.5 Average modulation constellation size for adaptive M-PAM modulation
for 30 user and 40 user cases with different desired BER values, no
shadowing effects. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 68
4.6 Average modulation constellation size for different numbers of users
with OCDMA and TDMA techniques, no shadowing effects. . . . . . 69
4.7 Average bit rate for different numbers of users with OCDMA and
TDMA techniques; no shadowing effects. . . . . . . . . . . . . . . . . 69
4.8 A typical indoor VLC system. . . . . . . . . . . . . . . . . . . . . . . 73
4.9 Discrete time version of a truncated LED response. . . . . . . . . . . 74
4.10 Block diagram of the transmitters for the proposed M-PAM joint op-
timal waveform design system. . . . . . . . . . . . . . . . . . . . . . . 75
4.11 Block diagram of the receiver. . . . . . . . . . . . . . . . . . . . . . . 77
4.12 Transmitted data rate for different numbers of samples per waveform
for 3 users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
4.13 SINR for different users. . . . . . . . . . . . . . . . . . . . . . . . . . 87
4.14 Illumination potential for different numbers of users. . . . . . . . . . 88
4.15 SINR comparison of JOW, OCDMA and TDMA for 3 users. . . . . . 89
4.16 SINR for imperfect CSI with different uncertainty variance, Lf = 7
and 3 users. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 91
viii
4.17 Diagram of DCO-OFDM with adjustable modulation index and loaded

bits. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
4.18 Signal constellation with clipping only, 4-QAM. . . . . . . . . . . . . 95
4.19 Signal constellation using (4.49), 4-QAM. . . . . . . . . . . . . . . . . 95
4.20 Signal constellation using (4.49), 16-QAM. . . . . . . . . . . . . . . . 96
4.21 BER comparison of simulation and theoretical results using M-QAM. 97
4.22 Throughput of DCO-OFDM with bit loading for different modulation
indexes. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 99
4.23 Throughput comparison of optimized DCO-OFDM and optimal M-
PAM for a single user. . . . . . . . . . . . . . . . . . . . . . . . . . . 99
5.1 Vulnerable area outside the room . . . . . . . . . . . . . . . . . . . . 103

5.2 SNR distribution in the indoor area . . . . . . . . . . . . . . . . . . . 109
5.3 SNR distribution in the outdoor area, daytime. . . . . . . . . . . . . 109
5.4 Gain of the PD requirements for different received power densities. . . 110
5.5 SNR distribution for outdoor areas by optimizing the LED modulation
efficiency using different sizes of aperture, γmin = 10 dB. . . . . . . . 111
5.6 SNR distribution for indoor areas by optimizing the LED modulation
efficiency, Aout = 1 m2 and Ar = 10−4 m2 . . . . . . . . . . . . . . . . 112
List of Tables
3.1 Parameters Used for Indoor Environment . . . . . . . . . . . . . . . . 27

3.2 Time consumption comparison of centralized and decentralized algo-
rithms in the large room with 50 users. . . . . . . . . . . . . . . . . 43
3.3 BER performance for CM-PAJO with 400 lx required illumination. . 52
4.1 Parameters Used for Large Indoor Environment . . . . . . . . . . . . 66

4.2 Parameters Used for Small Indoor Environment . . . . . . . . . . . . 85
4.3 Maximum data rates using M-PAM averaged over the 3 users . . . . 89
5.1 Parameters Used in Numerical Results . . . . . . . . . . . . . . . . . 108
ix
List of Acronyms
APD: avalanche photo diode

ACO-OFDM: asymmetrically clipped optical OFDM
BER: bit error rate
CSI: channel state information
CDMA: code division multiple access
CM-PAJO: multi-detector PAJO
DEPA: distributed equal power allocation
DCO-OFDM: DC-biased optical OFDM
FOV: field of view
FSO: free space optical
ISI: inter-symbol interference
ITS: intelligent transportation system
JOW: joint optimal waveform
LOS: line of sight
LED: light emitting diodes
LMS-DFE: least mean square decision feedback equalization
MAI: multiple access interference
M-PAM: M-ary pulse amplitude modulation
M-PPM: M-ary pulse position modulation
x
xi
M-VPPM: M-ary variable pulse position modulation

M-VPM: M-ary variable period modulation
M-PAM: M-ary pulse amplitude modulation
MIMO: multiple input multiple output
MISO: multiple input single output
MMSE: minimum mean squared error
NRZ: non-return-to-zero
OFDM: orthogonal frequency division multiplexing
OFDMA: orthogonal frequency division multiple access
OOC: optical orthogonal code
OOK: on-off keying
OCDMA: optical code division multiple access
PAJO: power allocation joint optimization
PDM-PAJO: partial distributed multi-detector PAJO
PADJO: power allocation disjoint optimization
PPNR: peak radiation power to noise ratio
PAPR: peak power to the average power ratio
PD: photodetector
PPM: pulse position modulation
RGB: red green blue
RZ: return-to-zero
RF: radio frequency
SNR: signal to noise ratio
SINR: signal to interference plus noise ratio
TDMA: time division multiple access
VLC: visible light communication
xii
WDM-PAJO: weighted distributed multi-detector PAJO

ZF: zero forcing
List of Symbols
ĥqk : channel gain between LED q and user k

hl (t): impulse response of LEDs
hqk (t): impulse response from LED q to user k
Ar : area of photodetector
2
σthermal : thermal noise variance
2
σshot : shot noise variance
σh2 : channel uncertainty variance
2
σofdm : OFDM signal variance
Pr : receiver optical power
Pb : background optical power
P max : peak transmitted optical power
pqk : allocated power from LED q to user k
Rb : transmitted bit rate
Rs : transmitted symbol rate
Rc : transmitted chip rate
M : modulation constellation size
Mofdm : modulation constellation size for OFDM
K: number of users
Q: number of lamps
xiii
xiv
NQ : number of total LEDs

Nsub : number of subcarriers for OFDM
Tofdm : symbol period for OFDM
V : number of detectors per user
G: gain of PD
Lf : sampling number of waveform
B max : required BER
No : noise power spectral density
nk : additive noise for user k
nclip : clipping noise for OFDM
P: power allocation matrix
C: CDMA code matrix
w: MMSE filter vector
f : waveform vector
H: channel matrix
F: waveform matrix
W: MMSE filter matrix
Ω: optimization thread
∅: dimming parameter
ρ: responsivity
τ : SINR weight for WDM-PAJO
Λ: computational burden
γ: signal to interference plus noise ratio
ς reflection coefficient of walls
: shadowing loss coefficient
η: illumination potential
xv
%/Nsub : modulation index for OFDM

φ: radiation angle
ψ: incident angle
α: nonlinear distortion coefficient for OFDM
ϕ transmittance percentage through normal glass
βq modulation efficiency for LED q
Chapter 1
Introduction
Visible light is defined as the light that is perceptible by the human eye, such as
the light from the sun (red, orange, yellow, green, blue, violet, and so on), which has
a wavelength ranging from 380 nm to 780 nm. Visible light communication (VLC)
is a kind of optical communication that uses the light within the visible wavelength
range to transmit signals. Due to the characteristics of light, VLC systems have
both advantages and disadvantages over radio frequency (RF) systems. Unlike RF
communication systems, there is no spectrum regulation in VLC, and many tech-
niques in RF communications cannot be adopted in VLC directly. In this chapter, we
first introduce reasons for developing VLC systems, and then, research achievements
contained in this dissertation are summarized.
1.1 Background
With the rapid development of handheld technology, high data rate wireless trans-
missions are demanded in our daily lives. Considering the spectrum of RF commu-
nications is so congested, and that the data transmission rate of RF communications
cannot satisfy huge demands for fast data transmissions, VLC has emerged as a new
1
2
possible technology for the next generation communications [3]. VLC systems are
built as dual systems (illumination and data transmission) and have potentially higher
privacy than RF communication systems due to the natural character of light. Light
emitting diodes (LEDs) that work as transmitters in VLC systems have many advan-
tages, such as ease of modulation, high power efficiency and long life expectancy [1].
Since LEDs have been widely used in indoor lighting systems, VLC systems are easy
to be built based on the existing lighting systems.
From a survey, over 80% of the Internet demands occur in indoor environments. In
addition, Wi-Fi systems cannot keep up the increasing demands of data transmissions.
Therefore, an alternative indoor high-speed wireless communication method needs to
be developed. VLC is a promising communication system that can provide high rate
wireless data transmissions and attracts more and more research attention recently.
VLC is not only for indoor communications, but can also be applied to outdoor
short-range communications, such as vehicle-to-vehicle communications. As the num-
ber of vehicles increases every year, urgent actions are needed to prevent and reduce
traffic accidents as well as improving road safety [4]. Since the vehicles’ headlights
and taillights are usually composed of LEDs, the vehicles can communicate with each
other using VLC. Then, an intelligent transportation system (ITS) can be built to
improve road safety and traffic flow based on VLC networks [5].
In addition, VLC can be used for many other applications, such as smart lighting,
mobile connectivity, healthcare, underwater communications, location-based services,
and so on. Applications of VLC have a great potential to increase in the next decades,
and these applications can change the pattern of people’s lives. According to the latest
market research report [6], the VLC market is expected to grow from USD 1.3 Billion
in 2017 to USD 14.91 Billion by 2022, at a compound annual growth rate (CAGR)
of 62.9% between 2017 and 2022.
3
Figure 1.1: Applications for indoor VLC systems.
In this dissertation, we focus on the indoor VLC systems to serve multiple users,
providing them high-speed data transmissions. As shown in Fig. 1.1, in an of-
fice room, multiple users require wireless connections at the same time, and some
may need high data rate transmissions, such as, video chatting, high-definition video
downloading, and so on. Therefore, indoor VLC systems are necessary for multiuser
applications with a high throughput performance.
Communication systems have both uplink and downlink data transmissions. Con-
sidering that users usually require much higher data transmissions over the downlink,
this dissertation only discusses the downlink systems.
1.2 Differences Between VLC and RF Communi-
cations
Due to the different characteristics between visible light and microwave, there are
many differences between VLC and RF systems. In this section, a brief introduction
4
of these differences is presented.
Modulation
The conventional modulation schemes adopted in RF communications cannot be

readily applied in VLC directly because the visible light is non-coherent. For RF
systems, amplitude, frequency, and phase can be used to modulate signals. But
for VLC systems, since intensity modulation and direct detection should be used,
only real non-negative signals can be transmitted. In the following section, some
modulation schemes for VLC are introduced.
Flicker and Dimming
Indoor VLC systems are built based on the existing lighting systems; therefore,
illumination is a very important function for VLC. For illumination, a non-flickering
and dimming controllable system is desired. Flicker and dimming both depend on
the light intensity, thus, transmitting data using intensity modulation has an effect
on flicker and dimming. When designing indoor VLC systems, dimming control and
flicker are two essential considerations to be taken into account.
Limitations of LEDs
For VLC systems, LEDs are usually used as transmitters. High power LEDs are
usually required for lighting; however, lighting LEDs have a low rise time that leads
to a bandwidth limit. The narrow bandwidth of LEDs limits the data transmission
rate.
A forward current drives LEDs, and a higher current should stimulate more op-
tical power. However, LEDs are non-linear devices due to a peak transmitted power
constraint. The peak power constraint may distort the signals. When the value of
5
the signal exceed the constraint, it must be clipped.
1.3 Modulation Schemes for VLC
In this section, some modulation schemes that can be used for VLC systems
are introduced. Unlike RF communications, only intensity modulation and direct
direction can be used in VLC.
1.3.1 On-Off Keying
On-off keying (OOK) is the simplest technique that can be used in VLC systems.
In OOK, the intensity of an optical source is directly modulated by the information
sequences which is usually binary. For a sequence, a bit “one” can be represented by
an optical pulse, referred to as “on”. On the contrary, a bit “zero”, referred to as
“off”, can be represented as a blank duration.
1.3.2 M -ary Pulse Amplitude Modulation
M -ary pulse amplitude modulation (M-PAM) can offer a higher bandwidth effi-
ciency than OOK, since more bits can be transmitted using one pulse in M-PAM. In
M-PAM, a pulse is sent in each symbol duration, where the pulse amplitude takes on
one of the M possible levels, typically {0, M1−1 , M2−1 , · · · , 1}. The number of bits per
symbol transmitted is log2 M .
An example of 4-PAM modulation is given in Figure 1.2 to help us understand
the principle of the M-PAM scheme. The data stream ready for transmission is
“111001001001” in this figure. For 4-PAM, we divide the stream into groups contain-
ing 2 bits. Here, the stream can be divided into “11”, “10”, “01”, “00”, “10”, “01”.
Converting these binary numbers into 4-ary ones, we get 0, 1, 3, 2, 1, 3. For 4-PAM,
6
1 1 1 0 0 1 0 0 1 0 0 1
0 1 3 2 1 3
Level 3
Amplitude
Level 2
Level 1
Time
Figure 1.2: 4-PAM modulation
the transmitter just needs to send the corresponding power levels.
1.3.3 Pulse Position Modulation
Pulse position modulation (PPM) that is illustrated in Fig. 1.3 is an orthogonal

modulation technique that requires less power than OOK for the same bit rate. The
PPM uses the pulse position information to modulate data. The example in Fig. 1.3
shows that the pulse in the 6th time slot in 16-PPM represents the data “0101”. To
achieve a same bit rate as OOK, since the PPM has more narrow pulse duration than
OOK, more bandwidth is required by PPM. Thus, since the LEDs are bandlimited,
the PPM scheme may introduce some intersymbol interferences (ISI). ISI is a form
of distortion of a signal in which one symbol interferes with subsequent symbols. In
addition, PPM requires high synchronization accuracy, which is difficult in practice;
therefore, PPM is not considered in this dissertation.
1.3.4 Orthogonal Frequency Division Multiplexing
Orthogonal frequency-division multiplexing (OFDM) is a method of modulating

digital data on multiple orthogonal subcarriers as shown in Fig. 1.4. OFDM is a pop-
7
Symbol 1 Symbol 2
PPM
OOK
0 1 0 1 1 1 1 0
Figure 1.3: Time waveforms for 16-PPM and OOK
ular modulation scheme that has been widely used in RF systems. The conventional
OFDM generates complex-valued bipolar signals, which needs to be modified in order
to become suitable for VLC systems. A real OFDM signal can be obtained by using
Hermitain symmetric data, which reduces the system bandwidth by a half. This ap-
proach has been widely accepted in the literature for the generation of a real OFDM
signal. The resulting waveform, however, is still bipolar in nature (it has positive
and negative parts). A number of techniques have been proposed for the creation of
a unipolar signal. For example, a DC bias can be add to the original bipolar signal.
This scheme is known as DC-biased optical OFDM (DCO-OFDM) [7]. Another ap-
proach is asymmetrically-clipped optical OFDM (ACO-OFDM), which has become
popular recently because of its high power efficiency [8]. In this scheme, only the
odd-indexed sub-carriers in the OFDM frame are modulated with information.
A characteristic of OFDM is a high peak to the average power ratio (PAPR). A
high PAPR results in a severely distorted transmitted signal. Since LEDs have peak
transmitted power constraint, the signals that exceed the constraint must be clipped.
In this dissertation, the performance of OFDM in VLC systems is discussed.
8
Subcarriers
Frequency
Figure 1.4: Orthogonal subcarriers in OFDM.
1.4 Multiple Access for VLC
For indoor VLC systems, multiuser wireless connection is a big issue that needs
to be considered. Often, more than one user needs Internet access simultaneously.
Therefore, a reliable multiple access technique is required. Since VLC uses intensity
modulation, some multiple access techniques are different than RF. In this section,
some multiple access techniques are introduced.
1.4.1 Time Division Multiple Access
Time division multiple access (TDMA) is a conventional multiple access scheme

that is used in the digital 2G cellular system such as Global System for Mobile
Communications (GSM) in RF. TDMA in VLC is the same as in RF communications.
The users in TDMA transmit signals in rapid succession, one after the other, each
using its own time slot as shown in Fig. 1.5. The advantages for TDMA are very
obvious. The structures of transmitter and receiver can be designed simply. There
is almost no multiple access interference (MAI) that is caused by the data for other
users if synchronization between different users can be guaranteed.
9
TDMA User 1 User 2 …… User K User 1 User 2 ……

Time
Figure 1.5: A TDMA stream divided into different time slots for different users.
1 2 3 4 5 6 7
Figure 1.6: An OOC code with the length of 7.
1.4.2 Optical Code Division Multiple Access
Code division multiple access (CDMA) can be employed to provide multiple access
for simultaneous users, which has been applied successfully in RF communications
for many years. To separate users, each user has a unique code, and the codes for
different users can be designed to be nearly orthogonal, therefore, the MAI can be
perfectly canceled. For VLC systems, due to intensity modulation, the CDMA codes
cannot be orthogonal, and the non-orthogonal codes cannot avoid MAI. To implement
the CDMA technique in VLC, choosing the proper optical CDMA (OCDMA) code
is a significant step. In this dissertation, we choose optical orthogonal codes (OOC)
for OCDMA. Fig. 1.6 shows an example of one of the OOC codes used.
1.4.3 Orthogonal Frequency Division Multiple Access
Orthogonal frequency-division multiple access (OFDMA) is a novel multiple access

scheme based on OFDM technique that is used in 4G LTE. As shown in Fig. 1.7,
multiple access in OFDMA can be achieved by assigning subsets of subcarriers to
individual users, which is a multi-user version of OFDM. Thus, OFDMA also has high
PAPR, and the peak power limited LEDs can introduce hard-clipped distortions.
10
User 1 Data Subcarrier User 2 Data Subcarrier
……
Frequency
Figure 1.7: Subcarriers for multiple users in OFDMA.
1.5 Literature Review and Dissertation Contribu-
tions
For a multiuser VLC system, one of the biggest challenges is how to reduce the
MAI and achieve a good communication performance. Three directions have emerged
to address this problem. One is to use a multiple input multiple output (MIMO)
technique with a precoding algorithm [9], [10], and [11]. However, since only the real
non-negative signals can be transmitted by LEDs, the procoding algorithms usually
need to add a DC bias. The second direction is to use color-shift-keying modulation
over red-green-blue (RGB) LEDs and CDMA to support multiple users [12]. The
third trend is to use OFDMA and discrete multi-tone (DMT) modulation to support
multiple users [13] and [14]. However, since the OFDMA and DMT have high PAPR,
severe distortion may be introduced. In this dissertation, we propose to use OCDMA
with resource allocation algorithms for multiple users. For our algorithms, the MAI
is minimized by optimizing the signal to interference plus noise ratio.
High-speed wireless data transmission is another hot topic in indoor VLC re-
search [15,16]. To increase the data transmission rate, there are three research trends.
One is to use high bandwidth efficiency pulsed modulation schemes, such as M-PAM,
M-ary pulse position modulation (M-PPM), M-ary variable pulse position modulation
11
(M-VPPM) and M-ary variable period modulation (M-VPM) [17–19]. These M-ary
modulation schemes have a (log2 M )-fold increase in data rate compared with OOK.
Equalization is the reversal of distortion caused by a bandlimited channel, which is
another research trend to reduce the ISI and increase the data rate [20–22]. Using an
OFDM technique to increase the bit rate is the third direction [23, 24]. For OFDM,
since complex modulation schemes, such as M -ary quadrature amplitude modulation
(M-QAM) can be used for subcarriers, the transmission rate can be increased. How-
ever, for OFDM, the high PAPR is a drawback. In this dissertation, we propose to use
M-PAM and equalization to achieve a high data rate for bandlimited VLC systems.
In general, this dissertation discusses multiuser indoor VLC systems. We propose
some algorithms and analyze the performance for indoor VLC systems. The main
contributions of this dissertation are listed below, along with the publications in which
the work appears:
• Multi-LED lamp and multi-detector receiver models are proposed to support

multiuser VLC systems [25].
• Impulse response of indoor VLC channel and received power distributions in

the indoor area are discussed [26].
• Centralized and decentralized power allocation algorithms are proposed for mul-
tiuser VLC systems [26–28].
• Some practical considerations, such as shadowing effects, illumination require-

ments, power quantization and computational burden are discussed in this dis-
sertation [28].
• An adaptive M-PAM modulation scheme with power allocation technique for

multiuser VLC systems is introduced [29].
12
• To reduce the ISI and MAI, a waveform design and minimum mean squared
error (MMSE) equalizer joint optimization algorithm is proposed [30, 31].
• Imperfect channel state information case is discussed in the waveform design

algorithm [31].
• DCO-OFDM is discussed in this dissertation. An optimal modulation index and

bit loading technique is applied in DCO-OFDM to maximize the transmission
bit rate [32].
• A comparison of the optimized DCO-OFDM and M-PAM modulation is given

[32].
• System vulnerability is analyzed [33].
With the help of the multi-LED transmitters and multi-detector receivers, the
centralized and decentralized power allocation algorithms can support multiple users
and achieve a good performance. Compared with the centralized algorithm, the de-
centralized algorithms have a much lower computational burden. In addition, some
practical considerations are taken into account, and with standard indoor illumina-
tion level, many users can be successfully served by using the proposed algorithms.
Although the lighting LEDs have limited 3 dB bandwidth, using M-PAM and wave-
form design algorithm, high data transmissions can still be achieved. Compared with
DCO-OFDM, we conclude that using M-PAM and waveform design algorithm can
provide a higher bit rate.
1.6 Dissertation Organization
The remainder of the dissertation is organized as follows.

13
Chapter 2 describes the multi-LEDs transmitter and multi-detector receiver mod-

els. Based on these models, the indoor channel impulse response is derived.
In Chapter 3, a centralized and four decentralized power allocation algorithms are
introduced to support multiple users in indoor VLC environments. In this chapter,
some practical considerations are discussed, such as shadowing effects, illumination
requirements, computational burden and power quantization.
In Chapter 4, an adaptive M-PAM modulation scheme is described to support
multiple users with the power allocation algorithm. To reduce the ISI caused by the
bandlimited LED, a waveform design and MMSE filter joint optimization algorithm
is described. MAI can also be reduced by the waveforms. An optimized DCO-OFDM
technique is introduced, and a comparison of DCO-OFDM and M-PAM with the
waveform design algorithm is discussed.
In Chapter 5, considering the physical characteristics of light, we analyze the
interception vulnerability of VLC systems.
Chapter 6 summarizes the dissertation. Conclusion and future work are also
described.
Chapter 2
System Description
White LEDs are now widely used as lighting sources for indoor illumination pur-
pose because of their high power efficiency and long life expectancy. For VLC systems,
white LEDs can also be used as transmitters. In this chapter, a multi-LED trans-
mitter and a multi-detector receiver models are introduced. With the help of the
proposed transmitter and receiver, the system robustness can be improved.
2.1 Transmitter and Receiver Models
We consider a transmitter fixture as illustrated in Fig. 2.1: a multi-LED lamp

model consisting of multiple LEDs with different inclination angles [26]. This struc-
ture is proposed to cover more illumination area and provide more power to the corner
areas. Furthermore, each LED can be configured to transmit light either bearing or
not bearing information, giving the system more control.
The multi-detector model used as a receiver is shown in Fig. 2.2, where each
photodetector has a different orientation that depends on its inclination angle [2].
The multi-detector receiver can capture the incident light from different directions. If
the incident light from one direction is blocked, with the help of this receiver structure,
14
15
Figure 2.1: Multi-LED lamp model, (a) side view, (b) bottom view
Figure 2.2: Multi-detector model structure, (a) 4-detector model, top and side view,
(b) 7-detector model, top and side view (similar to [2]).
the user can still receive light with data from other directions.
2.2 Channel Model
For indoor VLC systems, white LEDs work as transmitters and photo-detectors
work as receivers. Since the visible light is non-coherent, intensity modulation and
direct detection are employed in VLC systems. At the receiver, the received signal
can be represented as
h(t)
Optical current signal
x(t) Transmitted signal
y(t)
Noise n(t)
Figure 2.3: Basic indoor VLC channel model

16
y(t) = ρAr x(t) ∗ h(t) + n(t), (2.1)
where ρ represents the responsivity that measures the electrical output per optical
input. Ar is the area of the photodetector. ∗ is convolution, x(t) is the transmitted
optical intensity, n(t) is the additive noise, and h(t) is the indoor channel impulse
response.
Because of the principles of optics, the light rays from the transmitter can be
classified into two parts. They are the line of sight (LOS) rays and diffused rays, as
shown in Figure 2.4. These two components cause the multi-path effect in indoor
VLC systems. Thus, the indoor VLC channel gain from LED q to user k can be
approximated by [34]
(LOS) (Diff)
ĥqk = ĥqk + ĥqk , (2.2)
(LOS)
where ĥqk is the contribution due to the LOS, which depends on the distance be-
(Diff)
tween transmitter and receiver and on their orientation with respect to the LOS. ĥqk
is the diffused part, the intensity of which is less than the LOS part. The intensity of
the LOS rays and diffused rays follow the Lambertian law. The Lambertian radiant
intensity model can be defined as [35]

m+1


2π
cosm (φ) for φ ∈ [−π/2, π/2]
R0 (φ) = , (2.3)
 0 for |φ| ≥ π/2

where m is the Lambertian mode of the light source and φ is the radiation angle for
the transmitter as shown in Figure 2.5. The maximum radiated power is reached
when φ = 0. The Lambertian mode m is related to the LED’s semiangle Φ1/2 by
ln 2
m= . (2.4)
ln(cos Φ1/2 )
17
Transmitter
LOS ray
Diffused
surface
Diffused ray
FOV
Receiver
Figure 2.4: Light rays classification.
Transmitter
d 

FOV
Receiver Ar
Figure 2.5: LOS light rays model.
The detector effective area can be modelled as a function of the incident angle, ψ,
as [35] 
 Ar cos ψ
 −π/2 ≤ ψ ≤ π/2
Aef f (ψ) = , (2.5)
 0 | ψ |> π/2

We assume that the detector cannot be active beyond the field of view (FOV) angle
Ψc . Ar is the area of the photodetector at receiver. Therefore, the LOS link gain
18
between LED q and user k can be described as


(m+1)
(LOS)


2πl2
cosm (φ) cos(ψ) −Ψc ≤ ψ ≤ Ψc
ĥqk = , (2.6)
 0

elsewhere
where l is the distance between the transmitter q and the kth receiver. φ is the radi-
ation angle, and ψ represents the incident angle. The diffused part can be calculated
as
∞
Y
(Diff)
ĥqk = Li ς i , (2.7)
i=0
where ς is the wall reflection coefficient, and Li represents the ith bounce link atten-
uation,
(m + 1) cosm (φ0 ) cos(ψ0 )
L0 =
2πl02
cosm (φ1 ) cos(ψ1 )
L1 =
πl12 , (2.8)
..
.
cosm (φk ) cos(ψk )
Lk =
πlk2
where lk represents the distance of the kth bounce link. φk and ψk are radiation angle
and incident angle at kth bounce’s diffusion point, respectively. [36].
2.3 Additive Noise
The noise in this system can be modeled as thermal noise plus shot noise. The
thermal noise and shot noise can be represented as
2
σthermal = 4κTκ Rs /RL ,
(2.9)
2
σshot = 2qρPr Rs
19
where q is the electronic charge, κ is Boltzmann’s constant, Tκ is the absolute ther-

modynamic temperature, and Rs is the transmitted symbol rate. RL is the resistor
in the circuit of the receiver. Pr is the received optical power that can be calculated
as
Q
X
max
Pr = P Ar ĥqk + Pb , (2.10)
q=1
Pb is the received optical power of background light.
2.4 Summary
This chapter introduces multi-LED and multi-detector models and VLC channel
model, as well as the VLC channel can be classified into LOS and non-LOS compo-
nents. Both the LOS and non-LOS follow the Lambertian law.
Chapter 3
Power Allocation Algorithms for

Multiuser VLC Systems
In indoor VLC systems, one significant research challenge that has received some
attention in recent years is how to support multiple users with high data rates while
limiting the multiple access interference (MAI). So far, three popular research trends
have emerged. Multiple input and multiple output (MIMO) has been proposed to
use in VLC systems as a method for multiplying the capacity [37–39]. MIMO with
precoding is proposed to limit the MAI and improve the signal to interference plus
noise ratio (SINR) in [40–42]. The second trend is to use color-shift-keying modulation
over red-green-blue (RGB) LEDs and code division multiplexing access (CDMA) to
support multiple users [12]. The third direction is to use resource allocation schemes
to minimize the MAI. In the third trend, orthogonal frequency division multiple
access (OFDMA) and discrete multi-tone (DMT) modulation with transmitted power
allocation algorithms to limit the MAI were proposed in [14, 43, 44].
Due to the nature of white LEDs (their nonlinearity and the incoherent light they
transmit), it is not easy to implement a modulation requiring frequency-domain pro-
20
21
cessing. To avoid this problem, intensity modulation and direct detection (IM/DD)
with on-off keying (OOK) modulation is applied in this chapter. Then, direct-
sequence optical CDMA (OCDMA) with a time-space minimum mean squared error
(MMSE) filter is used to support multiple users. OCDMA has considerable advan-
tages compared with the recently popular orthogonal frequency-division multiplexing
(OFDM) technique [45–47]. Since OFDM has a high peak to average power ratio
(PAPR), some signals with high intensity would be distorted from the nonlinearity of
the LEDs. Furthermore, the structure of the receivers is simple for OCDMA systems
compared with OFDM.
In this chapter, we propose a centralized power allocation algorithm and several
decentralized power allocation algorithms for multiple users in indoor VLC environ-
ments. The algorithms we propose in this chapter have the following advantages
compared with other approaches
• All the transmitted power is used for both data transmission and illumination
(no extra light needed just for illumination).
• Compared with the OFDM technique, our algorithms do not need to address
the high PAPR.
• No DC bias is needed for the transmitted signals.
• The structures of transmitters and receivers are simple.
In addition, we propose to model the shadowing effects as path losses in this chapter.
Our adaptive algorithms can reallocate the transmit power and recompute the MMSE
filters coefficients to reduce the shadowing effects with the help of MIMO processing.
Some of the work presented in this chapter has been published in [26, 27, 29].
22
3.1 Transmitted and Received Signals
We assume that the indoor VLC network has N lamps, and there are Q LEDs
with different inclination angles for each lamp. Therefore, the number of total LEDs
is N × Q = NQ . We also assume that there are K users in the indoor environment,
and each user has V PDs with different orientations.
Let ik (t) be the signal that is intended for user k, which is represented as ik (t) =
dk · ck (t), where dk is the {0, 1} data, and ck (t) is the OCDMA code waveform for
user k. The qth LED sends a linear combination of the users’ data as
K
X
xq (t) = pqk ik (t), (3.1)
k=1
where pqk ∈ [0, pmax ] is the transmitted power of the qth LED allocated to transmit-
ting the data of user k. Assuming a peak radiation power limit of pmax from each
LED, the constraint K max
P
k=1 pqk ≤ p needs to be applied on the allocated powers.
These power levels are organized in a NQ × K matrix denoted as P. The elements in
matrix P represent the power allocation from each LED to each user.
The signal received by the vth detector of user k can be written as [25, 26]
NQ
(v)
X (v)
k = 1, . . . , K
rk (t) = ĥqkv xq (t) + nk (t), (3.2)
q=1 v = 1, . . . , V
(v)
where nk (t) is the noise experienced by the vth detector of user k. ĥqkv is the channel
gain from LED q to the vth detector of user k. In this chapter, shot noise from
ambient light and thermal noise are considered. Then, after chip matched filtering
23
and sampling, the `th sample of the discrete time signal received by PD v of user k is
NQ
(v)
X (v)
k = 1, . . . , K
rk [`] = ĥqkv xq [`] + nk [`]. (3.3)
q=1 v = 1, . . . , V
We design a linear time-space MMSE filter for user k, wk = (wk1 , wk2 , · · · , wkL )T ,
where wk` = (wk [1, `], wk [2, `], · · · , wk [V, `]), ` = 1, 2, · · · , L. Therefore, the
length of wk is V L, where L is the length of the OCDMA code. This time-space
MMSE filter can take advantage of the received signal from all the PDs. After the
linear MMSE filter, the received decision variable for user k can be represented as
L X
X V
(v)
yk = rk [`]wk [v, `] + bk , (3.4)
`=1 v=1
where bk is a constant for the linear MMSE estimator. From (3.1)-(3.4), the decision
variable for user k after MMSE filtering can be rewritten in a matrix form as
yk = g(CT DPT HTk )T wk + nTk wk + bk , (3.5)
where g(·) is a transformation to reshape the matrix into a V L-vector by concatenat-

ing the columns. In (3.5), D = diag(d1 , d2 , · · · , dK ), and nk is the noise vector. C, P
and Hk are the OCDMA, power allocation, and channel gain matrices, respectively.
They are represented as
 
 c1 [1] c1 [2] · · · c1 [L] 
 
 c2 [1] c2 [2] · · · c2 [L] 
 
C= . , (3.6)
 
 .. .. .. .. 
 . . . 
 
 
cK [1] cK [2] · · · cK [L]
24
 
 p11 p12 ··· p1K 
 
 p21 p22 · · · p2K 
 
P= . , (3.7)
 
 .. .. ... .. 
 . . 
 
 
pNQ 1 pNQ 2 · · · pNQ K
and  
 ĥ1k1 ĥ1k2 ··· ĥ1kV 
 
 ĥ2k1 ĥ2k2 · · · ĥ2kV 
 
Hk =  . . (3.8)
 
 .. .. .. ..
 . . . 

 
 
ĥNQ k1 ĥNQ k2 · · · ĥNQ kV
The time-space MMSE receiver in (3.5) can be derived as follows. The mean-
squared error Jk for user k is defined as
Jk = Ed,n {(g(CT DPT HTk )T wk + nTk wk + bk − dk )2 }, (3.9)
where Ed,n represents expectation with respect to the data vector d and the noise
∂Jk ∂Jk
nk . Solving for ∂b
= 0, and ∂wk
= 0, the MMSE receiver can be obtained as
wk = (G + σ 2 I)−1 g(CT Σk PT HTk )

, (3.10)
1 1 T T T T
bk = − g(C P Hk ) wk
2 2
where G = Ed {g(CT DPT HTk )g(CT DPT HTk )T }, and I is the identity matrix. σ 2
represents the noise variance. Σk = Ed {Ddk }.
From (3.5), the signal after the MMSE estimator consists of three parts: the target
(intended data) for user k, the MAI and the noise. Thus, the received signal for user
25
k after MMSE filtering can be represented as
yk = g(CT DZk PT ĤTk )T wk + g(CT DẐk PT ĤTk )T wk + nTk wk +b, (3.11)

| {z } | {z } | {z }
Target MAI Noise
where Zk is defined as a matrix with a 1 in its (k, k)th element and zeros in all other
places, and Ẑk = I − Zk .
3.2 Centralized Power Allocation Algorithm
In this section, we describe a centralized power allocation joint optimization algo-

rithm (CM-PAJO) and several decentralized algorithms.
For CM-PAJO, we assume that each LED serves all the users in this indoor en-
vironment. In order to eliminate the MAI, each lamp needs to exchange information
(channel information sent back from users) with other lamps and allocates power to
the users jointly.
From (3.11), the SINR for user k can be calculated as,
Signal
SINRk =
MAI + σ 2 wkT wk
Signal = A2r wkT Ed {g(CT DZk PT HTk )g(CT DZk PT HTk )T }wk ). (3.12)
MAI = A2r wkT Ed {g(CT DẐk PT HTk )g(CT DẐk PT HTk )T }wk )
The bit error rate for user k can be approximated by [34]
r !
1 SINRk
BERk ≈ erfc . (3.13)
2 2
To optimize the transmitted power allocation, we consider two optimization crite-

ria: to minimize the maximum BER among all the users or to minimize the average
26
of BER for all the users. Through optimization, we obtain the power allocation as
Fairness:P∗ = arg min max BERk (3.14)

P k
or
X
Min-BER:P∗ = arg min BERk , (3.15)
P
k
where P∗ is the optimal power allocation.
Algorithm 1: Optimal power allocation for “Fairness”

min max BERk ⇒ min y, s.t. BERk ≤ y, ∀ k;
Use method of Lagrange multiplier;
Equivalent objective function £(P, y, λi ) is created;
while i ≤ R, R is number of random initial values do
Initialization: random initial value Pi ;
SQP begins;
repeat
SQP algorithm;
until £(P, y, λi ) converges;
Get P∗i for initial value Pi ;
end
Output: Choose the P∗i that yields the smallest value of y
To find the optimal solutions to (3.14) and (3.15), an iterative method, the se-
quential quadratic programming (SQP) algorithm, can be used. For the “Fairness”
optimization in (3.14), the objective function can be reformulated into an equiva-
lent nonlinear programming problem by appending additional constraints of the form
BERk ≤ y for ∀ k, and then minimizing y over P. The method of Lagrange multi-
27
Table 3.1: Parameters Used for Indoor Environment
Number of lamps for small room 4

Number of lamps for large room 25
Number of LEDs in each lamp 25
PD number per user 1, 4, 7
Dimming parameters for all LEDs ∅=1
Responsivity 0.5A/W
Area of the PD 0.01 cm2
Wall reflection coefficient 0.8
Radiation optical power of each lamp 300 mW
LED semiangle 30o
Cyclic 7-length OOC code index [48] {1, 2, 4}
Cyclic 25-length OOC code index [48] {1, 2, 7} {1, 3, 10}
{1, 4, 12} {1, 5, 14}
Minimum access area am = 9.8 m2
Receiver bandwidth 70 MHz
pliers is used to tackle all constraints. Since the two optimizations are non-convex
problems, the optimal solution may be a local minimum. Therefore, we randomly
choose different initial values for optimization and choose the best solution from all
results. The steps for solving the power allocation algorithm for the “Fairness” cri-
teria is described in Algorithm 1. The steps for solving the “Min-BER” criteria are
similar.
To test the applicability of the system in different environments, we show results
for both small and large rooms. The parameters used to obtain the numerical results
are shown in Table 3.1. This is a baseline for all the numerical results in this chapter.
3.2.1 Single Detector Receiver
The single detector receiver case is considered first in this section.

For a small indoor environment, we consider two typical lamp and user positions,
28
5m 5m
2.5 m 2.5 m
2.5 m
2.5 m
5m
5m
1.
1.
77
77
m
m
Case 1 Case 2
Figure 3.1: Top-down view of the two typical user position cases for the small room.
The small circles represent the lamps and the squares represent the users.
0
10
−1
10
Case 1
Case 2
BER
−2
10
−3
10
Worst BER user, Fairness
Best BER user, Fairness
Worst BER user, Min−BER
Best BER user, Min−BER
−4
10
35 40 45 50 55 60 65 70
Peak radiation power to noise ratio (dB)
Figure 3.2: BER performance using “Fairness” and “Min-BER” for Case 1 and 2 for
CM-PAJO with a single detector and 7-length OOC codes, in the small room.
29
which are shown in Fig. 3.1. In Case 1, all the users are located in a corner near one of
the lamps. In Case 2, all the users are distributed in the room. The numerical results
for the BER of the CM-PAJO using the “Fairness” and “Min-BER” optimization
criteria from (3.14) and (3.15) for Cases 1 and 2 are shown in Fig. 3.2. The BER
curves can be represented as a function of the peak radiation power to noise ratio
(PPNR), which is defined as P max /σ 2 .1 Using the “Fairness” criterion, the BER
curves for all users are more similar than using the “Min-BER” criterion, as excepted.
At a BER of 10−3 , there is approximately a 3 dB required transmitted power gap
between the best and worst-case users for Min-BER both in Cases 1 and 2. Since
the Min-BER method minimizes the average BER for all users, the average BER
using Min-BER is slightly better than using “Fairness”, by 1 dB. But when equal
performance is desired, the “Fairness” method is preferable. Case 2 always has a
better BER than Case 1 because the users’ locations make better use of all lamps.
3.2.2 Multiple Detector Receiver
In this section, multi-detector receivers are applied and tested.

Fig. 3.3 shows the BER performance of CM-PAJO with different receiver inclina-
tion angles from 20 degrees to 60 degrees. We simulated 5 trials of 4 users random
distributed in the indoor environment. From the results, the FOV impacts the BER
performance more than the inclination angles. In general, with the help of our pro-
posed algorithm, the larger the FOV the better the BER performance for CM-PAJO.
In addition, the 7-detector CM-PAJO is always superior to the 4-detector one.
Fig. 3.5 shows that when the number of users increases (selected in the order
shown in Fig. 3.4), the BER performances of CM-PAJO become worse, due to an
1
Note that in VLC systems we use the transmitted power to receiver noise ratio as an SNR metric,
instead of the normal received power to receiver noise ratio [25–27].
30
−2
10
−3
10
BER
−4
10
4−detector, FOV=20 degrees
−5 7−detector, FOV=80 degrees
10
10 20 30 40 50 60
Inclination angle (degree)
Figure 3.3: Average BER performance of 4 users for CM-PAJO with different in-
clination angles and different FOV, the peak radiation power to noise ratio is 48
dB.
5m
6
5m
1
4 5
3 2
Figure 3.4: Top-down view of indoor environment. The small circles represent the
lamps and the squares represent the user 1 to user 6.
31
−1
10
−2
10
−3
10
BER
−4
10
−5
10
single-detector
4−detector
7−detector
−6
10
2 2.5 3 3.5 4 4.5 5 5.5 6
Number of users
Figure 3.5: Average BER performance of different users for CM-PAJO, with peak
radiation power to noise ratio is 48 dB, FOV= 80 degrees.
increase in MAI. Again the 7-detector CM-PAJO has better BER performance than
the 4-detector case.
To analyze the performance of the proposed CM-PAJO from a statistical point of
view, we simulate 40 trials of 4 users randomly distributed in the indoor environment.
The simulation results are shown in Fig. 3.6. For a peak radiation power to noise
ratio of 51 dB, more than 75% of the 160 users’ BER for both the 4-detector and
7-detector cases are lower than 10−3 .
3.3 Decentralized Power Allocation Algorithms
In a large room with many LED lamps, the centralized algorithm presented above
becomes prohibitively and unnecessarily complicated. In this section, we describe
four decentralized power allocation algorithms better suited to such environments.
For the decentralized algorithms, we define a circular access area for each lamp,
32
Figure 3.6: Histogram of BER performance for 4 randomly distributed users, with
peak radiation power to noise ratio of 51 dB, FOV= 80 degrees
which is shown in Fig. 3.7. This artificially-defined access area is smaller than the
actual illumination area of the lamps such that the lamps can serve only the users who
are in the access area. To cover the entire indoor area, there may be some overlap of
the access areas from different lamps. Each user must be served by at least one lamp,
and each lamp can serve more than one user. An example is shown in Fig. 3.8, where
there are 4 lamps and 5 users, and each lamp has an access area as drawn. Given the
locations of the users, users A and B are in the access area of lamp 1. Users B and
C are in the access of lamp 2. User D is in the overlap access area of lamps 3 and 4.
User E is in the access area of lamp 3. In this case, since user B is in the overlapping
access area of lamps 1 and 2, it can be served by these two lamps. Similarly, user D
can be served by both lamps 3 and 4.
Unlike the centralized algorithm, the decentralized VLC optimization can be di-
vided into parallel optimization threads. For each optimization thread, the transmit
power allocation and filter design work independently from the other threads. In
33
Figure 3.7: Illumination area and access area (the radius is R)
User A
User B User C
Lamp 1 Lamp 2
Access
area
User D
Lamp 3 Lamp 4
User E
Figure 3.8: Geometry structure of an example.
addition, when we calculate the SINR for each user, we only consider the messages
within the thread (so the MAI is assumed to be caused only by the users in the same
thread). We use OCDMA as our multiple-access scheme because it can allow each
thread to ignore other threads, even if they cause some interference. However, for
TDMA and OFDMA, interference can be catastrophic. Since each thread works indi-
vidually, there is no channel information exchange between the different optimization
threads. For all techniques, each lamp must know the data and channel state infor-
mation for the users in its access area, and all lamps must remain synchronized since
a user may receive its signal from more than one lamp.
34
Decentralized Equal Power Allocation (DEPA)
For DEPA, each lamp works independently and allocates the transmitted power
equally to the users in its access area. If there are no users in an access area, the
transmitted power is used for lighting only.
In the example displayed in Fig. 3.8, for DEPA, lamp 1 allocates equal transmitted
power to users A and B. Similarly, lamps 2 and 3 allocate power to each user in their
access areas equally. Since there is only one user in the access area of lamp 4, all the
power is allocated to that user.
Power Allocation Disjoint Optimization (PADJO)
All the lamps work independently in PADJO, and each lamp optimizes the power
allocated to the users in its own access area using (3.14) or (3.15). Since we assume
there are N lamps in the indoor environment, there are N optimization threads, and
all threads can work in parallel. Similar to DEPA, there is no channel information
exchange between lamps.
Using PADJO, all the lamps and users in the example shown in Fig. 3.8 can be
divided into four optimization threads. Thread 1 consists of lamp 1 and users A and
B. Thread 2 consists of lamp 2 and users B and C. Thread 3 consists of lamp 3 and
users D and E. Thread 4 contains lamp 4 and user D. The four optimization threads
work independently. Thus, when the algorithm calculates the SINR for each user in
a particular thread, it only consider the messages within the thread.
Weighted Decentralized Multi-detector Power Allocation Joint Optimiza-

tion (WDM-PAJO)
For WDM-PAJO, all the lamps work independently. They need to know how
many access points serve each users, yet there is still no channel information exchange
35
between lamps. Thus, there are N threads for WDM-PAJO. The SINR for each user
is weighted by τk to normalize for the extra power received by users that are served
by multiple lamps. The algorithm calculates
p
P∗Ω(i) = arg min max Q τk · SINRk , ∀ i, (3.16)
W P (i)
k∈ΩW
which is similar to the PADJO, except it accounts for the number of lamps that
(i)
serve user k, denoted as τk . ΩW represents the ith WDM-PAJO optimization thread.
P∗ (i) is the optimal power allocation matrix for the lamps in the ith thread using
ΩW
WDM-PAJO.
Similar to PADJO, all the lamps and users in Fig. 3.8 can be divided into four opti-
mization threads for WDM-PAJO. In this example, when we optimize the transmitted
power in thread 1 using (3.16), τA = 1, τB = 2 and τD = 2, because there are two
lamps that serve users B and D. In this case, the optimization threads 1, 2, 3 and 4,
(1) (2)
can be represented as ΩW = {lamp 1, user A, user B}, ΩW = {lamp 2, user B, user C},
(3) (4)
ΩW = {lamp 3, user D, user E} and ΩW = {lamp 4, user D}, respectively.
Partial Decentralized Multi-detector Power Allocation Joint Optimization

(PDM-PAJO)
In PDM-PAJO, the lamps and users are divided into different optimization threads
depending on the users’ locations. Different from PADJO, the lamps that serve the
same users can exchange channel information in PDM-PAJO. Therefore, the lamps
that work together form an optimization thread.
For PDM-PAJO, the optimization process for a thread is similar to the CM-PAJO
36
case, which can be described as
p
P∗Ω(i) = arg min max Q SINRk , ∀ i, (3.17)
P P (i)
k∈ΩP
(i)
where ΩP represents the ith PDM-PAJO optimization thread, which contains some
lamps and users. P∗ (i) is the optimal power allocation matrix for the lamps in the
ΩP
ith thread using PDM-PAJO.

For the example shown in Fig. 3.8, all the users and lamps can be divided into two
optimization threads using PDM-PAJO. Given the locations of the users, the two opti-
(1)
mization threads can be represented as ΩP = {lamp 1, lamp 2, user A, user B, user C},
(2)
ΩP = {lamp 3, lamp 4, user D, user E}. Thus, lamps 1 and 2 can work together to
support user B by optimizing the transmitted power. When the algorithm calculates
the SINR for user A, the MAI is assumed to be caused by the messages from both
lamps 1 and 2 to user B. Although users C and A are in the same optimization thread,
the algorithm ignores user C when calculating the MAI for user A, since they are not
in the same access area.
In general, DEPA, PADJO and WDM-PAJO require no coordination between
lamps. PDM-PAJO requires some coordination, and CM-PAJO requires the most,
depending on the physical location of the users.
3.3.1 Performance Comparison
We compare the performance of the proposed CM-PAJO and our four decentral-
ized algorithms using the multi-detector model. We test a large indoor environment
described in Table 4.2 to compare the CM-PAJO, PDM-PAJO, WDM-PAJO, PADJO
and DEPA. In this chapter, we consider the minimum access area case2 for all the
2
The minimum access area means the minimum value of the access area for which the entire
indoor floor surface is covered. The access area of all lamps is assumed equal.
37
12.5 m
1.7
7 m 2.5 m
2.5 m
12.5 m
Figure 3.9: Top-down view of the positions of lamps and users in a large indoor
environment. The small circles represent the lamps and the squares represent the
users.
64
Required peak radiation power to noise ratio (dB)
62
60
400 lux illumination
58
56
54
52
CM−PAJO, 1−detector
50 PDM−PAJO, 1−detector
WDM−PAJO, 1−detector
48 CM−PAJO, 7−detector
PDM−PAJO, 7−detector
46 WDM−PAJO, 7−detector
PADJO, 1−detector
44 DEPA, 1−detector
5 10 15 20 25 30 35 40 45 50
Number of users
Figure 3.10: Peak radiation power to noise ratio (PPNR) required for a BER of 10−3
in the large indoor environment using the minimum access area needed cover the
room with 25-length OOC codes.
38
algorithms. [27] discusses the effect of the size of the access area on some of these
algorithms. The geometric position of the lamps and users are shown in Fig. 3.9.
From the results in Fig. 3.10, we see that the optimized algorithms do much better
than DEPA in general, showing the advantage of resource optimization. CM-PAJO
is the optimal power allocation algorithm that can spend about 10 dB less trans-
mitted power than DEPA to achieve the same BER performance. For decentralized
algorithms, PDM-PAJO and WDM-PAJO only need 2 dB more power than the CM-
PAJO to get the same BER. In addition, using 7 PDs can save as much as 2 dB
transmitted power for both centralized and decentralized algorithms over single PD
cases. If there is no background light, when the proposed VLC system satisfies stan-
dard 400 lx illumination,3 it can support up to 40 users when using the 7-detector
CM-PAJO algorithm.
The BER of the proposed PDM-PAJO technique for different access area radii
in the small indoor environment is shown in Fig. 3.11. We also show the BER per-
formance of the CM-PAJO algorithm for comparison. From the simulation results,
when the radius increases from 1.77 m to 2.26 m, the BER of PDM-PAJO converges
to the CM-PAJO. When R = 1.77 m (the minimum radius of the access area), the
PDM-PAJO has less than a 2 dB power penalty compared with CM-PAJO.
Fig. 3.12 shows the BER of the WDM-PAJO technique with different access area
radii, also in a small indoor environment, compared with PDM-PAJO and CM-PAJO.
When the access areas have a small radius, the BER performance of WDM-PAJO
and PD-PAJO are almost the same, both yielding less than a 2 dB power penalty
compared with CM-PAJO. However, as the radius of the access areas increases, the
BER curves diverge. The weight τk for all k in (3.16) approaches K, and when all
the weight are the same the WDM-PAJO algorithm ceases to work well.
3
400 lx is a standard illumination level for office spaces [49]. The conversion between illuminance
and power can be found in [50].
39
−1
10
−2
BER 10
−3
10
CM−PAJO
PDM−PAJO, R=2.26m
PDM−PAJO, R=2.15m
−4 PDM−PAJO, R=1.91m
10
PDM−PAJO, R=1.77m
46 48 50 52 54 56
Figure 3.11: Average BER of 4 users for CM-PAJO and PDM-PAJO for different
radii of access area in the small indoor environment with single detector.
0
10
−1
10
−2
10
BER
−3
10
CM−PAJO
PDM−PAJO, R=1.77m
WDM−PAJO, R=1.77m
WDM−PAJO, R=1.91m
−4 WDM−PAJO, R=2.15m
10
WDM−PAJO, R=2.26m
46 48 50 52 54 56
Figure 3.12: Average BER of 4 users for CM-PAJO and WDM-PAJO for different
radii of access area in the small indoor environment with single detector.
40
Figure 3.13: Histogram of BER performance for 4 randomly distributed users, with
peak radiation power to noise ratio of 61 dB
To analyze the performance of the proposed PDM-PAJO and WDM-PAJO from

a statistical point of view, we simulate 40 trials of 4 users randomly distributed in
the small room. From the simulation results shown in Fig. 3.13, more than 75% of
the 40 trials’ BER for PDM-PAJO and WDM-PAJO are lower than 10−4 . Since some
of the lamps in the PDM-PAJO algorithm exchange feedback information from the
users, while in WDM-PAJO there is no information exchange, PDM-PAJO is better
able to allocate power to users and often results in a lower BER.
3.4 Computational Burden Comparison
The goal for seeking a decentralized power allocation algorithm is to reduce the
computational burden of the centralized algorithm, CM-PAJO, especially for a large
indoor environment. To estimate the computational burden, we use the maximum
number of variables per thread (optimization problem size) as the metric. The vari-
ables to be calculated per thread are the power allocated from the LEDs to the users
and the time-space MMSE filter coefficients, which are represented as pqk and wk [v, `]
41
(defined in (3.1) and (3.4), respectively). The number of variables in each thread is
the size of the optimization problem, which implies the computational burden.
Since in CM-PAJO all lamps need to share the channel feedback information from
all the users and work together to solve for the power allocation, there is only one
optimization thread. Therefore, the optimization problem size for CM-PAJO can be
derived as
ΛCM = (NQ + V L)K. (3.18)
Since the proposed decentralized algorithms use parallel processing, the compu-
tational burden per thread for them is much lower than for CM-PAJO. The actual
optimization problem size depends on the users’ positions in the indoor environment.
In this chapter, we consider the users to be uniformly distributed in the indoor envi-
ronment. The optimization problem size of the decentralized algorithms also depends
on the access area. We consider the minimum access area case for calculating the
computational burden.
For DEPA, the transmitted power for each user is the same. Thus, the optimiza-
tion problem size is smaller than the other decentralized algorithms because there is
no need to calculate the transmitted power for each user, only the filter coefficients
at the detectors. The optimization problem size of DEPA can be calculated as
V LK
ΛDEP A = . (3.19)
N
Since for both PADJO and WDM-PAJO, the lamps all work independently, and
there is no channel information shared among the lamps, the optimization problem
42
5
10
4
10
Computational Burden
3
10
2
10
CM−PAJO
1 PDM−PAJO
10
WDM−PAJO
PADJO
0
DEPA
10
0 10 20 30 40 50
Number of users
Figure 3.14: Computational burden with the minimum access area in the large indoor
environment, 7 PDs per user, and length of OCDMA code of 7.
size of PADJO and WDM-PAJO can be assumed to be the same. Thus
ΛP ADJO = ΛW DM
(3.20)
(NQ + V L)K 1
= = ΛCM .
N N
The optimization problem size per thread for PDM-PAJO can be written as
KNQ
ΛP DM = + V LK
N (3.21)
(N − 1)V LK
= ΛW DM +
N
Numerical results on optimization problem size are shown in Fig. 3.14. With the
help of parallel processing, the four decentralized algorithms have much lower com-
putational burden than the CM-PAJO algorithm. As the number of users increases,
the advantage of using a decentralized algorithm becomes more obvious.
43
Table 3.2: Time consumption comparison of centralized and decentralized algorithms

in the large room with 50 users.
Running Time/Thread, s
CM-PAJO 2.62 × 104
PDM-PAJO 1.93 × 102
WDM-PAJO 8.65 × 100
PADJO 1.02 × 101
DEPA 1.65 × 10−2
We compare the running time per thread for the centralized and decentralized
optimization algorithms. The optimization is performed using the SQP solver in
MATLAB running on a PC with an Intel i5 processor and a 2G memory. The 50
uniformly distributed users case is tested. The results, which are the average of 5
trials, are shown in Table 3.2. We find that the decentralized algorithms need much
less time than the centralized algorithm. Since there is no optimization for DEPA,
the time consumed for DEPA is smallest. PADJO and WDM-PAJO need a similar
running time that is about 0.04% of the centralized algorithm. PDM-PAJO takes
about 25 more time than WDM-PAJO and PADJO since the optimization threads
for PDM-PAJO usually contain more lamps and users; for WDM-PAJO and PADJO,
each thread only contains a single LED. Taking into consideration the computational
burden and BER performance of the centralized and decentralized algorithms we
propose, PDM-PAJO and WDM-PAJO both provide a reasonable trade-off between
BER performance and computational burden.
3.5 Practical Considerations
In this section, several practical considerations of our proposed VLC design such as
shadowing effects, illumination requirements, dimming control, beamwidth selection,
44
0
10
Normalized data rate −1

10
Adaptive CM−PAJO
CM−PAJO without shadowing info.
−2 Equal power allocation
10
User is only served by the closest lamp
No shadowing
0 1 2 3 4 5 6
Shadowing effects (dB)
Figure 3.15: Normalized data rate of the user that is blocked under different shadow-
ing conditions for a BER = 10−3 . 4 users are in the small indoor environment, and a
single detector is used with length-7 OOC codes.
and nonlinear effects of LEDs are discussed.
3.5.1 Shadowing Effects
Shadowing is a common phenomenon that can be regarded as a kind of path loss

as in RF communication systems [51]. In our work, we assume that the shadowing
effects in VLC systems are caused by objects that block the light. Since the light can
be partially blocked, we define the shadowing effects as a power loss. We assume the
shadowing losses are generated from the one lamp that is closest to the user. The
shadowing loss coefficient for user k is denoted as k ∈ [0, 1]. When k = 0 the light
is totally blocked, and when k = 1 there is no shadowing for user k. In this work,
we represent the power loss due to k in dB.
To test the effects of shadowing on our system, we assume a 4 users system
45
in the small indoor environment, and only one of them is affected by shadowing.
Fig. 3.15 shows the maximum data rate of the affected user normalized to that of
the non-shadowed case. In this chapter, we design our algorithms to be adaptive,
so the system reallocates the transmitted power when the environment and users’
positions change. Fig. 3.15 compares the adaptive CM-PAJO, the CM-PAJO without
shadowing information, the DEPA, and the case that each user is only served by the
closest lamp. From the numerical results, although the data rate of all schemes
decreases with increasing shadowing effects, the adaptive CM-PAJO has significantly
better performance.
For the decentralized algorithms, if the shadowed users are supported by more
than one lamp, the decentralized power allocation algorithms can also adjust the
power assignment to provide those users good communication service. However, if
a user is only served by one lamp, the decentralized algorithms cannot alleviate the
shadowing effect. We can usually adjust the size of the access area to make sure each
user can be served by more than one lamps using the decentralized algorithms.
3.5.2 Illumination Requirements and Dimming Control
Dimming can be used to satisfy different illumination requirements for different

purposes. The effective dimming level depends on the radiation power and the ratio
of the OCDMA code weight to the code length, η, which determines the illumination
potential. In this work, we assume the OCDMA codewords have been specified (not
adaptive), and η is fixed. Thus, the dimming level can only be adjusted by changing
the radiation power. The Illumination Engineering Society of North America provides
some illumination level standards for indoor environments [49]. For example, the
illumination level for an office building should be greater than 400 lx. For hotels and
restaurants, 100 lx illumination is enough.
46
To ensure the room is uniformly illuminated in space, we assume that there are
Kv virtual users uniformly distributed in the room, and the virtual users need illumi-
nation only (no communications). Thus, the total number of users is Ktot = K + Kv ,
where K is the number of real users who need both data and illumination. Under
this assumption, we can define the illumination tolerance at user k as ∆k , and require
that
|Ar η ĥTk pmax
dim + Pb − Preq | ≤ ∆k , (3.22)
where ĥk = (ĥ1k1 , ĥ2k1 , . . . , ĥNQ k1 )T . We denote ĥqk1 as the channel gain from LED q
to the detector of user k that is pointed towards the ceiling. Pmax
dim is the dimmed peak
power vector, which can be represented as Pmax

dim = ∅P
max
= P max (∅(1) , ∅(2) , . . . , ∅(NQ ) )T ,
where ∅ = (∅(1) , ∅(2) , . . . , ∅(NQ ) ), and ∅(q) is the dimming parameter for LED q. To
satisfy specific illumination requirements, the dimming parameters can be adjusted in
the range of [0, 1] for dimming control. Thus, the peak power constraint for different
LEDs may be different. Pb and Preq represent the received power from background
light and the required illumination, respectively. The tolerance ∆k limits the differ-
ence between the required illumination and the actual illumination.
To make sure the illumination throughout the room is as spatially constant as
possible, the dimmed transmitted power of each LED can be controlled to minimize
the illumination tolerance among all the users (real and virtual). Thus, the optimal
dimming parameters ∅∗ can be found by
[∅∗ ] = arg min max ∆k . (3.23)

∅ k
Then, the dimmed peak power vector pmax

dim can be used as a peak power constraint
for each LED, in either the centralized or one of the decentralized power allocation
algorithms described above.
47
40
Required illumination, 300 lx
35
Average illumination tolerance (%)

30
25
20
15
10
0
0 10 20 30 40 50
Number of virtual users
Figure 3.16: Average illumination tolerance for different number of virtual users.
The illumination tolerance can be used as a criterion to evaluate how uniformly

an illumination can be provided by the VLC system. Numerical results on the aver-
age illumination tolerance of all the indoor area with different numbers of uniformly
distributed virtual users are shown in Fig. 3.16. As expected, the more virtual users,
the lower the average illumination tolerance that can be achieved, since more virtual
users can represent the space in the room more fully. However, more virtual users
can introduce more computational burden when we calculate the illumination toler-
ance. For this result, we conclude that 16 uniformly distributed virtual users can
fully represent the entire space in the small indoor environment.
The illumination tolerance ∆k affects the BER performance in multiuser indoor
VLC systems, which is assumed to be in the range of 9% to 60%. From the sim-
ulation results, we observe that if the tolerance of illumination increases, the BER
performance curve converges to the no-illumination-constraint case. Simulations are
48
−1
10
−2
10
−3
10
BER
No constraints
∆k=60%
∆k=30%
−4
10 ∆k=20%
∆k=15%
∆k=13%
−5
10 ∆k=12%
∆k=9%
48 49 50 51 52
Total radiation power to noise power ratio (dB)
Figure 3.17: BER comparison with different lighting tolerances, with 4 users and 16
virtual users for the 25-LED lamp case, 400 lx illumination requirements
shown for 4 users with data and illumination requirements and 16 virtual users with
only illumination requirement in Fig. 3.17. From these results, the BER with 60%
tolerance is quite close to the BER without constraints. Note that this variation in
the room lighting may be unpleasant for a human eye. The evaluation of this aspect
of the design is beyond the scope of this dissertation.
Fig. 3.18-(a) shows a contour plot of the illumination distribution for 4 users with
both data transmission and illumination requirements, plus 16 virtual users with illu-
mination requirements only. Fig. 3.18-(b) shows the illumination distribution without
data transmission requirements. Comparing these two figures, the illumination distri-
bution in (a) is still smooth and flat. That is to say, setting illumination constraints
prevents the lighting system from creating too dark and too bright spots in the room,
and the illumination requirements at all the user locations are satisfied.
49
45 45
400lx
40 40
35 35
width of the room

width of the room
350lx
30 30
25 25
300lx
20 20
15 15
250lx
10 10
5 5
200lx
10 20 30 40 10 20 30 40
length of the room length of the room
(a) (b)
Figure 3.18: Illumination distribution comparison of (a) data transmission case and
(b) no data transmission case. The red dots identify the real users, and the blue dots
represent the virtual users, with 10% tolerance.
The semiangle of the LEDs is another factor that affects the dimming control
accuracy. In Fig. 3.19, we compare the optimal illumination tolerances for different
semiangles using our multiple-LED lamp and a single-LED lamp in which there is
only one LED per lamp. For small semiangle LEDs (less than 15 degrees) in the
multiple-LED case, the beam width of the LEDs is too narrow, and all the area on
the floor cannot be illuminated. Thus, some areas of the floor would be very dark,
and other areas would be bright. Because of that, the illumination tolerance defined
in (3.22) is large. If large semiangle LEDs are used, the illumination area of each
LED is relatively large, but the intensity of the illumination would not be as high
as in the small semiangle cases. It is not easy to control the illumination level for
a particular area as accurately with large semiangle LEDs. Therefore, to make sure
the illumination distribution is uniform for different requirements, the semiangle of
the LEDs cannot be too large or too small. From the numerical results in Fig. 3.19,
a 20-degree semiangle LED is the best choice for the proposed multiple-LED lamp
model to have the lowest illumination tolerance if 16 uniformly distributed virtual
users are modeled in the small room. The single-LED lamp has a similar behavior as
50
35
400 lx
350 lx
30 300 lx
Minimal illumination tolerance (%) 25 Single−LED lamp
20
15
10
Multiple−LED lamp
5
0 20 40 60 80 100
Semiangle (degree)
Figure 3.19: Minimum illumination tolerance under different illumination require-

ments for different LED’s semiangle in the small indoor environment; 16 virtual users.
the multiple-LED lamp case. There is an optimal choice for the semiangle, which is
around 60 degrees for the single-LED lamp. Compared with the multiple-LED lamp,
the single-LED lamp cannot provide high accuracy illumination control.
We also take the background illumination (BI) in the indoor environment into ac-
count in the form of background power Pb in (3.22). We assume that the background
power also introduces shot noise. If the required illumination level in the room is
assumed to be fixed around 400 lx [49], the more background light there is, the less
radiation power the LED lamps need to emit. Fig. 3.20 shows the BER performance
of the CM-PAJO algorithm under different background light conditions. In this re-
sult, we assume the background light is uniformly distributed, and the background
illumination on different PDs is the same. We note that increasing the background
light decreases the number of users the system can support.
51
−2
10
−3
10
−4
10
BER
−5
10
−6
10
10
−7 BI=0 lx
BI=80 lx
BI=100 lx
−8
10
7 8 9 10 11 12 13
Number of users
Figure 3.20: BER performance for CM-PAJO under different background light illumi-
nation conditions in the small indoor environment with 400 lx required illumination
and 7-detector model, 4 users with length-7 OOC codes.
If the background light comes from a window or another room, our multiple PDs
system has advantages over the single PD case. Since we take advantage of the signals
from different PDs, the space-time MMSE filter can improve the SINR. We now model
the background light from a window as a point light source on a wall. We suppose the
window is located on one wall of the small room at (0, 2.8, 2.8). Numerical results
for this case are shown in Table 3.3. In this case, there are four randomly distributed
users, and the background light adds shot noise. The results indicate that our multi-
detector system is robust against background light interference from a window by
using our MIMO technique.
52
Table 3.3: BER performance for CM-PAJO with 400 lx required illumination.
BER×10−5 BI=0 lx BI=80 lx BI=100 lx

Number of PDs, V = 1 11.3 57.3 132
Number of PDs, V = 4 2.19 2.39 2.57
Number of PDs, V = 7 1.54 1.67 1.81
3.5.3 Transmitted Power Quantization
Although we assume on-off CDMA coding and OOK modulation, since each LED
transmits the sum of signals meant for the various users, the signal itself is no longer
on-off pulsed. In this section, we assume each LED of the multiple-LED lamp is a
LED-array that is composed of many micro-LEDs (µLED) [52].
The optical power from the LEDs is driven by an input electrical signal that carries
information. Due to the structure of the LEDs and the principles of generating light,
the relation between the output optical power and the input current can be modeled
as a nonlinear function. To diminish the effect of the nonlinearity of LEDs on our
system, each µLED in the LED-arrays can only be controlled as on or off, and these
µLEDs can be clustered into different groups, where each group can be controlled to
be on or off. For example, if the µLEDs in an LED-array can be divided into 7 groups
with the same number of µLEDs, there are 8 levels of intensity that can be emitted,
from level 0 to level 7. For level 0, no group is lit; for level 7, all the groups are
switched on. A design trade-off needs to be found between the quantization errors
and the structural complexity, which is outside the scope of this study. Fig. 3.21
shows a possible LED grouping scheme for 8 quantization levels. In this figure, the
LEDs in the LED-array are divided into 7 groups with the same number of LEDs.
Thus, there are 8 levels of intensity that can be emitted, from level 0 to level 7. For
level 0, no group is; for level 7, all the groups are switched on. A design trade-off
53
ⅠⅡⅢ ⅣⅤⅥ Ⅶ Ⅷ ⅨⅩ Ⅺ Ⅻ XIII XⅣ
A
B
Group C
1
Group D
2 E
Group
3 F
Group G
4
H
Group
5 I
Group J
6
K
Group
7 L
M
N
Figure 3.21: LED grouping scheme for 8 quantization levels
needs to be found between the quantization errors and the structural complexity,
which is outside the scope of this study.
Numerical results for the system performance of different quantization levels are
shown in Fig. 3.22. For the same scenario as shown in Fig. 3.1, Case 2, we conclude
that 8 quantization levels are sufficient in our system.
3.6 Summary
In this chapter, we present a multiuser MIMO indoor visible light communication

system that is robust against shadowing, dimming, background radiation, and LED
nonlinearity. In this system, a centralized power allocation scheme and four decen-
tralized algorithms are proposed. To enhance the SINR for each user, a multiple PDs
model is employed at the receiver. The BER performance and computational burden
of the algorithms are analyzed. Compared to the centralized power allocation algo-
rithms, the four proposed decentralized power allocation algorithms all have much
54
−1
10
5 users
4 users
3 users
−2 2 users
10
2 users, unquantized
−3
BER 10
−4
10
−5
10
−6
10
2 3 4 5 6 7 8
Quantization level
Figure 3.22: The average BER performance for different quantization levels with 2,
3, 4 and 5 users using length-7 OOC codes in the small environment, semiangle is 30
degrees, no dimming control.
lower computational burden. Considering the BER performance of the centralized

and all decentralized algorithms, PDM-PAJO and WDM-PAJO are the best choices.
When some users are affected by shadowing, our proposed adaptive MIMO power
allocation algorithms can reallocate the transmitted power to reduce the shadowing
effects. From the simulation results, the data rate of the shadowed user using adaptive
CM-PAJO is about twice as high as the algorithm without knowing the shadowing
information when the shadowing loss coefficient αk is 3 dB. The algorithms proposed
in this chapter can adjust the dimming parameters for each LED to accommodate
the illumination requirements. From the numerical results, our proposed MIMO algo-
rithm can support multiple users with high communication performance in both small
and large indoor environments within strict illumination requirements. In addition,
the nonlinearity of LEDs is also considered in this chapter and can be solved by using
55
micro-LED arrays.
Chapter 4
Modulation Schemes for VLC

Systems
In this chapter we first propose a robust and high-rate multiuser system design
based on M-PAM. Since light emitted from LEDs is non-coherent, an M-ary intensity
modulation that has a high bandwidth efficiency such as M-PAM is a good choice [17].
Then, a joint optimization of waveform and MMSE filter is proposed to reduce ISI and
MAI simultaneously. In the end, a comparison between DCO-OFDM and M-PAM
with designed waveform is given.
The proposed algorithm can adjust the modulation constellation size for each user
to maximize the bit rate under different channel environments such as shadowing, light
dimming, and the impact of multiple access interference. In our MISO approach,
multiple LED lamps coordinate to provide users with maximum data rates. We
compare OCDMA using our adaptive M-PAM with TDMA. The OCDMA technique
can offer a higher bit rate when the number of users is larger than the length of the
OCDMA code.
To increase the transmission throughput, ISI is one of the biggest challenges. We
56
57
propose a joint optimal waveform design for visible light communication system using
M-ary pulse amplitude modulation to support multiple users. The transmitted wave-
forms and minimum mean squared error filters are jointly optimized to minimize the
intersymbol and multiple access interferences. Based on different channel conditions,
the designed waveforms and modulation constellation size can be adaptively adjusted
to guarantee the highest data rate.
A comparison between our optimized M-PAM and DCO-OFDM for LED-based
communication systems is given. Considering the bandwidth limit and constrained
peak transmitted power characteristics of LEDs, bit loading with an optimized mod-
ulation index is used for the DCO-OFDM.
Part of the work presented in this chapter has been published in [30, 53].
4.1 Adaptive M-PAM for Multiuser MISO Indoor
VLC Systems
4.1.1 Background
To support multiple users, MISO processing and OCDMA can be applied [45, 54–
56]. Multiple LED lamps transmit CDMA coded signals in a coordinated manner
to support multiple users, making the system robust against channel shadowing.
In addition, to diminish the MAI and improve the SINR, the transmitted power
from each LED can be optimally allocated to users and optimally detected using a
MMSE filter at the receivers, as presented in Section 3.2. In this section we adopt
an adaptive M-PAM modulation scheme instead of the OOK previously used. The
adaptive M-PAM modulation algorithm selects a different constellation size for each
user to optimize the transmitted data rate in a fair manner. Users with better channel
58
downlink quality can benefit from a larger modulation constellation size and/or be
allocated a lower portion of the total LED power so that all users can maintain a
preset communication performance level. We show that CDMA is able to provide
higher data rates than TDMA for the same performance when the number of users
is larger than the code length.
Recently, some significant research has been directed towards designing modula-
tion schemes for VLC systems [17]. M-PAM was explored in [57] to yield a (log2 M)-
fold increase in the data rate compared with OOK. Instead, OFDM can be used to
increase the data rate and efficiently combat ISI [24, 58]. Furthermore, researchers
have proposed adaptive modulation schemes for VLC based on OFDM [59]. The
drawback of OFDM is that it has a relative high PAPR, making it more sensitive
to the nonlinear distortion of the LEDs than pulsed techniques such as PAM. An
M-ary variable period modulation (MVPM) scheme for VLC was proposed in [19];
MVPM has been proven capable of reducing the slot duration to increase the data
transfer rate in VLC system. However, it is difficult to keep all the users synchro-
nized. In addition, the narrow time slot may introduce ISI from multipath in the
indoor channel. A MIMO-PPM technology was proposed in [60] to improve the data
rates without reducing the reliability of the link. However, the multiuser case was not
considered in [60]. Furthermore, PPM is bandwidth inefficient and very sensitive to
external interference that may cause a complete data corruption. To alleviate these
drawbacks, we propose a MISO CDMA VLC system using an adaptive M-PAM mod-
ulation scheme with synchronized symbol rate across users and LED lamps. Channel
state information at the transmitter is assumed known perfectly so that when the
downlink channel conditions change due to motion or shadowing, the proposed algo-
rithm can adjust the modulation constellation size to optimize the bit rate adaptively.
59
4.1.2 Adaptive M-PAM
We assume all LED lamps are synchronized with each other and all contribute to
the data transmission for all users in the access area of interest. The VLC channel
between LED q and user k is completely characterized by ĥqk and known at the
transmitters. Using M-PAM modulation, we assume the amplitude of the transmitted
symbol for user k is sk ∈ {0, Mk1−1 , Mk2−1 ..., 1}, and each symbol carries log2 Mk bits,
where Mk is the modulation constellation size for user k. Since we assume the binary
data is equally likely, the ak are uniformly distributed. Thus, the transmitted signal
for the qth LED can be represented as
K
X
xq (t) = pqk sk ck (t), (4.1)
k=1
where pqk is the power allocated to the qth LED for user k and ck (t) is the OCDMA
codeword for user k. Similar to the work in Section 3.2, the received signal for user
k after MMSE filtering can be represented as
yk = sT Bk Cwk + nTk wk , (4.2)
where s = (s1 , s2 , . . . , sK )T is the transmitted symbol vector; the MMSE filter for user
k is defined as wk ; C represents the CDMA code matrix, which can be represented
as C = (c1 , c2 , · · · , ck )T , where ck is the CDMA code for user k; nk is the noise at
the user k, which can be modeled as Gaussian distributed noise with variance σ 2 . To
60

facilitate the formulation, we define the matrix Bk = diag ĥTk · P , where
 
 p11 p12 · · · p1K 
 
 p21 p22 · · · p2K 
 
P= . (4.3)
 
 .. .. .. .. 
 . . . 

 
 
pQ1 pQ2 · · · pQK
represents the power allocation matrix. After some calculations, the MMSE filter for
user k can be represented as
−1 T
wk = CT Bk Σs Bk C + σ 2 I C Bk qk , (4.4)
where I is the identity matrix of the same size as the OCDMA code matrix C, and
Σs is the correlation matrix for the transmitted symbol, which can be calculated as
 
 E{s1 · s1 } E{s1 · s2 } · · · E{s1 · sK } 
 
 E{s2 · s1 } E{s2 · s2 } · · · E{s2 · sK } 
 
Σs =  , (4.5)
 
.. .. .. ..

 . . . . 

 
 
E{sK · s1 } E{sK · s2 } · · · E{sK · sK }
and
qk = (E{sk · s1 }, E{sk · s2 }, · · · , E{sk · sK })T , (4.6)
where E{sk · sv } can be calculated as

2
 2Mk −Mk


6(Mk −1)
k=v
E{sk · sv } = . (4.7)

1

k 6= v
4
61
The SINR for user k can be represented as [27]1
A2r wk T CT Bk Zk Σs Zk BTk Cwk

γk = , (4.8)
wkT CT Bk Ẑk Σs Ẑk BTk Cwk + σ 2 wkT wk
where the matrix Zk is defined as a matrix with a ‘1’ in its (k, k)th element and zeros
in all other places, and Ẑk = I − Zk .
From (4.5)-(4.8), we conclude that the SINR for user k depends on the power
allocation scheme and the M-ary modulation constellation size of all users. Therefore,
the SINR is a function of M = (M1 , M2 , · · · , MK )T and the power allocation matrix
P.
The bit error rate (BER) for user k when using M-PAM can be represented ap-
proximately as [34]
r r
Mk − 1 γk 1 γk
BERk ≈ erfc ≈ erfc . (4.9)
Mk log2 Mk (Mk − 1)2 log2 Mk (Mk )2
In this chapter, our adaptive M-PAM scheme can adjust the modulation constel-
lation size for different users, i.e., choose the optimal constellation size to optimize
the throughput for all users. The bit rate for user k can be represented as
(k)
Rb = Rs · log2 Mk , (4.10)
where Rs is the symbol rate, assumed to be the same for all users. To optimize the
throughput fairly, the optimization cost function we use is given by
(k)
[P∗ , M∗ ] = arg max min Rb , (4.11)
P,M k
1
For all SINR expressions in this chapter, the responsivity of the receiver is ignored. It is nonethe-
less accounted for in the simulation results.
62
BER Channel H
M-PAM Feedback Power Feedback
Controller Allocation
s1 ...
M1
s2
Sum
M2
s1 (t )
...
...
...
... sK
MK
LED 1
Power
c1 Allocation
c2 ...
cK
Sum
s2 (t )
...
LED 2
...
Power
Allocation
...
Sum
sN (t )
...
LED N
Figure 4.1: Block diagram of the proposed adaptive M-PAM algorithm using our
MISO technique
where P∗ and M∗ are the optimal solutions for power allocation and modulation con-
stellation size, respectively. When doing the optimization, a peak transmitted power
constraint must be considered. To satisfy the communication quality, a constraint on
the BER should also be taken into account. Thus, the optimization constraints can
be represented as
K
X
pqk ≤ P max and pqk ≥ 0 ∀q, and BERk ≤ B max , (4.12)
k=1
where P max is the peak transmitted power. B max is the desired BER for each user,
which guarantees the communication quality. Usually B max is chosen as 10−3 , since
forward error correction (FEC) can then be applied to lower it to within application-
specific standard values [61].
A block diagram of the proposed adaptive M-PAM algorithm is shown in Fig. 4.10.
63
The power allocation P and the M-ary modulation constellation size vector M is
computed jointly. Depending on the channel state information and BER fed back
from the users, the central controller optimizes these to maximize the bit rate for each
users. The binary data d = (d1 , d2 , . . . , dK ) are M-ary modulated , then modulated
with OCDMA codes (c1 , c2 , . . . , cK ), and finally transmitted by the LEDs. If the
channel or BER feedback information is changed, the controller adaptively adjusts
the constellation size to maintain the desired performance.
4.1.3 OCDMA vs TDMA using M-PAM
Since the OCDMA codes are not perfectly orthogonal, it is not evident a priori
whether OCDMA is a more efficient method to support multiple users in indoor VLC
systems than an orthogonal multiple-access scheme, such as TDMA [62]. In this sec-
tion, we analyze the throughput achievable with our optimized adaptive M-PAM al-
gorithm using OCDMA vs. TDMA. We compare the SINR, modulation constellation
size and the bit rate achievable using OCDMA and TDMA. To keep the comparison
fair, we assume the OCDMA and TDMA options use the same bandwidth, i.e., the
pulse width, Tc , for both OCDMA and TDMA is the same, as shown in Fig. 4.2.
For TDMA, if the number of users increases, the symbol rate for each user decreases,
since each time slot can only be used by one user at a time. For OCDMA, the symbol
rate for each user only depends on the length of the codeword L. Thus, we can write
the symbol rate R
bs and R
es using TDMA and OCDMA, respectively, as
1 es = 1 .
R
bs = , R (4.13)
K · Tc L · Tc
64
TDMA User1 User2 ... UserK User1 User2 ...

User1
OCDMA User2
Tc
... UserK
Figure 4.2: OCDMA and TDMA comparison using M-PAM, K < L case (L = 7 and
K = 5).
Therefore, the bit rate for user k using TDMA and OCDMA can be expressed as
(k) log2 M̂k (k) log2 M̃k

R̂b = , R̃b = , (4.14)
K · Tc L · Tc
where M̂k and M̃k are the modulation constellation sizes for user k using TDMA and
OCDMA, respectively.
To compare the SINR for TDMA and OCDMA, in the following analysis we
assume the average transmitted power is P̄ , and the channel gain ĥ from the lamp to
all users is the same. Then, we can roughly represent the SINR for each user using
TDMA as
A2r ĥ2 P̄ 2
γ̂ = . (4.15)
σ2
65
Similarly, the SINR for each user using OCDMA can be roughly represented as
A2r ĥ2 P̄ 2 ω
γ̃ = , (4.16)
(K − 1)ĥ2 P̄ 2 λ + σ 2 K 2 L
where λ is the upper-bound on the cross-correlation value for OCDMA codes used,
and ω is the code weight. Note that this expression is a worst case since the MMSE
filter would remove much of the MAI.
From (4.15) and (4.16), we obtain
γ̃ σ2
= . (4.17)
γ̂ (K−1)ĥ2 P̄ 2 λ 2
+ σ 2 Kω L
ω
Since K ≥ 1 and ω < L, we conclude γ̃ < γ̂. In other words, the modulation
constellation size for TDMA is greater than or equal to that of OCDMA.
From (4.14), we see that the bit rate is related to the number of users K and the
length of the code L for TDMA and OCDMA, respectively. Comparing the bit rate,
we get
R̃b K log2 M̃k
= · , (4.18)
R̂b L log2 M̂k
and thus
OCDMA log2 M̂k
K ≷ ξ · L where ξ = . (4.19)
TDMA log2 M̃k
Since for both M̃k and M̂k can be chosen from small values such as {2, 4, 8, 16}, we can
safely assume that ξ ≈ 1. Therefore, we can conclude that, when the number of users
is larger than the length of OCDMA code, the OCDMA technique can offer a higher
bit rate than using TDMA. The highest data rate is achieved when the minimum
length code needed to support the number of users is chosen.
66
Table 4.1: Parameters Used for Large Indoor Environment
Size of room 12.5 m × 12.5 m × 3 m

Semiangle of LEDs 60o
Area of PD, Ar 1 cm2
Peak optical power per lamp 300 mW
Noise variance σ 2 = 2 µW
Modulation constellation sizes 2, 4, 8, 16
B max ≤ 10−3
OOC code index L = 25 : {1, 2, 7} {1, 3, 10}
{1, 4, 12} {1, 5, 14}
L = 19: {1, 2, 6} {1, 3, 9}
{1, 4, 11}
L = 13: {1, 2, 5} {1, 3, 8}
4.1.4 Numerical Results
In this section, the performance of the proposed adaptive M-PAM algorithm is

shown using simulation. To test the applicability of the algorithm in different scenar-
ios, we show results for different users cases in a large indoor environment, i.e., an
empty and unfurnished room. Unless otherwise noted, the parameters used to obtain
the numerical results are shown in Table 4.2. We assume the users are randomly
dispersed in the room. The geometric position of the lamps and users (K = 40 case)
is shown in Fig. 4.3.
The shadowing effects are also taken into account in this dissertation, since it is
common for objects such as furniture and pedestrians to partially block the light from
an LED lamp. We model the shadowing effect as an optical power loss from the one
lamp that is closest to the user. Define k ∈ [0, 1] as the shadowing loss coefficient
for user k. When k = 0 the light is totally blocked, and when k = 1 there is no
shadowing effects for user k. In this work, we represent the power loss due to k in dB.
A feedback channel from each user to the LED controller informs the system of the
channel gain experienced so that the algorithm can adjust the M-PAM modulation
67
12.5 m
1.7
7m
2.5 m
2.5 m
12.5 m
Figure 4.3: Top-down view of indoor environment. The small circles represent the
lamps and the squares represent the users.
constellation size adaptively to optimize the bit rate when the channel is experiencing
shadowing.
Simulation results under different shadowing conditions are shown in Fig. 4.4. We
compare the average modulation constellation size for OCDMA and TDMA under
different shadowing loss assuming one quarter of all users are suffering from the
shadowing effect. The average modulation constellation size for TDMA is uniformly
higher than using OCDMA, as expected due to the lower SINR of OCDMA because
of the MAI it experiences.
For higher quality communications, a lower desired BER can be used, inevitably
leading to a smaller modulation constellation size, as evident from (4.9). Simulation
results in Fig. 4.5 show the performance for various values of B max . As expected, the
algorithm must sacrifice data rate to obtain a better BER performance.
Although Fig. 4.6 shows that TDMA has a larger modulation constellation size
than OCDMA, the throughput also depends on the relation between the bit rate and
the symbol rate, given in (4.10). Numerical results showing the average bit rate using
68
16
OOC,L=25,K=30
OOC,L=25,K=40
Average Modulation constellation size

TDMA
15
14
13
12
11
10
0 0.5 1 1.5 2
Shadowing effects (dB)
Figure 4.4: Average modulation constellation size for adaptive M-PAM for 30 and 40
user cases with shadowing effects.
16
OOC,L=25,K=40
15 OOC,L=25,K=30
TDMA
14
13
12
11
10
7 −6 −5 −4
10 10 10
Desired BER
Figure 4.5: Average modulation constellation size for adaptive M-PAM modulation
for 30 user and 40 user cases with different desired BER values, no shadowing effects.
69
OOC,L=13
16
OOC,L=19

OOC,L=25
15 TDMA
14
13
12
11
10
9
10 20 30 40 50 60 70 80
Number of users
Figure 4.6: Average modulation constellation size for different numbers of users with
OCDMA and TDMA techniques, no shadowing effects.
30
OOC,L=13
OOC,L=19
OOC,L=25
25
TDMA
Average bit rate for each user (Mb/s)
OCDMA,OOC optimal selection
20
15
10
0
10 20 30 40 50 60 70 80
Number of users
Figure 4.7: Average bit rate for different numbers of users with OCDMA and TDMA
techniques; no shadowing effects.
70
OCDMA and TDMA are given in Fig. 4.7, and prove that we can obtain a higher
bit rate using OCDMA if we choose the best OCDMA code. In this chapter, we use
OOC codes with length 13, 19 or 25 to support multiple users. These OOC codes
can support up to 26, 57 and 100 users, respectively [48]. The results show that the
average bit rate for OCDMA is higher than TDMA when the number of users is larger
than 15, 20, and 28 when using the length 13, 19 and 25 OOC codes, respectively. We
can get a higher bit rate using OCDMA than TDMA by choosing the right OCDMA
codes. The highest throughput obtainable for this scenario is labeled ‘OCDMA, OOC
optimal selection’ in Fig. 4.7.
4.2 M-PAM Joint Optimal Waveform Design for
Multiuser VLC Systems over ISI Channel
4.2.1 Background
ISI is one of the biggest challenges for high-speed data transmissions. The low rise-
time of lighting LEDs and multi-path propagation are two factors leading to ISI. For
indoor VLC systems, multi-path propagation comes from reflections of the light from
the ceiling, walls, furniture and other reflective surfaces and limits systems operating
at bandwidth above 100 MHz [63]. The 3 dB modulation bandwidth of commercial
lighting LEDs is limited to a few tens of MHz [64]. Thus, the bandlimited LED is
the dominant factor to introduce ISI.
Equalization that can be realized by either hardware or software is an attractive
solution to mitigate ISI and obtain a high data rate. Some researchers have proposed
a pre-equalization circuit to increase the data rate and have achieved up to 340 Mb/s
transmission using OOK with a BER of 2×10−3 [65]. Using red-green-blue LEDs and
71
a hardware equalization circuit, Gb/s data rate can be achieved [64]. As for software
equalizers, the zero forcing (ZF) algorithm is a popular signal processing method to
mitigate the effects of ISI in RF communications [66, 67]. ZF has also been applied
in optical wireless communication [68]. A least mean squared error equalizer using a
training sequence for indoor VLC systems was proposed in [20], and both linear and
decision feedback equalizers were discussed.
A high bandwidth efficiency modulation, such as M-PAM is a good choice to
provide high-speed connections for VLC systems since the light emitted from the
LEDs is non-coherent, and intensity modulation should be used. Using M-PAM, a
(log2 M )-fold increase in the data rate compared to OOK can be achieved. Recently
an adaptive M-PAM scheme was proposed to provide higher data rate for multiuser
VLC systems [21, 53, 69–71]. Instead, OFDM can be used to increase the data rate
while combating ISI [23, 24]. However, optical OFDM systems experience a rela-
tive high peak to average power ratio (PAPR) that can result in a severe nonlinear
distortion of the transmitted signals because of the LED peak transmitted power con-
straint. Researchers have shown that M-PAM with equalization can provide better
performance than OFDM for VLC systems [32, 70].
Another important research topic in indoor VLC is how to choose a multiple access
technique. MAI can be a factor limiting the throughput of multiuser systems. TDMA
is one approach that can be used due to its small operational complexity [72, 73]. A
SDMA technique using angular diversity of the transmitters to support and separate
multiple users in indoor VLC environments was proposed in [62], but the bandlimited
characteristic of the LEDs was not considered. Recently, we proposed an OCDMA
indoor VLC system using a resource allocation algorithm to support multiple users
and reduce the MAI [27].
In this section, we solve both problems, high-speed transmission and multiple
72
access, by using a joint optimal waveform (JOW) design algorithm for MISO multiuser
systems. Recently, we proposed a waveform design algorithm in [30], yet MAI was
not considered. In this chapter, we expand on this idea to design waveforms capable
of reducing both ISI and MAI. In this system, unique waveform designs and MMSE
filters for different users are optimized jointly. In addition, to achieve a high data rate,
an M-PAM modulation scheme is applied. In this work, the proposed JOWs have
two functions: separating users and reducing ISI. Similar to OCDMA, the proposed
JOWs are discrete time sequences that are unique to users. But superior to OCDMA,
JOWs combat ISI and MAI simultaneously and can be adaptively redesigned when the
channel state changes. The JOWs can be optimized for different transmitted symbol
rates by maximizing the SINR. For a fixed required BER, by changing the JOW and
M -ary modulation constellation size, we can find the maximum data rate. The JOWs
can be designed to allow dimming of the light by changing the illumination level,
determined by the average waveform power. However, the illumination level is difficult
to change in OCDMA or TDMA. Channel state information (CSI) uncertainty is taken
into account in this chapter by modeling the channel impulse response as a Gaussian
random process.
To address the real-time computational complexity of performing the requisite
optimization, an off-line waveform design algorithm is then proposed, using a pre-
established waveform table. In practice, the proper waveforms can be selected from
the table based on the real channel gains. The performance of the off-line algorithm
can be estimated by using the channel uncertainty model.
4.2.2 Channel Model
In a typical indoor VLC system, LEDs are used to transmit data and illuminate
the indoor area at the same time. Since the light from the LEDs is non-coherent,
73
LED Lamps
Users
Figure 4.8: A typical indoor VLC system.
IM/DD are employed. Considering the multipath propagation in the indoor channel
and bandwidth limit of the LEDs, the optical link can experience severe ISI.
In this section, we assume there are Q LED lamps and K users with one pho-
todetector (PD) per user receiving signals, and each lamp serves to transmit downlink
signals to all the users. In this work, we consider the bandwidth of the indoor channel
is limited by the LED rise time. [74]. Therefore, the overall channel impulse response
from LED q to user k can be modeled as
q = 1, . . . , Q
hqk (t) = ĥqk hl (t), (4.20)
k = 1, . . . , K
where hl (t) is the impulse response of the LEDs, which can be modeled as a lowpass
filter. We assume the impulse response of all LEDs is the same. ĥqk is the LOS
channel gain from LED q to user k, which can be calculated from (2.6). hqk =
(hqk [1], hqk [2], · · · , hqk [Lh ])T represents the discrete time version of the truncated
channel impulse response from LED q to user k. Fig. 4.9 shows an example of a
truncated channel impulse response that lasts Th seconds. If Rc represents the sam-
pling rate, the length of the discrete time truncated channel impulse response can be
74
1/ Rc
 Nl ... -1 0 1 ... ... N n

u
Lh
t
Th
Figure 4.9: Discrete time version of a truncated LED response.
calculated as
Lh = Th Rc , (4.21)
We further assume that the impulse response of the LEDs can be modeled and
estimated perfectly. However the estimation of the channel gains may be affected
by shadowing, multipath, or noise. In addition, objects moving around users may
interfere with the reflected light, which can introduce uncertainty in the assumed
impulse response. In this chapter, we model the channel impulse response from LED
q to user k as
h∗qk = hqk + ∆hqk , (4.22)
where ∆hqk = (∆hqk [1], ∆hqk [2], · · · , ∆hqk [Lh ])T is the uncertainty in modeling the
channel impulse response. In this chapter, we assume the elements in ∆hqk are
independent of each other, and each element can be modeled as a Gaussian random
variable with zero mean and variance σh2 [75, 76].
75
s1[i ]
M-PAM f11 x1[n]
s2 [i ]
M-PAM f12 ∑ D/A LED 1
...
sK [i ]
M-PAM f1K
f 21 x2 [n]
f 22 ∑ D/A LED 2
...
f2 K
...
fQ1 xQ [n]
fQ 2
... ∑ D/A LED Q
fQK
Figure 4.10: Block diagram of the transmitters for the proposed M-PAM joint optimal
waveform design system.
4.2.3 M-PAM Joint Optimal Waveform Design
Transmitted and Received Signals
A block diagram of the proposed M-PAM joint optimal waveform design system is
shown in Fig. 4.10. The designed transmitted waveform and MMSE filter for different
users are jointly optimized. After M-PAM modulation, the M -ary amplitude symbol
stream for user k at time instant i can be represented as sk [i] ∈ {0, Mk1−1 , Mk2−1 , · · · , 1},
and each symbol carries log2 Mk bits, where Mk is the modulation constellation size
for user k. We assume the sk [i] are uniformly distributed. After the waveform design,
the transmitted sequence from LED q can be represented as
∞
K X
X
xq [m] = sk [i]fqk [m − iLf ], (4.23)
k=1 i=−∞
76
where fqk = (fqk [1], fqk [2], · · · , fqk [Lf ])T is the designed waveform for LED q and user
k. Lf is the number of samples used to represent the waveform. The channel gain
from each LED to each user could be different, and the waveform for each LED and
user is unique.
As shown in Fig. 4.11, after chip matched filtering and sampling, the received
signal for user k can be represented as
Q ∞
X X
rk [m] = xq [m + j]hqk [−j] + nk [m]. (4.24)
q=1 j=−∞
After applying the MMSE filter, the received signal for user k can be written in matrix
form as
Q
X
yk [i] = wkT Hqk xq + wkT nk + bk , (4.25)
q=1
where the MMSE filter for user k is represented by wk = (wk [1], wk [2], · · · , wk [Lw ])T
with length Lw . nk = (nk [1], nk [2], · · · , nk [Lf ])T is the additive Gaussian noise for
user k with zero mean and variance σn2 = N0 Rc , where N0 is the noise spectral density.
bk is a constant needed for the MMSE estimator. Hqk is a Toeplitz matrix, which can
be represented as
T
Lw −1 Lw −1
Hqk = SL(hqk , ),· · · ,hqk ,· · ·, SR (hqk , ) , (4.26)
2 2
where SL (x, m) and SR (x, m) are two functions that operate as m circular shifts on
x, to the left and right, respectively. The vector of transmitted samples that affect
yk [m] is denoted xq = (xq [−Nu ], · · · , xq [0], · · · , xq [Nl ])T . Nl + Nu + 1 = Lh , where Lh
describes the length of successive samples that blur together. Nl and Nu represent
past and future samples that contribute to ISI, respectively as shown in Fig. 4.9.
P
From (4.23), the mth element of the vector xq can be calculated as k sk [bm/Lf c]fqk [
77
y1
PD A/D w1 User 1
y2
PD A/D w2 User 2
...
yK
PD A/D wK User K
Figure 4.11: Block diagram of the receiver.
mod(m, Lf )], where bm/Lf c represents the largest integer less than m/Lf , which is
the number of successive M-PAM data that blurs together, and mod(m, Lf ) is the
remainder of m/Lf .
Waveform Design Algorithm with Imperfect CSI
For the JOW algorithm, the CSI must be known at the transmitters. However,
in practice, we cannot estimate the CSI perfectly. To account for the imperfect CSI,
we substitute the imperfect channel model h∗qk for hqk in (4.26). The received signal
for user k after the MMSE filter with channel uncertainty can be represented as
Q
X
yk [i] = wkT (Hqk + ∆Hqk )xq + wkT nk + bk , (4.27)
q=1
which consists of four parts: the target (intended data) for user k, the uncertainty
caused by the imperfect CSI, the ISI plus MAI, and the noise. Thus, (4.27) can be
78
rewritten as
Q Q
X X
yk [i] = wkT Ĥqk x̂q + bk + wkT ∆Hqk xq
q=1 q=1
| {z } | {z }
Target Uncertainty
(4.28)
Q Q
X X
+ wkT H̃qk x̂q + wkT Hqk x̃q + wkT nk ,
q=1 q=1
| {z }
| {z } Noise
ISI+MAI
where H̃qk = Hqk |hqk [0]→0 , hqk [0] is the peak value of hqk , Ĥqk = Hqk − H̃qk , x̂q =
xq |(fqi =0,i6=k) , and x̃q = xq |fqk =0 .
The mean-squared error, Jk , for user k is defined as
Jk = Es,n,∆h {(yk [i] − sk [i])2 }, (4.29)
where Es,n,∆h {·} represents expectation with respect to the transmitted symbols
(s1 , s2 , · · · , sK ), the noise and the channel uncertainty, which are statistically in-
dependent. Substituting (4.27) into (4.29), we obtain
Q Q Q Q
X X X X
Jk =wkT Hqk Σ(qp) HTpk wk + σh2 wkT Σ(qp) wk
q=1 p=1 q=1 p=1
Q
X 2Mk2 − Mk
− 2wkT Hqk eq + σn2 wkT wk + , (4.30)
q=1
6Mk − 6
Q
X
− bk + 2bk wkT Hqk mq + b2k
q=1
79
where Σ(qp) = Es {xq xTp }. The (m, n)th element of Σ(qp) can be calculated as
 K
X 2M 2 − Mk
 k

 fqk [u]fpk [v]
6Mk − 6


k=1


 , i=j
(qp)
σmn = 1 XX . (4.31)
+ fqk [u]fpz [v]



 4 k6=z z6=k


 1 PK PK

4 k=1 z=1 fqk [u]fpz [v] , i 6= j
where i = bm/Lf c and j = bn/Lf c; u = mod(m, Lf ) and v = mod(n, Lf ). eq =

Es {sk [i] · xq } and mq = Es {xq }.
∂Jk ∂Jk
Solving for ∂bk
= 0 and ∂fk
= 0, the MMSE filter for user k can be obtained as
Q
−1 X
w k = Tk + σn2 I Hqk eq
q=1
Q
(4.32)
1 X
bk = − wkT Hqk mq ,
2 q=1
where
Q Q Q Q
X X X X
Tk = Hqk Σ(qp) HTpk + σh2 wkT Σ(qp) , (4.33)
q=1 p=1 q=1 p=1
and I is the identity matrix.

The signal to interference plus noise ratio (SINR) for user k can be calculated as
Signal
SINRk = , (4.34)
Uncertainty + Interference + Noise
80
where
Q Q Q
!2
X X X 1
Signal = wkT Hqk Σ̂(qp) HTpk wk + wkT Hqk mq −
q=1 p=1 q=1
4
(4.35)
Q Q
XX
Uncertainty = σh2 wkT Σqp wk
q=1 p=1
Noise = σn2 wkT wk (4.36)
Q Q
X X
Interference = wkT H̃qk Σ̂(qp) H̃Tpk wk
q=1 p=1
Q Q
X X
+ wkT Hqk Σ̃(qp) HTpk wk (4.37)
q=1 p=1
Q Q
X X
+ 2wkT H̃qk Σ̄(qp) HTpk wk
q=1 p=1

where Σ̂(qp) = Es x̂q x̂Tp , Σ̃(qp) = Es x̃q x̃Tp , Σ̄(qp) = Es x̂q x̃Tp , all of which can be
calculated similarly as the element of Σ(qp) in (4.31). Substituting (4.32) into (4.34),
we can find that F = (F1 , F2 , · · · , FQ ), M = (M1 , M2 , · · · , MK ) and Rc are the
only variables needed to find the SINRk , where Fq = (fq1 , fq2 , · · · , fqK ). We denote
SINRk = γk (F, M, Rc ). Then, for M-PAM modulation, the BER for user k can be
approximated by [34]
s !
Mk − 1 γk (F, M, Rc )
BERk ≈ erfc , (4.38)
Mk log2 Mk (Mk − 1)2
where erfc(·) is the complementary error function, which is defined as erfc(x) =

R ∞ −u2
√2 e du.
π x
For different data rates, the waveform design algorithm can adaptively adjust the
waveforms for each user to minimize the ISI. For a fixed data rate and modulation
81
constellation size, the optimal waveforms can be obtained by maximizing the mini-
mum SINR of all the users, through which each user can achieve a fair performance.
The optimization cost function is
F∗ = arg max min γk (F, M, Rc ), (4.39)

F k
where F∗ is the optimal value for F. When optimizing the waveforms, a peak trans-
mitted power constraint must be considered, which can be represented as
K
X
∀ i, k and q, fqk [i] ≤ P max , and fqk [i] ≥ 0, (4.40)
k=1
where P max represents the peak LED transmitted power. After the optimization pro-
cess (finding the optimal waveforms in (4.39)), the SINR for all the users are similar.
(k)
The transmitted data rate for user k can be calculated by Rb = Rc (log2 Mk )/Lf ,
where we assume the sampling rate, Rc , for each user is the same. Rc /Lf represents
the transmitted symbol rate.
The maximum data rate for each user is constrained by the required BER, B max ,
since the communication quality needs to be taken into account. For a fixed SINR,
the modulation constellation size determines the BER and the transmitted data rate.
Therefore, to maximize the data rate for each user, we need to solve the following
problem:
Mk∗ = max Mk , ∀ k = 1, · · · , K
(4.41)
s !
Mk − 1 γk (F∗ , M, Rc )
s.t. erfc < B max ,
Mk log2 Mk (Mk − 1)2
where Mk∗ is the optimal value for Mk to maximize the data rate for user k.
The steps for solving (4.41) and getting the optimal waveforms are described in
82
Algorithm 2. We use the genetic algorithm (GA), a powerful heuristic searching

method, to find the optimal waveforms [77].
Algorithm 2: Optimal waveforms and the highest data rate

Initialize: Rc , Lf ;
repeat
for k = 1, 2, · · · , K do
for Mk = 2, 4, 8, 16 do
while Constraint in (4.41) is satisfied do
GA begins;
Initialization for GA;
Generate random individuals for F (1st GEN);
repeat
Evaluate the fitness function, γk (F, M, Rb );
Select the individuals by checking the constraint (4.40);
Match, mutate and crossover;
Generate the next generation;
until γk converges;
Get F∗ ;
GA ends;
Calculate BER using (4.38);
(k)
Calculate Rb ;
end
end
end
Increase Rc ;
(k)
until Rb converges;
(k)
Output: The maximum Rb , optimal M, and F∗
Illumination and Dimming Control
For VLC systems, illumination control is an important consideration. The trans-

mitted optical power can provide wireless access as well as illumination. The maxi-
mum illumination level in the room depends on the peak transmitted power and the
illumination potential, defined as the ratio of the highest average transmitted power
to the peak LED power. For this work, the illumination potential using JOW can be
83
represented as
Q K X Lf
1 X X
ηJ = fqk [l]. (4.42)
2QKLf P max q=1 k=1 l=1
In the optimization process, ηJ is an optimization constraint that can be adjusted to

satisfy the specific illumination requirements by changing the values of waveforms. In
(4.42), the coefficient 1/2 comes from the uniform distribution of sk [i]. The illumina-
tion level can be controlled by finding the optimal waveform as shown in Algorithm
??, inserting (4.42) as an additional constraint.
For OCDMA, the illumination potential is fixed, and depends on the codewords.
The illumination potential using optical orthogonal codes (OOC) as waveforms in an
OCDMA system can be calculated as

 W/2Lc ,
 K≤W
ηC = , (4.43)
 K/2Lc ,

K>W
where W is the weight for the OCDMA codeword, and Lc is the length of the code.
Therefore, if a certain OCDMA code is selected, for a certain number of active users,
the illumination potential of using OCDMA cannot be changed. JOW has a more flex-
ible illumination potential than OCDMA due to the optimally designed waveforms.
When using the proposed JOW algorithm in indoor VLC systems, the illumination
level can be adjusted by designing for a specific illumination potential.
For TDMA, only one user is served per time slot, and the data for each user can
be sent directly. Barring any DC offset, the illumination potential for TDMA is a
constant, which can be represented as ηT = 1/2. Compared with JOW, we can state
that TDMA is at least as power efficient as JOW, i.e., ηJ ≤ ηT .
84
Off-Line Waveform Design
The proposed waveform design algorithm is a time consuming process due to

the non-linear and non-convex optimization. In this section, we propose an off-line
solution to make the system adapt in real-time. In the off-line method, we calculate
the waveforms for multiple users in advance. Given the number of lamps in the
space, a table of the waveforms for different numbers of users and channel gains can
be created. In typical VLC systems, only a few users can be served by any one lamp.
In practice, depending on the number of users and the channel gains, the proper
waveforms for the users can be selected from the pre-established tables.
Since the LED impulse response, hl (t), can be estimated perfectly, the only factor
that can affect the off-line solution’s performance is the channel gains. The more
channel gain choices are used to create the table, the better the performance the
off-line waveforms algorithm can achieve. One table is created for each possible value
of K. We assume that the initial channel gains that are used to create the table can
be represented as a matrix
 
 µ11 µ12 ··· µ1K 
 
 µ21 µ22 · · · µ2K 
 
UK =  . , (4.44)
 
 .. .. .. .. 
 . . . 
 
 
µLT 1 µLT 2 · · · µLT K
where each row represents one set of the initial channel gains for the K users in the
table. LT is the number of sets of channel gains, which decides the size of the table.
To create the table, µik can be used to replace ĥqk , ∀ q in (4.20). Then, the waveform
lookup table can be created by using the proposed algorithm in this chapter.
During operation, the table corresponding to the number of active users K is first
85
Table 4.2: Parameters Used for Small Indoor Environment
Size of the small room 5m×5m×3m

Locations of the lamps (1.25,1.25,3),(1.25,3.75,3)
(3.75,1.25,3),(3.75,3.75,3)
Responsivity 0.5 A/W
Area of the photodetector 0.01 cm2
Radiated optical power per lamp 3W
LED semiangle 60o
Noise spectral density 1 × 10−9 mW/Hz
3 dB bandwidth of LEDs 20 MHz
Modulation constellation size 2, 4, 8, 16
BER requirement, B max 10−4
selected. Then, based on the real estimated channel gains for the multiple users, the
proper waveforms can be selected by using the following criteria
K
X 2
i∗q = arg min µik − ĥqk , ∀ q, (4.45)
i
k=1
where i∗q is the index of the channel gains selected for LED q. The performance of
the off-line algorithm is essentially equivalent to a channel uncertainty with
K
1 X
σh2 = (µi∗ k − ĥqk )2 . (4.46)
K k=1
4.2.4 Numerical Results and Discussions
In this section, numerical results of the performance of the proposed system are
shown. To test the applicability of the system, we show results for an indoor envi-
ronment with four LED lamps. This JOW does not use ηJ as a constraint. Unless
otherwise noted, the parameters used to obtain the numerical results are shown in
Table 4.2.
86
200
190
Maximum bit rate per user (Mbps)

180
170
160
150
140 2−PAM
4−PAM
130 8−PAM
16−PAM
120
2 4 6 8 10 12 14
Number of samples per waveform
Figure 4.12: Transmitted data rate for different numbers of samples per waveform for
3 users.
Perfect CSI
In this chapter, adaptive M-PAM is used together with JOW to enhance the
transmitted data rate. Only 2, 4, 8, and 16-PAM are considered in this work for our
numerical results. The number of samples per waveform, Lf , is an adjustable param-
eter, which needs to be sufficiently large to reduce the ISI and MAI. Fig. 4.12 shows
the numerical results of the optimized data rate with different numbers of samples
per waveform using M-PAM to satisfy a BER 10−4 . The channel gains for the 3 users
are 0.036, 0.032 and 0.025, respectively. In general, as the number of samples per
waveform increases, a higher data rate can be supported by using M-PAM. However,
the data rate achieves a limit when the number of samples per waveform is 11. Thus,
considering the computational burden and design complexity, the optimal number of
samples per waveform is 11 for this case. From the results, 8-PAM can provide the
87
20
Symbol rate=8 Msps, Lf=3
18 Symbol rate=10 Msps, Lf=3
16 Symbol rate=8 Msps, Lf=5

Symbol rate=10 Msps, Lf=5
14
ISI limit
12
SINR (dB)
10
MAI limit
8
ISI limit
6
4
MAI limit
2
2 3 4 5 6 7
Number of users, K
Figure 4.13: SINR for different users.
highest data rate among the other modulation schemes. Since we have a BER con-
straint to guarantee the communication quality, the larger modulation constellations
require a higher SINR. When the system cannot provide a high enough SINR for
the current M-PAM to satisfy the BER requirement, a lower level modulation needs
to be used. We envision an adaptive procedure that adjusts the constellation size
depending on the channel quality and received SINR.
More users can introduce more MAI. Fig. 4.13 shows the SINR for up to 7 users.
The channel gains for these 7 users are 0.036, 0.032, 0.032, 0.028, 0.025, 0.021, and
0.018. If the number of users K exceeds the number of samples per waveform Lf ,
the MAI dominates over the ISI. In Fig. 4.13 the numerical results show that the
SINR drops sharply when K is larger than Lf , and the system enters the MAI limited
region. Thus, depending on the number of active users in this room, we can select
the minimum number of samples per waveform.
88
0.5
0.4 Lf=11
Lf=3 Lf=5 Lf=7 Lf=9
Illumination potential 0.3
0.2
TDMA
JOW
0.1
OCDMA, OOC, Lc=7
OCDMA, OOC, Lc=13
0
2 3 4 5 6 7 8 9 10
Number of users, K
Figure 4.14: Illumination potential for different numbers of users.
The illumination potential is shown in Fig. 4.14. From the result, TDMA has
the highest illumination potential since it only serves one user per time slot. The
illumination potential using TDMA is the expected value of the data, which is the
maximum achievable. OCDMA and JOW follow a similar principle to support mul-
tiple users. Depending on the codewords or waveforms, the power efficiencies for
illumination of OCDMA and JOW are different. For OCDMA, this illumination po-
tential is increasing as the number of users increases. Eventually, when the number
of active users is equal to the length of the selected OCDMA code, the illumination
potential for OCDMA can reach its maximum value since the value of the sum of the
unmodulated codewords is P max . In Fig. 4.14, when the number of users is lower
than 7, a OOC code with length Lc = 7 is enough. However, when the number of
users is greater than 7, a length Lc = 13 OOC code is needed, since the length 7
OOC code cannot support that many users. Therefore, a longer code length has a
lower illumination potential for OCDMA. Comparing OCDMA and JOW, JOW can
89
20
JOW, Lf=7, No ηJ constraint
JOW, Lf=7, ηJ=21%
18
OCDMA, Lc=7, ηC=21%
TDMA, ηT=50%
16
SINR (dB)
14
12
10
8
60 80 100 120 140
Symbol rate (Msps)
Figure 4.15: SINR comparison of JOW, OCDMA and TDMA for 3 users.
Table 4.3: Maximum data rates using M-PAM averaged over the 3 users
Data rate (Mbps) 2-PAM 4-PAM 8-PAM 16-PAM

No ISI, ideal channel 532 476 403 352
JOW, Lf = 7 151 180 189 172
OCDMA, Lc = 7 78 121 137 164
No-equalization 24 44 61 69
provide higher illumination potentials for most cases, and the illumination potential
of JOW can achieve 80% of the maximum value.
Fig. 4.15 shows numerical results of SINR for JOW, OCDMA and TDMA tech-
niques. In this result, ηJ is used as an optimization constraint. For both OCDMA
and TDMA, MMSE filters are applied at the receivers. For the same transmitted
symbol rate, JOW has a higher SINR than TDMA and OCDMA since the optimized
waveforms can reduce ISI and MAI together with the MMSE filter. Since there is
90
no MAI for TDMA, the SINR for TDMA is higher than OCDMA. When the JOW
has the same illumination potential as OCDMA, the SINR for JOW is greater than
OCDMA.
Since a higher modulation level requires a higher SINR to satisfy the commu-
nication quality (BER requirement), the larger modulation constellation size cannot
always provide higher data transmission rates. In Table. 4.3, numerical results for the
maximum data rate with different modulation constellation sizes are shown. 2-PAM
can provide the highest data rate for the ideal channel. Since there is no ISI for the
ideal channel (MAI without ISI), using higher levels modulation does not increase the
throughput. For the case where the LED bandlimit is applied but no equalization is
used, a higher level modulation can provide a higher transmission data rate. With
the help of JOW, ISI can be reduced; therefore, the optimal modulation constellation
size for JOW is 8 for this case. OCDMA has more ISI than JOW, thus, 16-PAM
needs to be used for OCDMA to achieve the maximum data rate. In general, as the
ISI increases, the optimal modulation constellation size increases.
Imperfect CSI
The imperfect CSI case is also considered in this chapter. Fig. 4.16 shows the
comparison of the perfect and imperfect CSI cases. For the imperfect CSI case, if the
channel uncertainty is known, there is about a 4 dB SINR penalty compared with the
perfect CSI case when the channel uncertainty is −20 dB (the channel uncertainty is
36% of the average channel gain).
In Fig. 4.16, numerical results also show cases with and without knowing the
channel uncertainty information, i.e., the uncertainty variance. From the results, the
algorithm that knows the channel uncertainty variance can obtain a higher SINR
than if it does not know the variance. When the variance of the channel uncertainty
91
16
Perfect CSI
14
12
SINR (dB) 10 Symbol rate=80 Msps
Symbol rate=100 Msps

8
4
Not Knowing σ2h
2
knowing σ2h
0
−70 −60 −50 −40 −30 −20
σ2h (dB)
Figure 4.16: SINR for imperfect CSI with different uncertainty variance, Lf = 7 and
3 users.
is −20 dB, knowing the variance can provide about 2 dB SINR advantage over not
knowing the variance.
The same method used to estimate the effect of imperfect CSI can be used to
evaluate the off-line algorithm. The difference between ĥqk and µik can be regarded
as a known channel uncertainty. From this results, if the variance of the difference
is around −20 dB, the performance of the off-line algorithm can provide about 4 dB
less SINR compared to the regular (on-line) algorithm.
92
4.3 Comparison of DCO-OFDM and M-PAM
4.3.1 Background
Recently, orthogonal frequency division multiplexing (OFDM) has been employed

in OWC systems due to its resistance to inter-symbol interference (ISI) and high
spectral efficiency [23, 24]. Since intensity modulation and direct detection are used
in OWC systems, the transmitted signal should be non-negative. Therefore, the
conventional OFDM cannot be applied directly in OWC. DC-biased optical OFDM
(DCO-OFDM) is a popular optical OFDM technique that can be applied in OWC
that use incoherent light [7]. Hermitian symmetric data is used to make the DCO-
OFDM signal real. Because of the nonlinearity of LEDs, the DCO-OFDM signal
must be clipped, distorting the signal.
In this section, we compare the performance of DCO-OFDM and M-PAM tech-
niques for OWC systems. For DCO-OFDM, we consider the clipping noise caused by
the LEDs’ nonlinearity (clipping at both zero and peak current). We optimize the
modulation index and the bits loaded on each subcarrier to maximize the transmitted
bit rate. In this section, to simplify the analysis, we consider single user operation.
4.3.2 Optimized DCO-OFDM
In this section, we describe how we optimize DCO-OFDM. For VLC systems, due
to the nonlinearity of the LEDs, the DCO-OFDM signals may be clipped by the LEDs.
The optimized DCO-OFDM scheme maximizes the transmitted bit rate by optimizing
the modulation index and the bits loaded on all subcarriers. A block diagram of the
optimized DCO-OFDM is shown in Fig. 4.17. In this diagram, Xi [m] is the data to
be modulated by ith subcarrier at the mth time instant after M-QAM. We assume
93
X 1[ m ] xofdm [1, m]
X 2 [ m] xofdm [2, m]
...
QAM
...
Modulation
=
X Nsub /2 [m] Index
Binary
data XN /2 1 [m] IFFT P/S
D/A Clip LPF
sub
sofdm (t )
DC Bias LED
QAM*
...
XN 1 [ m]
sub
X N [ m] xofdm [ N , m]
sub
Figure 4.17: Diagram of DCO-OFDM with adjustable modulation index and loaded
bits.
that there are Nsub subcarriers. To make the OFDM signal real, XNsub +1−i [m] is the
conjugate of Xi [m], XNsub +1−i [m] = Xi∗ [m]. After modulation, the real OFDM signal
for the kth subcarrier component, xofdm [k, m], can be represented as
Nsub
X j2πki
xofdm [k, m] = Xi [m]e Nsub , ∀ k = 1, 2, · · · , Nsub (4.47)
i=1
After converting the parallel data to a serial stream, adding a DC offset and the D/A
converter, the electrical signal sofdm (t) can be represented as
∞ Nsub
% X X
sofdm (t) = xofdm [k, m]g(t − k − mTofdm ) + sdc , (4.48)
Nsub n=−∞ k=1
where the term %/N is referred to as the modulation index. g(t) is the signal pulse
function, and Tof dm is the duration of the pulse. sdc is the DC bias, which is set to
sdc = Imax /2, where Imax is the saturation current to drive the LEDs. When Nsub
is large (usually greater than 64), the analog signal sofdm (t) can be modeled as a
Gaussian random process.
In order to prevent the LEDs from damage, the drive current should remain in
the range of [0, Imax ]. Considering the bandlimited characteristic of LEDs, we model
94
the LED as a clipping component and a lowpass filter in series as shown in Fig. 4.17.
Therefore, the signal outside the range [0, Imax ] is clipped.
After matched filtering and sampling at the receiver, the received signal can be
modeled as [78]
yclip [m] = αsofdm [m] ∗ h[m] + nclip [m], (4.49)
where h[m] is the discrete time version of the impulse response of the LED. Since the
clipping effect is a non-linear operator, the constant coefficient α can be found by
using the Bussagang theorem, [78]:
!
I
α = 1 − erfc pmax , (4.50)
8σs2
√ R∞ 2
where erfc(x) = 2/ π x e−y dy, and σs2 is the variance of the OFDM signal, sofdm (t).
We model the clipping noise, nclip [i], as a zero mean Gaussian variable with a variance
estimated using
Z 0 Z ∞
2 2
σclip = (αx) f (x)dx + (αx − Imax )2 f (x)dx, (4.51)
−∞ Imax
where f (·) is the probability density function (pdf) of the samples αsofdm [m].
In Fig. 4.18, the constellation of the clipped signal and the original signal are
shown. From the plot, the clipping effect not only introduces noise, but also causes
distortion. Since the clipping effect limits the peak power of the transmitted signals,
the constellation of the clipped signals are shrunk. Using the model in (4.49), the
constellations of the modeled clipped signals are illustrated in Figs. 4.19 and 4.20.
From the results, (4.49) can perfectly model the clipping effect.
At the receiver, the signal to noise ratio (SNR) for the ith subcarrier can be
95
1.5
0.5
−0.5
−1
−1.5
−1.5 −1 −0.5 0 0.5 1 1.5
Figure 4.18: Signal constellation with clipping only, 4-QAM.
1.5
0.5
−0.5
−1
−1.5
−1.5 −1 −0.5 0 0.5 1 1.5
Figure 4.19: Signal constellation using (4.49), 4-QAM.

96
−1
−2
−3
−4
−4 −3 −2 −1 0 1 2 3 4
Figure 4.20: Signal constellation using (4.49), 16-QAM.
calculated as
(i) (%αHi E{|Xi |})2
γofdm = 2 2
, (4.52)
N σofdm + σclip
where Hi is the LED response for the ith subcarrier. E{·} represents the expectation
2
operation, and σofdm is the variance of the receiver additive Gaussian noise in the
ith subcarrier. Given the SNR, we can calculate the bit error rate (BER) for each
subcarrier by using the approximate expression [34]
q v 
(i)
Mofdm − 1 u 3γ (i)
u
ofdm
BERi ≈ q q erfc t (i)
 , ∀ i, (4.53)
(i)
Mofdm log2
(i)
Mofdm 2Mofdm − 2
(i)
where Mofdm is the modulation constellation size for the QAM used in the ith sub-
carrier. The simulation and theoretical results using (4.53) are shown in Fig. 4.21.
With an increasing modulation index, the SNR increases, thus the BER decreases.
However, when the clipping effects dominates the noise, increasing the modulation
97
0
10
−1
10
−2
10
BER
−3
10
Simulation, M=16
−4
10 Theoretical, M=16
Simulation, M=32
Theoretical, M=32
−5
10
0 0.05 0.1 0.15
Modulation Index/Pmax
Figure 4.21: BER comparison of simulation and theoretical results using M-QAM.
index can make the BER worse, as expected.

The throughput for the DCO-OFDM can be calculated as
Nsub /2
Nsub X (i)
Rb = log2 Mofdm . (4.54)
2Tofdm i=1
To optimize the throughput, we can choose the optimal Tof dm , %, and the number of
bits loaded onto each subcarrier. In this section, for each subcarrier, the subcarrier
bit loading is constrained by the BER requirement and the LED bandwidth filter.
Usually, with the help of forward error correction (FEC), the BER requirement for
each subcarrier can be set at 10−3 .
4.3.3 Numerical Results and Comparison
In this section, numerical results of the comparison of DCO-OFDM and M-PAM

are shown. To obtain a fair comparison, the same parameters are used for DCO-
98
OFDM and M-PAM. We assume that the 3 dB bandwidth of the LED, f3dB = 20 MHz.
The additive noise power spectral density is No = 10−9 mW/Hz. The BER require-
ment for communication quality is B max = 10−3 . To simplify the problem, an ideal
channel response (zero loss) is considered in this section. In addition, the forward
current to optical power conversion ratio of the LED is assumed to be unity. Thus, the
saturation current constraint implies the constraint on the peak transmitted optical
power.
With the help of bit loading, the throughput of DCO-OFDM can be maximized.
Adjusting the modulation index, a compromise is reached between the signal power
and clipping noise power. In Fig. 4.22, the throughput with different modulation
indexes using bit loading is shown. From the results, the number of subcarriers does
not seem to affect the maximum throughput. In this result, the peak transmitted
power is 10 mW.
Fig. 4.23 shows a comparison between the optimized DCO-OFDM and the op-
timized M-PAM using JOW. In this figure, Rb /f3dB is used to measure the spectral
efficiency for both M-PAM and DCO-OFDM. For JOW, the waveform design algo-
rithm and MMSE equalizer can use more than the 3 dB bandwidth of the LED, and
the clipping distortion caused by the nonlinearity of the LED can affect the perfor-
mance of the DCO-OFDM; therefore, using M-PAM with JOW can provide a better
performance than DCO-OFDM. From the results, the M-PAM using JOW can pro-
vide an 80% higher throughput than the optimized DCO-OFDM described in Section
4.3.2. With the help of waveform design, M-PAM can achieve a higher bit rate than
using only the MMSE equalizer. If there is no equalization technique for M-PAM, the
optimized DCO-OFDM can support about five times higher data rate than M-PAM
when the transmitted power is 8 mW.
99
8
x 10
2.5
1/Tofdm=0.97 MHz
1/Tofdm=3.23 MHz
1/Tofdm=4.84 MHz
2
1/Tofdm=0.48 MHz
1/Tofdm=1.59 MHz
Throughput (bps) 1/Tofdm=2.38 MHz
1.5
1
Nsub=64
0.5
Nsub=128
0
0 0.2 0.4 0.6 0.8 1 1.2 1.4
Modulation Index
Figure 4.22: Throughput of DCO-OFDM with bit loading for different modulation
indexes.
20
DCO−OFDM
18 JOW, Lf=5, Lw=15
JOW, Lf=5, Lw=13
16
MMSE Equalizer, Lw=15
Spectral Efficiency (bps/Hz)
14
MMSE Equalizer, Lw=13
12 M−PAM, No Equalization
10
0
0 2 4 6 8
Peak Received Power (mW)
Figure 4.23: Throughput comparison of optimized DCO-OFDM and optimal M-PAM

for a single user.
100
4.4 Summary
In this chapter, we first propose an adaptive M-PAM algorithm using OCDMA

and MISO techniques to support multiple users. Depending on the SINR at the
receivers, the proposed algorithm can choose the optimal power allocation and M-
PAM constellation size for the users to optimize the bit rate fairly. It is able to adapt
to different shadowing effects and desired BERs. Compared with the same algorithm
using TDMA, the proposed algorithm using OCDMA can offer a higher bit rate when
the number of users is larger than the length of the OCDMA code.
Then, a joint optimal waveform design algorithm using adaptive M-PAM modu-
lation scheme is proposed to provide high data data transmission rates for multiple
users. In this algorithm, the waveforms transmitted from each LED lamp to all the
users are uniquely designed to reduce the ISI and MAI. Together with the MMSE
filters at the receivers, the transmitted waveforms are optimized to provide the max-
imal data rates. The high level modulation can provide high data transmission rates.
The modulation constellation size can be adaptively adjusted under different SINR
conditions and BER requirements. In this work, the channel uncertainty and imper-
fect CSI cases are discussed. From numerical results, the proposed robust algorithm
can get a higher SINR than the non-robust algorithm by knowing the channel un-
certainty information. Compared with OOC codes, the proposed JOW can provide
higher SINR. Considering the illumination and dimming control, JOW is more flex-
ible than CDMA. The illumination level can be easily adjusted in JOW by setting
a illumination constraint. The waveform length is another design parameter, we can
find the optimal waveform length to maximize the data rate. In conclusion, the JOW
using adaptive M-PAM can provide high data transmission rate over ISI channel for
multiple users. Compared with CDMA, JOW has prominent advantages in SINR and
101
illumination.
In the last section, we compare the performance of DCO-OFDM and M-PAM
for LED-based communication systems for a single-user operation. Considering the
LED bandlimited characteristic and the constrained transmitted power, we propose
an optimized DCO-OFDM by choosing the optimal modulation index with an appro-
priate bit loading algorithm. For M-PAM, our waveform design joint optimization
algorithm is used to reduce the ISI caused by the bandlimited LED. In this section,
from the simulation and theoretical results, we conclude that M-PAM with waveform
design and MMSE equalization can provide an 80% higher data transmission rate
than DCO-OFDM with bit loading and an optimal modulation index.
Chapter 5
Interception Vulnerability Analysis
VLC is considered as an alternative to Wi-Fi systems to provide wireless network

access in indoor areas due to some critical shortcomings of Wi-Fi. A major-draw
back of Wi-Fi systems is its susceptibility to eavesdropping, which has limited its use
in security sensitive environments. RF signals that are used in Wi-Fi systems can
penetrate through walls, and this leakage can expose the information carried by these
signals to hackers. Unlike Wi-Fi systems, the signals in VLC systems are blocked
by any opaque object, and therefore, the signals cannot pass through the walls. We
can assume the RF leakage from the VLC modulator is controlled. As a result, 100%
security against outside eavesdroppers can be guaranteed in VLC systems if no light
escapes the room. However, this is usually not the case in real scenarios: the light
leaks out through glass doors and windows, which can impose security risks on the
VLC users.
In [79], the security risks of VLC systems due to the light leaked from under the
door and through the key hole have been analyzed for various room setups. In [80],
Mostafa and Lampe analyze the security of VLC systems in an indoor area against
an internal hacker and, in [81], they propose a multiple input and single output MISO
102
103
3m
Vulnerable Zone
5m
5m
Figure 5.1: Vulnerable area outside the room
system to increase the privacy of VLC systems through pulse shaping. [82] presents
a high security CDMA scheme for VLC systems, and derives an information leakage
expression. Pan et. al. investigate the secrecy performance of a VLC system with a
group of randomly distributed eavesdroppers in [83]. There have been more general
analyses on the security of free space optical systems that can also be extended to
VLC systems [84].
To the best of our knowledge, there is no study on the information leakage of VLC
systems through a window and its security risks. Fig. 5.1 shows an example of the
information leakage through a window. We define the area with high security risks
as vulnerable zone. In this work we present an analysis for the strength of the VLC
signals leaked through windows, and based on this, we calculate the specifications of
devices required for eavesdropping. We divide the area outside the window into high
risk and low risk regions based on the access to the line of sight (LOS) signals, and
calculate the size of the minimum aperture in each region to achieve the minimum
SNR required for detection of the information. We then propose a technique to
reduce the size of the high-risk region by optimizing the modulation efficiency that
controls the percentage of the optical power for carrying the data. This algorithm
104
can provide a flexible highly secure communication link for indoor users and minimize
the possibility of data interception in outdoor areas.
5.1 System Performance and Security
5.1.1 System Performance
The criteria used to evaluate the performance of the VLC system are the BER and
the data rate. The relationship between these two metrics is discussed in this section.
Since IM/DD is employed in VLC, M-PAM can be used to increase the data rate for
a fixed transmitter bandwidth. Using M-PAM, a (log2 M )-fold increase in the data
rate compared with OOK can be achieved. The choice of modulation constellation
size M depends on the SNR at the receiver. In this chapter, an adaptive technique
is assumed, where the largest constellation size allowed by the SNR of the intended
user is used to get the highest data rate [29]. No waveform optimization is included
here.
We assume the desired-user SNR is determined by the channel quality, transmitted
power, and the size of photo-detector (PD). For a given channel, the SNR for user k
can be calculated as 2
PQ
ρP max Ar `=1 β` ĥ`k
γk = (5.1)
σn2
where ρ represents the the responsivity of the PD. P max is the peak transmitted power
and Q is the number of LEDs. Ar is the area of the intended receiver’s PD, which
is small in the VLC system, where the receiver is in the room. ĥ`k is the channel
gain from LED ` to user k, which can be calculated from (2.2). β` represents the
modulation efficiency that describes the percentage of the `th LED’s optical power
used for data transmission. The value of β` does not affect the total emitting power
105
of the LEDs. σn2 represents the noise power at the receiver, which includes thermal
noise and shot noise.
5.1.2 System Security
The VLC system has high security because of the characteristics of light that
it cannot penetrate opaque objects like walls. It keeps the communication system
secure in the wireless physical-layer. When the eavesdropper cannot have a good
quality channel, the transmitted information cannot be reconstructed [79].
The size of the receiver aperture (the lens placed in front of the PD) is a signifi-
cant factor to be considered by eavesdroppers. The smaller the eavesdropper receiver
aperture becomes, the harder it is to expose it, since a large receiver aperture makes
the eavesdropping equipment easily spotted. However, a too small size receiver aper-
ture cannot support a sufficiently high SNR to intercept the signal. The required
eavesdropper SNR depends on the desired modulation constellation size and BER.
Consider the case that the eavesdropper is outside the window during the daytime.
The additional shot noise from sunlight helps to lower the received SNR, and therefore
improves the security of the communications. In this chapter, we assume the sunlight
illumination is sufficiently bright to make the shot noise caused by background light
the dominant noise source. For this case, the eavesdropper minimum receiver aperture
required to obtain a BER = 10−3 can be calculated as
2qRs Φsun γk
Ar ≥ , (5.2)
ρ(ϕP max Q
P 2
`=1 β` ĥ`k )
where Φsun is the power density of the sunlight on the ground, which can be calculated
from the sunlight illumination. ϕ is the transmittance percentage through normal
glass. Rs is the transmitted symbol rate, the value of which is assumed to be the
106
same as the bandwidth. For the indoor area or the outside area at night, we assume
the background light is negligible, and therefore the thermal noise is dominant. For
this case, the minimum receiver aperture (for both intended and unintended users)
can be calculated using p
2
γk σthermal
Ar ≥ . (5.3)
ρϕP max Q
P
`=1 β` ĥ`k
If a very small size of receiver aperture is necessary, a high gain PD, such as an
avalanche photo diode (APD), can be used. The actual PD size can be calculated
roughly as
Ar
Âr = , (5.4)
G
where G is the gain of the APD, and Âr is the effective aperture size.
5.1.3 LED Modulation Efficiency Optimization
In order to reduce the leakage of information and enhance the indoor communica-
tion performance, we can optimize the modulation efficiency of each LED. With the
help of the lamp model shown in Fig. 2.1, the security and indoor data transmission
quality can be considered at the same time.
We assume that the smaller the eavesdropper receiver aperture becomes, the
harder it is to expose it. On the other hand, a larger aperture can collect more
optical power and achieve a higher SNR to capture the data. To minimize the possi-
bility of interception and guarantee the indoor communications, we can optimize the
modulation efficiency by
max min γkin

β1 ,β2 ,··· ,βQ kin
s.t. max (γkout ) ≤ γmin , (5.5)
0 ≤ β` ≤ 1, ` = 1, · · · Q,
107
where kin is any user in the indoor area, and kout represents any eavesdropper outside
the window. γmin is the minimum SNR for data detection. γkin and γkout represent
the SNR for the any indoor user and outdoor eavesdropper, respectively, which can
be calculated by using (5.1)
PQ 2
max
ρP Ain `=1 β` ĥ`kin
γkin = 2
,
σin
2 (5.6)
max
PQ
ρP ϕAout `=1 β` ĥ`kout
γkout = 2
σout
where Ain and Aout represent the desired user aperture and eavesdropper aperture,
2
respectively. σin represents the noise power for indoor users, which can be assumed as
2 2 2 2
σin = σthermal ; σout is the noise power for the outdoor eavesdropper, which is σthermal
2 2
at night. σout = σshot in the daytime.
To find the optimal modulation efficiency, an iterative method, the SQP algorithm,
can be used. Since the maximin optimization is non-convex, only a locally optimal
solution of the modulation efficiency can be obtained.
5.2 Simulation Results and Analysis
To test the system vulnerability, we set up a scenario as an empty and unfurnished

room with dimensions of 5 m × 5 m × 3 m. There is only one window of 2 m × 2 m
in size, and the low edge of the window is 1 m off the floor. The parameters used to
obtain the simulation results are shown in Table 5.1. All measurements are simulated
at the floor level.
108
Table 5.1: Parameters Used in Numerical Results
LED lamp placement (2.5, 2.5, 2.9) m

Peak radiation power of each LED 500 mW
Number of LEDs per lamp 25
Semiangles of LEDs 20o
Intended user aperture, Akin 10−4 m2
Eavesdropper aperture, Akout 1 m2
Transmittance percentage through window, ϕ 1
Sunlight illumination 1.2 × 105 lx
2
Thermal noise power, σthermal 6.56 × 10−15 W
2
Background light noise power, σshot 1.12 × 10−13 W
5.2.1 No LED Modulation Efficiency Optimization
In this section we consider the vulnerability of the system when all LEDs transmit
data (β` = 1, ∀ `), unconcerned with security issues.
For a typical office room, the illumination level should be around 400 lx [49]. To
make this requirements, the SNR distributions in the indoor area and the outdoor
area are shown in Figs. 5.2 and 5.3, respectively. In these two results, we assume
the size of the PD at the intended receiver is 10−4 m2 . For the indoor area, the
thermal noise is dominant, and for the outdoor area, the shot noise caused by the
background sunlight is dominant. In general, the SNR in the indoor area is more
than 38 dB and is distributed symmetrically around the LED lamp. The high indoor
SNR can support a reliable and high speed wireless data transmission by using a
large modulation constellation size. For instance, an SNR of 38 dB allows us to use
an 16-PAM.
The SNR distribution outside the room varies dramatically depending on where
the receiver is placed and the background light levels. The area just outside the
window has a higher SNR than other places. The high SNR area outside the window
is referred to as the vulnerable zone in Fig. 5.3. In this plot, we choose a 20 dB SNR
109
45
SNR (dB)
40
35
400
400
200 200
Width (cm) 0 0 Length (cm)
Figure 5.2: SNR distribution in the indoor area
Indoor Outdoor
Vulnerable Zone
Window
20
0
0
−20
SNR
(dB)
−40 100
−60 200
−80
300
Width (cm)
−100
0 400
200
400
Length (cm) 600 500
Figure 5.3: SNR distribution in the outdoor area, daytime.

110
120
Required SNR=32dB
110 Required SNR=20dB
100
90
80
Gain (dB)
70 Sunlight dominant
60
50 No background light
40
30
20
0 0.2 0.4 0.6 0.8 1
Received Power Density (W/m2) −5
x 10
Figure 5.4: Gain of the PD requirements for different received power densities.
as a threshold, corresponding to a constellation size of M = 8. The area with SNR

greater than 20 dB is defined as the vulnerable zone. In the vulnerable area, the LOS
light dominates.
Fig. 5.4 shows the relationship between the required gain of the receiver and the
received power density. Two extreme cases are compared in this result. The sunlight
dominant case represents an outdoor scenario in daylight. The indoor area and the
outdoor area at night can be assumed to experience no background light. From
the results in Fig. 5.4, for the same received power density, the sunlight-dominant
case needs around 30 dB higher gain than the no-background light case to get the
same SNR. Thus, the background sunlight indeed makes it more difficult for the
eavesdropper to detect the information.
111
20
−20
−40
SNR (dB)
−60
Aout=1 m2, Daytime

−80
Aout=10−2 m2, Daytime
−100
Aout=10−4 m2, Daytime
−120 Aout=1 m2, Nighttime 0
0 200
200
400 400
Length (cm) 600 Width (cm)
800 600
Figure 5.5: SNR distribution for outdoor areas by optimizing the LED modulation
efficiency using different sizes of aperture, γmin = 10 dB.
5.2.2 LED Modulation Efficiency Optimization to Diminish
Vulnerability
In this section we consider the vulnerability of the VLC system to eavesdropping

when the LED modulation efficiency optimization is adjusted. In this part, we test
two environment: daytime and nighttime cases. We assume that the sunlight only
affects the outside background light. For the modulation efficiency optimization, we
assume the detector of the eavesdropper is on the floor.
Fig. 5.5 shows the SNR distribution outside the window after applying the optimal
modulation efficiency of each LED in the lamp. For the optimization results, we
design γmin = 10 dB. If the eavesdropper detector uses a Aout = 1 m2 aperture, it
can only obtain up to a 10 dB SNR. However, a detector with a 1 m2 aperture is
very easily detected. From (4.38), a 10 dB SNR is the minimum value to support a
112
60
40
20
SNR (dB)
−20
−40 Sunlight, γmin=10 dB

No sunlight, γmin=10 dB
0
−60
0 10 20 Width (cm)
30 40 50 50
Length (cm)
Figure 5.6: SNR distribution for indoor areas by optimizing the LED modulation
efficiency, Aout = 1 m2 and Ar = 10−4 m2 .
communication system with a 10−3 BER by using 2-PAM. The outdoor eavesdropper
is unable to demodulate a 16 to 32-PAM signal that is used in indoor areas. If the
eavesdropper uses a detector with a smaller aperture, such as 10−2 m2 or 10−4 m2
as shown in Fig. 5.5, up to -10 dB and -30 dB SNR can be obtained, which are too
small to recover the data. No matter whether there is sunlight or not, our algorithm
can limit the outdoor SNR to a very low level. The SNR distributions of the sunlight
and no sunlight cases are similar.
After optimizing the modulation efficiency, the SNR distribution in the indoor area
is shown in Fig. 5.6. Although we assume the sunlight cannot affect the illumination
in the indoor area, the optimization of the modulation efficiency depends on the
outside sunlight. The strong sunlight can introduce more background shot noise than
the case without sunlight. Thus, it is more difficult for the eavesdropper to capture
the data under a strong sunlight circumstance. Compared with the nighttime case,
113
the sunlight can allow the system to use a higher modulation efficiency for a better
indoor wireless connection. Considering the results of Fig. 5.4 and 5.5, for both
the daytime and nighttime, we can guarantee a similar outdoor SNR distribution by
setting the same γmin , which is secure. With the help of sunlight, daytime can allow
up to a 40 dB enhancement of the indoor SNR compared with the nighttime.
5.3 Summary
In this chapter, we quantitatively analyze the potential vulnerability of the system

from eavesdropping by calculating the SNR in different areas outside a window. We
calculate the required gain and aperture size for the receiver in different areas in order
to illustrate how big the eavesdropper equipment must be to successfully recover
the signal. We show that the area where the LOS signal can be received is most
vulnerable. We also show that the daytime sunlight greatly reduces the possibility of
eavesdropping.
Based on our analysis, we propose a solution to improve the VLC system security
by implementing a modulation efficiency optimization algorithm that can minimize
the possibility of outdoor eavesdropping and maximize the minimum SNR in the in-
door area to guarantee a reliable indoor data transmission. The results show that by
doing this we can maintain indoor communication system performance while greatly
reducing the SNR outside the room. If the eavesdropper uses small detection equip-
ment, the SNR is not high enough to recover the data. Otherwise, to obtain an
acceptable SNR, a large size equipment should be used, which makes the eavesdrop-
per easy to detect.
Chapter 6
Conclusions and Future Work
6.1 Summary and Conclusions
In this dissertation, we investigate the multiuser MIMO indoor VLC systems.

Multi-LED transmitter and multi-detector receiver models are proposed and applied.
Following the Lambertian rule, the indoor VLC channel model is derived.
Then, considering multiple users cases, we propose a centralized and several decen-
tralized power allocation algorithms. In the centralized power allocation algorithm,
all the LED lamps in the room are coordinated and controlled by a central controller;
each LED lamp supports all the users within the indoor area. For standard indoor
office illumination level (400 lx), about 40 users can be supported with bit error rates
less than 10−3 using on-off keying and 70 MHz bandwidth of receivers at 5 × 10−7
W/Hz spectral density. The decentralized power allocation algorithms we propose
have similar bit error rate performance and less computational burden compared to
the centralized algorithm. In our decentralized algorithms, users are supported by
a subset of the LEDs, and so the optimization problem size of the decentralized al-
gorithms can be reduced, by as much as 90% for the large room scenario. For each
114
115
receiver, multiple photodetectors with different orientations are employed to improve

the signal to interference plus noise ratio. In addition, some practical considerations,
such as shadowing effects, illumination requirements, dimming control and transmit-
ted power quantization are taken into account. From numerical results, the proposed
adaptive power allocation algorithms can adjust the transmitted power to reduce
shadowing effects. Furthermore, the transmitted power can be dimmed to satisfy the
illumination requirements within a range of tolerance. In this dissertation, the system
uses the OCDMA technique to support multiple users. Time-space minimum mean
squared error filters at the receivers are designed to diminish the effect of multiple-
access interference in the indoor VLC system.
In addition, we discuss modulation schemes for VLC systems. We first propose
an adaptive M-PAM algorithm to support multiple users. The proposed algorithm
can adjust the modulation constellation size for each user to maximize the bit rate
under different channel environments such as shadowing, light dimming, and the
impact of multiple access interference. In our MISO approach, multiple LED lamps
coordinate to provide users with maximum data rates. We compare OCDMA using
our adaptive M-PAM with TDMA. The OCDMA technique can offer a higher bit rate
when the number of users is larger than the length of the OCDMA code. Second, a
joint optimal waveform design for visible light communication systems using M-ary
pulse amplitude modulation is proposed to support multiple users. The transmitted
waveforms and MMSE filters are jointly optimized to minimize the intersymbol and
multiple access interferences. Based on different channel conditions, the designed
waveforms and modulation constellation size can be adaptively adjusted to guarantee
the highest data rate. When the 3 dB bandwidth of LEDs is 20 MHz and the peak
transmitted power is 3 Watts, the maximum bit rate per user can be achieved about
200 Mbps using the proposed waveform design algorithm. Compared with OCDMA
116
and TDMA, the algorithm proposed in this dissertation can provide about 8 dB higher
signal to interference plus noise ratio when the transmitted symbol rate is 80 Mb/s.
In addition, the proposed algorithm can flexibly change the illumination level by
adjusting the optimization constraint. From numerical results, the proposed waveform
design algorithm can provide about a 20% higher illumination than OCDMA. Taking
the computational complexity into account, an off-line waveform design solution is
proposed, and the performance of the off-line solution can be estimated. Finally, a
comparison between the optimized DCO-OFDM and M-PAM with JOW is discussed.
From numerical results, the M-PAM with waveform design and MMSE equalizer can
provide an 80% higher data rate than the optimized DCO-OFDM with the same BER
for single user.
In the last part of the dissertation, the vulnerability of VLC systems in the physical
layer is discussed. we first analyze the potential vulnerability of the system from
eavesdropping outside the room. By setting up a signal to noise ratio threshold, we
define a vulnerable area outside of the room through a window. We compute the
receiver aperture needed to capture the signal and what portion of the space is most
vulnerable to eavesdropping. Based on the analysis, we propose a solution to improve
the security by optimizing the modulation efficiency of each LED in the indoor lamp.
The simulation results show that the proposed solution can improve the security
considerably while maintaining the indoor communication performance.
6.2 Future Work
As future work, the mobility of users should be considered. A faster way of

finding the optimized power allocation for users needs to be investigated. Optical
CDMA codewords reuse strategy can be explored. To reduce the clipping distortion
117
from LEDs, a new kind of optical OFDM technique that is suitable for LED-based
communication systems can be studied.
Bibliography
[1] T. Komine and M. Nakagawa, “Adaptive detector arrays for optical communi-
cation receivers,” IEEE Trans. Consum. Electron., vol. 50, no. 1, pp. 100–107,
2004.
[2] Z. Chen, D. Tsonev, and H. Haas, “Improving SINR in indoor cellular visible
light communication networks,” in 2014 IEEE International Conf. on Commun.,
2014, pp. 3383–3388.
[3] M. Noshad and M. Brandt-Pearce, “Application of expurgated PPM to indoor

visible light communications part I: Single-user systems,” J. Lightw. Technol.,
vol. 32, no. 5, pp. 875–882, 2014.
[4] P. Luo, Z. Ghassemlooy, H. L. Minh, E. Bentley, A. Burton, and X. Tang, “Fun-

damental analysis of a car to car visible light communication system,” in 2014
International Symposium on Commun. Systems, Networks Digital Signal Proces.
(CSNDSP),, 2014, pp. 1011–1016.
[5] S. Ucar, B. Turan, S. C. Ergen, O. Ozkasap, and M. Ergen, “Dimming support

for visible light communication in intelligent transportation and traffic system,”
in NOMS 2016 - 2016 IEEE/IFIP Network Operations and Management Sym-
posium, April 2016, pp. 1193–1196.
118
119
[6] “Free space optics (FSO) and visible light communication (VLC) / Li-Fi tech-
nology market by component (LED, photo detector, microcontroller, and soft-
ware), transmission type (unidirectional and bidirectional), application, and ge-
ography - global forecast to 2022, http://www.marketsandmarkets.com/Market-
Reports/visible-light-communication-market-946.html,” in Markets and Mar-
kets, Feb. 2017.
[7] M. Zhang and Z. Zhang, “An optimum DC-biasing for DCO-OFDM system,”
IEEE Commun. Lett., vol. 18, no. 8, pp. 1351–1354, Aug 2014.
[8] J. Armstrong and A. J. Lowery, “Power efficient optical OFDM,” Electronics

Letters, vol. 42, no. 6, pp. 370–372, March 2006.
[9] Y. Hong, J. Chen, Z. Wang, and C. Yu, “Performance of a precoding MIMO

system for decentralized multiuser indoor visible light communications,” IEEE
Photon. J., vol. 5, no. 4, 2013.
[10] Z. Yu, R. Baxley, and G. Zhou, “Multi-user MISO broadcasting for indoor visible
light communication,” in IEEE International Conf. on Acoustics, Speech and
Signal Processing (ICASSP) 2013, 2013, pp. 4849–4853.
[11] J. Chen, Y. Hong, X. You, H. Zheng, and C. Yu, “On the performance of MU-
MIMO indoor visible light communication system based on THP algorithm,”
in 2014 IEEE International Conf. on Commun. in China (ICCC), 2014, pp.
136–140.
[12] S. H. Chen and C. W. Chow, “Color-shift keying and code-division multiple-

access transmission for RGB-LED visible light communications using mobile
phone camera,” IEEE Photon. J., vol. 6, no. 6, pp. 1–6, 2014.
120
[13] C. Chen, D. Tsonev, and H. Haas, “Joint transmission in indoor visible light
communication downlink cellular network,” in 2013 IEEE Global Commun. Conf.
(GLOBECOM), 2013, pp. 1127–1132.
[14] D. Bykhovsky and S. Arnon, “Multiple access resource allocation in visible light
communication systems,” J. Lightw. Technol., vol. 32, no. 8, pp. 1594–1600,
2014.
[15] Y. Wang, L. Tao, X. Huang, J. Shi, and N. Chi, “Enhanced performance of a

high-speed WDM CAP64 VLC system employing volterra series-based nonlinear
equalizer,” IEEE Photonics Journal, vol. 7, no. 3, pp. 1–7, June 2015.
[16] T. Yamazato, A. Ohmura, H. Okada, T. Fujii, T. Yendo, S. Arai, and K. Ka-

makura, “Range estimation scheme for integrated I2V-VLC using a high-speed
image sensor,” in 2016 IEEE International Conference on Communications
Workshops (ICC), May 2016, pp. 326–330.
[17] K.-I. Ahn and J. K. Kwon, “Capacity analysis of M-PAM inverse source coding
in visible light communications,” J. Lightw. Technol., vol. 30, no. 10, pp. 1399–
1404, 2012.
[18] R. Lee, K. Yun, J.-H. Yoo, S. Y. Jung, and J. K. Kwon, “Performance analysis
of M-ary PPM in dimmable visible light communications,” in 2013 Fifth Inter-
national Conference on Ubiquitous and Future Networks (ICUFN), July 2013,
pp. 380–383.
[19] S. He, G. Ren, Z. Zhong, and Y. Zhao, “M-ary variable period modulation for
indoor visible light communication system,” IEEE Commun. Lett., vol. 17, no. 7,
pp. 1325–1328, 2013.
121
[20] T. Komine, J. Lee, S. Haruyama, and M. Nakagawa, “Adaptive equalization

system for visible light wireless communication utilizing multiple white LED
lighting equipment,” IEEE Trans. Wireless Commun., vol. 8, no. 6, pp. 2892–
2900, 2009.
[21] P. Haigh, Z. Ghassemlooy, S. Rajbhandari, I. Papakonstantinou, and

W. Popoola, “Visible light communications: 170 Mb/s using an artificial neu-
ral network equalizer in a low bandwidth white light configuration,” J. Lightw.
Technol., vol. 32, no. 9, pp. 1807–1813, 2014.
[22] H. Li, X. Chen, B. Huang, D. Tang, and H. Chen, “High bandwidth visible light
communications based on a post-equalization circuit,” IEEE Photon. Technol.
Lett., vol. 26, no. 2, pp. 119–122, 2014.
[23] A. Azhar, T. Tran, and D. O’Brien, “A Gigabit/s indoor wireless transmission us-
ing MIMO-OFDM visible-light communications,” IEEE, Photon. Technol. Lett.,
vol. 25, no. 2, pp. 171–174, 2013.
[24] H. Elgala, R. Mesleh, and H. Haas, “Indoor optical wireless communication:

potential and state-of-the-art,” IEEE Commun. Mag., vol. 49, no. 9, pp. 56–62,
2011.
[25] J. Lian and M. Brandt-Pearce, “Multiuser multidetector indoor visible light com-
munication system,” in Opto Electron. and Commun. Conf., 2015.
[26] J. Lian, M. Noshad, and M. Brandt-Pearce, “Multiuser MISO indoor visible light
communications,” in 2014 Asilomar Conf. on Signals, Systems and Computers,
2014 (Invited paper), pp. 1729–1733.
122
[27] J. Lian and M. Brandt-Pearce, “Distributed power allocation for indoor visible
light communications,” in 2015 IEEE Global Commun. Conf. (GLOBECOM),
2015.
[28] ——, “Multiuser MIMO indoor visible light communication system using spatial
multiplexing,” Journal of Lightwave Technology, vol. 35, no. 23, pp. 5024–5033,
Dec 2017.
[29] ——, “Adaptive M-PAM for multiuser MISO indoor VLC systems,” in 2016
IEEE Global Communications Conference (GLOBECOM), Dec 2016, pp. 1–6.
[30] ——, “Joint optimal waveform design for multiuser VLC systems over ISI chan-
nel,” in 2016 IEEE International Conf. on Commun., May 2017, pp. 1–6.
[31] J. Lian, M. Noshad, and M. Brandt-Pearce, “M-PAM joint optimal waveform

design for multiuser VLC systems over ISI channel,” submitted to Journal of
Lightwave Technology.
[32] ——, “Comparison of DCO-OFDM and M-PAM for LED-based communication

system,” to be submitted to IEEE Commun. Lett.
[33] J. Lian, X. Wang, M. Noshad, and M. Brandt-Pearce, “Optical wireless inter-

ception vulnerability analysis of visible light communication system,” Submitted
to 2018 ICC.
[34] Z. Ghassemlooy, W. Popoola, and Rajbhandari, Optical Wireless Communica-

tions: System and Channel Modeling with Matlab. CRC Press, 2013.
[35] J. Kahn and J. Barry, “Wireless infrared communications,” Proceedings of the

IEEE, vol. 85, no. 2, pp. 265–298, Aug 1997.
123
[36] K. Lee and H. Park, “Channel model and modulation schemes for visible light
communications,” in 2011 IEEE 54th International Midwest Symposium on Cir-
cuits and Systems (MWSCAS), July 2011, pp. 1–4.
[37] K. Ying, H. Qian, R. J. Baxley, and S. Yao, “Joint optimization of precoder and
equalizer in MIMO VLC systems,” IEEE J. on Selected. Areas in Commun.,
vol. 33, no. 9, pp. 1949–1958, Sept 2015.
[38] L. Zeng, D. C. O’Brien, H. L. Minh, G. E. Faulkner, K. Lee, D. Jung, Y. Oh,

and E. T. Won, “High data rate multiple input multiple output (MIMO) optical
wireless communications using white LED lighting,” IEEE J. on Selected Areas
in Commun., vol. 27, no. 9, pp. 1654–1662, December 2009.
[39] M. Biagi, A. M. Vegni, S. Pergoloni, P. M. Butala, and T. D. C. Little, “Trace-

orthogonal PPM-space time block coding under rate constraints for visible light
communication,” J. Lightw. Technol., vol. 33, no. 2, pp. 481–494, Jan 2015.
[40] H. Ma, L. Lampe, and S. Hranilovic, “Coordinated broadcasting for multiuser

indoor visible light communication systems,” IEEE Trans. on Commun., vol. 63,
no. 9, pp. 3313–3324, Sept 2015.
[41] T. V. Pham, H. Le-Minh, and A. T. Pham, “Multi-user visible light communica-

tion broadcast channels with zero-forcing precoding,” IEEE Trans. on Commun.,
vol. 65, no. 6, pp. 2509–2521, June 2017.
[42] R. Feng, M. Dai, H. Wang, B. Chen, and X. Lin, “Linear precoding for multiuser
visible-light communication with field-of-view diversity,” IEEE Photon. J., vol. 8,
no. 2, pp. 1–8, April 2016.
124
[43] B. Lin, X. Tang, Z. Ghassemlooy, C. Lin, and Y. Li, “Experimental demonstra-

tion of an indoor VLC positioning system based on OFDMA,” IEEE Photon. J.,
vol. 9, no. 2, pp. 1–9, April 2017.
[44] H. Kazemi and H. Haas, “Downlink cooperation with fractional frequency reuse
in DCO-OFDMA optical attocell networks,” in 2016 IEEE International Con-
ference on Communications (ICC), May 2016, pp. 1–6.
[45] M. Guerra-Medina, O. Gonzalez, B. Rojas-Guillama, J. Martin-Gonzalez, F. Del-

gado, and J. Rabadan, “Ethernet-OCDMA system for multi-user visible light
communications,” Electron. Lett., vol. 48, no. 4, pp. 227–228, 2012.
[46] C. He, L. liang Yang, P. Xiao, and M. A. Imran, “DS-CDMA assisted visible light
communications systems,” in 2015 IEEE International Workshop on Computer
Aided Modelling and Design of Communication Links and Networks (CAMAD),
2015, pp. 27–32.
[47] S. Xie and C. Zhang, “Code division multiple access based visible light commu-
nication in vehicle adaptive cruise control under emergency situation,” in 2013
IEEE International Conf. on Info. and Auto., 2013, pp. 219–224.
[48] H. Ghafouri-Shiraz and M. M. Karbassian, Optical CDMA Network Principle,

Analysis and Applications. Wiley Press, 2012.
[49] I. E. S. of North America, IESNA lighting handbook, 9th ed.
[50] M. Noshad and M. Brandt-Pearce, “Hadamard-coded modulation for visible light

communications,” IEEE Transactions on Communications, vol. 64, no. 3, pp.
1167–1175, March 2016.
125
[51] J.-H. Jung, J. Lee, J.-H. Lee, Y.-H. Kim, and S.-C. Kim, “Ray-tracing-aided
modeling of user-shadowing effects in indoor wireless channels,” IEEE Trans.
Antennas and Propagation, vol. 62, no. 6, pp. 3412–3416, 2014.
[52] D. Tsonev, H. Chun, S. Rajbhandari, J. McKendry, S. Videv, E. Gu, M. Haji,

S. Watson, A. Kelly, G. Faulkner, M. Dawson, H. Haas, and D. O’Brien, “A 3
Gb/s single LED OFDM-based wireless VLC link using a gallium nitride µLED,”
IEEE Photon. Technol. Lett., vol. 26, no. 7, pp. 637–640, 2014.
[53] J. Lian and M. Brandt-Pearce, “Adaptive M-PAM for multiuser MISO indoor
VLC systems,” in 2016 IEEE Global Commun. Conf. (GLOBECOM), 2016, pp.
1–7.
[54] Z. Zheng, T. Chen, L. Liu, and W. Hu, “Experimental demonstration of femtocell

visible light communication system employing code division multiple access,” in
Optical Fiber Commun. Conf. (OFC), 2015, pp. 1–3.
[55] Y. Chen, Y. Chang, Y. Tseng, and W. Chen, “A framework for simultaneous

message broadcasting using CDMA-based visible light communications,” IEEE
Sensors J., vol. 15, no. 12, pp. 6819–6827, 2015.
[56] Q. Wang, Z. Wang, and L. Dai, “Multiuser MIMO-OFDM for visible light com-
munications,” IEEE Photon. J., vol. 7, no. 6, pp. 1–11, 2015.
[57] J. Grubor, S. Randel, K.-D. Langer, and J. Walewski, “Broadband information

broadcasting using LED-based interior lighting,” J. Lightw. Technol., vol. 26,
no. 24, pp. 3883–3892, 2008.
[58] B. Li, J. Wang, R. Zhang, H. Shen, C. Zhao, and L. Hanzo, “Multiuser MISO
transceiver design for indoor downlink visible light communication under per-
LED optical power constraints,” IEEE Photon. J., vol. 7, no. 4, pp. 1–15, 2015.
126
[59] L. Wu, Z. Zhang, J. Dang, and H. Liu, “Adaptive modulation schemes for visible
light communications,” J. Lightw. Technol., vol. 33, no. 1, pp. 117–125, 2015.
[60] M. Biagi, A. M. Vegni, S. Pergoloni, P. M. Butala, and T. D. C. Little, “Trace-

orthogonal PPM-space time block coding under rate constraints for visible light
communication,” J. Lightw. Technol., vol. 33, no. 2, pp. 481–494, 2015.
[61] T. Mizuochi, Y. Miyata, T. Kobayashi, K. Ouchi, K. Kuno, K. Kubo, K. Shimizu,

H. Tagami, H. Yoshida, H. Fujita, M. Akita, and K. Motoshima, “Forward error
correction based on block turbo code with 3-bit soft decision for 10-Gb/s opti-
cal communication systems,” IEEE J. of Selected Topics in Quantum Electron.,
vol. 10, no. 2, pp. 376–386, 2004.
[62] Z. Chen and H. Haas, “Space division multiple access in visible light commu-
nications,” in 2015 IEEE International Conf. on Commun. (ICC), 2015, pp.
5115–5119.
[63] K. Lee, H. Park, and J. R. Barry, “Indoor channel characteristics for visible light
communications,” IEEE Commun. Lett., vol. 15, no. 2, pp. 217–219, February
2011.
[64] X. Huang, J. Shi, J. Li, Y. Wang, and N. Chi, “A Gb/s VLC transmission using
hardware preequalization circuit,” IEEE Photon. Technol. Lett.,, vol. 27, no. 18,
pp. 1915–1918, 2015.
[65] Y. Wang, L. Tao, Y. Wang, and N. Chi, “High speed WDM VLC system based on
multi-band CAP64 with weighted pre-equalization and modified CMMA based
post-equalization,” IEEE Commun. Lett.,, vol. 18, no. 10, pp. 1719–1722, 2014.
127
[66] A. Schmidt, R. Schober, and W. Gerstacker, “Single-carrier frequency-division

multiple access transmission with physical layer network coding over ISI chan-
nels,” in 2015 European Wireless Conf., May 2015, pp. 1–7.
[67] S. H. Song and K. B. Letaief, “Diversity analysis for linear equalizers over ISI
channels,” IEEE Trans. on Commun., vol. 59, no. 9, pp. 2414–2423, September
2011.
[68] D. Lee and J. Kahn, “Coding and equalization for PPM on wireless infrared
channels,” in 1998 IEEE Global Commun. Conf. (GLOBECOM), 1998, pp. 201–
206.
[69] M. Zhang, M. Shi, F. Wang, J. Zhao, Y. Zhou, Z. Wang, and N. Chi, “4.05-gb/s
RGB LED-based VLC system utilizing PS-manchester coded nyquist PAM-8
modulation and hybrid time-frequency domain equalization,” in 2017 Optical
Fiber Commun. Conf. and Exhibition (OFC), March 2017, pp. 1–3.
[70] D. J. F. Barros, S. K. Wilson, and J. M. Kahn, “Comparison of orthogonal

frequency-division multiplexing and pulse-amplitude modulation in indoor opti-
cal wireless links,” IEEE Transactions on Communications, vol. 60, no. 1, pp.
153–163, January 2012.
[71] J. Du, W. Xu, H. Zhang, and C. Zhao, “Visible light communications using
spatial summing PAM with LED array,” in 2017 IEEE Wireless Communications
and Networking Conference (WCNC), March 2017, pp. 1–6.
[72] J. M. Luna-Rivera, R. Perez-Jimenez, J. A. Rabadan-Borjes, J. F. Rufo-Torres,

V. Guerra, and C. Suarez-Rodriguez, “Multiuser scheme for indoor visible light
communications using RGB LEDs,” in 2014 International Work Conf. Bio-
inspired Intel. (IWOBI), 2014, pp. 119–123.
128
[73] M. Biagi, A. M. Vegni, and T. D. C. Little, “LAT indoor MIMO-VLC-localize,

access and transmit,” in 2012 International Workshop on Optical Wireless Com-
mun. (IWOW), 2012, pp. 1–3.
[74] Y. Zhou, S. Liang, S. Chen, X. Huang, and N. Chi, “2.08 Gbit/s visible light com-
munication utilizing power exponential pre-equalization,” in 2016 25th Wireless
and Optical Communication Conference (WOCC),, 2016, pp. 1–3.
[75] K. Ying, H. Qian, R. J. Baxley, and S. Yao, “Joint optimization of precoder and
equalizer in MIMO VLC systems,” IEEE J. Selec. Areas in Commun., vol. 33,
no. 9, pp. 1949–1958, Sept 2015.
[76] K. H. Park, Y. C. Ko, and M. S. Alouini, “On the power and offset allocation
for rate adaptation of spatial multiplexing in optical wireless MIMO channels,”
IEEE Trans. on Commun., vol. 61, no. 4, pp. 1535–1543, April 2013.
[77] K. Deb, A. Pratap, S. Agarwal, and T. Meyarivan, “A fast and elitist mul-
tiobjective genetic algorithm: NSGA-II,” IEEE Transactions on Evolutionary
Computation, vol. 6, no. 2, pp. 182–197, Apr 2002.
[78] J. Armstrong, “OFDM for optical communications,” Journal of Lightwave Tech-

nology, vol. 27, no. 3, pp. 189–204, Feb 2009.
[79] J. Classen, J. Chen, D. Steinmetzer, M. Hollick, and E. Knightly, “The spy

next door: Eavesdropping on high throughput visible light communications,” in
Proceedings of the 2nd International Workshop on Visible Light Communications
Systems. ACM, Sep 2015, pp. 9–14.
[80] A. Mostafa and L. Lampe, “Physical-layer security for indoor visible light
communications,” in 2014 IEEE International Conference on Communications
(ICC), June 2014, pp. 3342–3347.
129
[81] ——, “Pattern synthesis of massive LED arrays for secure visible light com-
munication links,” in 2015 IEEE International Conference on Communication
Workshop (ICCW), June 2015, pp. 1350–1355.
[82] D. Li, L. Zhang, and J. Qiu, “High security chaotic multiple access scheme
for VLC systems,” in 2016 26th International Telecommunication Networks and
Applications Conference (ITNAC), Dec 2016, pp. 133–135.
[83] G. Pan, J. Ye, and Z. Ding, “On secure VLC systems with spatially random
terminals,” IEEE Communications Letters, vol. 21, no. 3, pp. 492–495, March
2017.
[84] F. J. Lopez-Martinez, G. Gomez, and J. M. Garrido-Balsells, “Physical-layer

security in free-space optical communications,” IEEE Photonics Journal, vol. 7,
no. 2, pp. 1–14, April 2015.

Multiuser MIMO Indoor Visible Light Communications: University of Virginia

Uploaded by

Copyright:

Available Formats

Multiuser MIMO Indoor Visible Light Communications: University of Virginia

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Multiuser MIMO Indoor Visible Light Communications: University of Virginia

Uploaded by

Copyright:

Available Formats

Multiuser MIMO Indoor Visible Light

Jie Lian, Author

Prof. Maite Brandt-Pearce, PhD Advisor

Prof. Stephen G. Wilson , Committee Chair

Prof. Daniel S. Weller

Prof. Malathi Veeraraghavan

Prof. Stephen D. Patek

Accepted for the School of Engineering and Applied Science:

Dean, School of Engineering and Applied Science

Visible light communications (VLC) is an energy efficient and cost-effective solu-

some practical considerations, such as shadowing effects, illumination requirements,

3 Power Allocation Algorithms for Multiuser VLC Systems 20

4 Modulation Schemes for VLC Systems 56

4.2.4 Numerical Results and Discussions . . . . . . . . . . . . . . . 85

5 Interception Vulnerability Analysis 102

6 Conclusions and Future Work 114

1.1 Applications for indoor VLC systems. . . . . . . . . . . . . . . . . . . 3

4.2 OCDMA and TDMA comparison using M-PAM, K < L case (L = 7

4.17 Diagram of DCO-OFDM with adjustable modulation index and loaded

5.1 Vulnerable area outside the room . . . . . . . . . . . . . . . . . . . . 103

3.1 Parameters Used for Indoor Environment . . . . . . . . . . . . . . . . 27

4.1 Parameters Used for Large Indoor Environment . . . . . . . . . . . . 66

5.1 Parameters Used in Numerical Results . . . . . . . . . . . . . . . . . 108

APD: avalanche photo diode

M-VPPM: M-ary variable pulse position modulation

WDM-PAJO: weighted distributed multi-detector PAJO

ĥqk : channel gain between LED q and user k

NQ : number of total LEDs

%/Nsub : modulation index for OFDM

Figure 1.1: Applications for indoor VLC systems.

1.2 Differences Between VLC and RF Communi-

of these differences is presented.

The conventional modulation schemes adopted in RF communications cannot be

Flicker and Dimming

the signal exceed the constraint, it must be clipped.

1.3 Modulation Schemes for VLC

1.3.1 On-Off Keying

1.3.2 M -ary Pulse Amplitude Modulation

Figure 1.2: 4-PAM modulation

the transmitter just needs to send the corresponding power levels.

1.3.3 Pulse Position Modulation

Pulse position modulation (PPM) that is illustrated in Fig. 1.3 is an orthogonal

1.3.4 Orthogonal Frequency Division Multiplexing

Orthogonal frequency-division multiplexing (OFDM) is a method of modulating

Figure 1.3: Time waveforms for 16-PPM and OOK

Figure 1.4: Orthogonal subcarriers in OFDM.

1.4 Multiple Access for VLC

1.4.1 Time Division Multiple Access

Time division multiple access (TDMA) is a conventional multiple access scheme

TDMA User 1 User 2 …… User K User 1 User 2 ……

Figure 1.6: An OOC code with the length of 7.

1.4.2 Optical Code Division Multiple Access

1.4.3 Orthogonal Frequency Division Multiple Access

Orthogonal frequency-division multiple access (OFDMA) is a novel multiple access

User 1 Data Subcarrier User 2 Data Subcarrier

Figure 1.7: Subcarriers for multiple users in OFDMA.

1.5 Literature Review and Dissertation Contribu-

• Multi-LED lamp and multi-detector receiver models are proposed to support

• Impulse response of indoor VLC channel and received power distributions in