Content uploaded by Alain Richard Ndjiongue
Author content
All content in this area was uploaded by Alain Richard Ndjiongue on Oct 26, 2017
Content may be subject to copyright.
Visible Light Communications (VLC) Technology
A. R. Ndjiongue†, H. C. Ferreira†and T. M. N. Ngatched‡
†Department of Electrical and Electronic Engineering Science,
University of Johannesburg,
P.O. Box 524, Auckland Park, 2006,
Johannesburg, South Africa.
‡Faculty of Engineering and Applied Science,
Memorial University,
St. John’s, NL A1B 3X5,
Newfoundland, Canada
Emails: {arrichard,hcferreira}@uj.ac.za, tngatched@grenfell.mun.ca
Abstract
This article reviews the visible light communications (VLC) technology. It presents
the VLC communication system: the transmitter, the channel, and the receiver. The
single and multi-channel transceivers are presented. The channel for a system that uses
a single light emitting diodes (LED) and the matrix representing the multi-colour chan-
nel are discussed. Various modulation schemes are reviewed. Basic techniques used to
implement a VLC system are highlighted and different causes of impairment are un-
derlined. A word is given on the standardisation of the VLC technology. Numerous
applications of VLC technology are given. Challenges for VLC practical implementation
and commercialisation are reviewed. Advances in VLC and future research issues are
discussed.
Index terms: VLC technology, VLC, Modulation methods, Dimming control, MIMO, Optical wireless
communications.
1 Introduction
Visible light communications (VLC) is a communication technology in which the visible spectrum is
modulated to transmit data. Due to the propagation distance of the light emitting diodes (LEDs), VLC
is a short-range communication technology.
In the electromagnetic spectrum, the visible spectrum covers between 350 nm and 800 nm of wave-
length and the frequencies are comprised between 4.3 x 1014 Hz and 7.5 x 1014 Hz. In VLC technology,
LEDs are used because their currents intensity are easily modulated, with respect to their counterparts
i.e. incandescent and fluorescent light bulbs. LEDs are based on a doping process, consequently their
efficiency and their durability are improved, and they have longer lifetime in comparison to the incan-
descent and fluorescent light bulbs [1]. In any lighting application (general lighting, signage, displays,
vehicles’ lights to mention only a few), it is predicted that LEDs are going to overtake the usual light
bulbs. They are going to provide double applications, namely lighting and communication. As in the
case of any communication technology, transmission in VLC technology is generally characterised by
a transmission matrix, which is a mathematical representation of the channel impulse response. The
size of this matrix varies with the number of groups of LEDs and with the number of LEDs per group.
With multiple blocks of multiple LEDs, very high data rate transmission can be performed. When the
transmission is corrupted by noise and interferences from unwanted sources, equalization techniques for
channel pre-compensation, knowing the channel behaviour, help to recover the symbols. VLC technology
faces many implementation challenges: some are related to the communication system design and others
are related to the practical implementation. To properly implement a VLC communication system, some
1
Wiley Encyclopedia of Electrical and Electronics Engineering, Published Online: 15 JUN 2015, DOI: 10.1002/047134608X.W8267,
Copyright © 2000 John Wiley & Sons, Inc. All rights reserved.
constraints have to be met: the lighting constraints related to the average optical power and the commu-
nication objective related to the throughput. During transmission, LED flickering is to be avoided and
under dimming conditions, the data rate has to be reduced considerably [1], [2] and [3].
VLC technology has been around for a while. The history started in 1880, when Alexander Graham
Bell invented the photophone [4]. This instrument was used to transmit speech by modulating the
sunlight. In the 1960s, optical communications were born. Light amplification by stimulated emission of
radiation (LASER) and light emitting diodes (LEDs) were invented [5]. Later, in 2003, recent work began
on VLC technology. Natagawa Laboratory, in Keio University, Japan, used LEDs to transmit data. In
2006, the center for information communication technology research (CICTR), Pen State, USA, proposed
the first combination of power line communications (PLC) and white LED to provide broadband access
for indoor applications. Since then, there have been numerous research activities on VLC. Among them,
light fidelity (Li-Fi), founded by Harald Haas from the University of Edinburgh in the United Kingdom
[4], is one of the most interesting achievement of VLC for the several years.
This article is organised as follows: The structure of the VLC transceiver, the transfer matrix, the
channel response and the signal-to-noise ratio (SNR) are presented in Section 2. Section 3 presents
the latest suggestions of the standardisation organisations (SDOs) on VLC technology. The techniques
proposed by the IEEE 802.15.7 standards are highlighted and the efforts of the Japanese organisation on
VLC technology named visible light communications consortium (VLCC) are presented. In Section 4,
the VLC modulation methods are detailed. Due to the fact that VLC technology is characterised by a
high SNR [3], techniques applied to systems with high SNR work in VLC systems. In this section, we
present some examples. Pulse position modulation (PPM) and on-off keying (OOK), suitably used for
low and medium data rate applications, are presented. Some complex modulation such as the orthogonal
frequency division multiplexing (OFDM) and color shift keying (CSK), deployed for high data rate
communication systems, are also presented. Section 4 also highlights the use of the spatial diversity in
VLC technology. Section 5 focuses on the challenges of VLC practical implementation. Dimming and
LEDs control methods are presented for the above-mentioned modulation techniques. This is followed
by the potential applications of VLC technology presented in Section 6. In Section 7, the article presents
the challenges of the commercialisation of VLC technology. Followed by the recent advances in VLC
technology in Section 8. Section 9 looks at the unsolved problems in VLC and discusses future directions.
2 VLC communication systems
Figure 1: Basic transmission blocks in a VLC systems
VLC technology is part of the set of optical wireless communications (OWC). Hence the physical optical
principles can be applied to the VLC systems. In fact, the carrier in VLC is the visible rays used for
illumination. VLC is typically characterised by a non negative and non-coherent signal transmission. It
respects the communication principle in which three main parts are considered: a transmitter, a channel
and a receiver. Fig. 1 shows the basic blocks of a VLC transmission system. It is made of the transmitter,
the channel and the receiver, and , for a system corrupted by the additive white Gaussian noise (AWGN),
the transmission is always governed by
ri=Hsi+ωi,(1)
2
where riand siare the received and the transmitted sets of symbols respectively, His the channel
response and ωithe channel noise. A suitable model for VLC communication systems is depicted in
Fig. 2. It shows two electrical domains and one optical domain. The modulated signal, added to a DC
voltage is used to power the LED, this constitutes the transmitter. The LED in its operation produces
the light and at the same time, convey the information through the channel. The receiver is made of the
the photodetector (PD) and the demodulator. The PD detects the light and produce an electrical signal
composed of the message plus noise. Part of the noise here is produced by the channel even though in
the model, we represent the total noise in the electrical domain. This is due to the fact that the PD
converts both the message and the optical noise into an electrical current.
Figure 2: Model of VLC communication system
In most modulation schemes highlighted in this article, the model exploited does not inject any DC
offset. It uses the message signal to power the LED. nevertheless, in the case of frequency shift keying
(FSK) and OFDM techniques for example, the model in Fig. 2 is well indicated to be used.
2.1 The VLC transmitter
In VLC systems, the transmitter groups, in one module, the data source, the modulation module and
the LED. The last two elements are the very important elements in a VLC transmitter. Two types of
LEDs are used in VLC systems: The single-colour LED and the multicolour LEDs. The multicolour LED
groups in one package multiple single-colour LEDs. The most used multicolour LED is the red-green-blue
(RGB) LED. In multi-carrier systems, each of the colour LEDs included in the package represents an
antenna, corresponding to one channel. There are as many channels in the system as there are LEDs in
the package. Hence, a given number of colour-LEDs will provide the same number of distinct channels.
Consequently, the RGB-LED transmitter is seen as a special multichannel transmitter that can be used
to deploy multicarrier modulation techniques. For example, with a single RGB-LED, a three-by-three
multiple inputs-multiple outputs (3×3 MIMO) technique is applicable over the VLC channel [6], [7]. Fig. 3
depicts the two common types of VLC transmitter: Fig. 3-a, a single VLC transmitter and Fig. 3-b, a 3
channel VLC transmitter.
Figure 3: VLC transmitters: a) the single LED transmitter, b) a 3 channels VLC transmitter
3
2.2 The VLC channel
In communication, the channel represents the space between the transmitter and the receiver. It is
characterised by its ability to transmit the carrier signal, and, it is influenced by many factors such as
attenuation, interference and noise. In VLC technology, the channel is the space between the LED and
the PD. It is mathematically represented by its transfer function H(see (1)). Two main types of channels
are considered in VLC communication systems: the single VLC channel involving a single LED and a
single PD, and the multichannel VLC systems in which the transmitter is made of multicolour LEDs.
In this second case, the PD is made of more than one detector, each of them being sensitive to a colour
from the transmitter.
2.2.1 A single VLC channel (single input-single output system (SISO))
In single VLC channel, one LED and one PD are used to achieve transmission. The capacity CSISO of
the transmission link is given by [8]
CSI SO =log2(1 + g2Pt
σ2B),(2)
where Pt, independent of the illumination, denotes the transmitter power, Bthe transmission bandwidth,
σ2the variance of the total noise in an AWGN channel, and gthe channel gain. The quantity g2Pt/σ2B
represents the SNR characterising the channel. The distribution link is organised in two different types:
a line-of-sight (LOS, direct and non-direct) link, or a non-line-of-sight (NLOS) link.
•LOS VLC link
Figure 4: Non-direct line-of-sight (ndLOS) distribution of a single ray
In LOS link, there is a straight link without obstacle between the LED and the PD (see Fig. 4). We
distinguish the direct LOS (dLOS) in which the LED is linked to the PD with 0oincidence (ϕ= 0), and
the non-direct LOS (ndLOS) in which the incidence is not null (ϕ6= 0). dLOS and ndLOS links are
similar in terms of model. The LOS system has been studied and more information can be found in the
literature in [9], [10], [11] and [12]. In LOS VLC link, (1) becomes
ri=HLOS si+ωi,(3)
where HLOS is the LOS channel response. This model (3) was used in [13] to describe the VLC transmis-
sion system. The diffuse link model of a LOS VLC transmission is represented in Fig. 4. The bandwidth
in this situation can be determined by the summation of the LOS and diffuse component of the received
signal [14]. The transmission gain g(LOS), studied and presented in [11], [13], [14] and [15], is given by
g(LOS)= [ (ξ+ 1)A
2πd2].cosξ(ϕ).Tf(α).g(α).cos(α),(4)
4
where the incidence angle ϕis given by 0 ≤ϕ≤Φm,Tf(ϕ) is the transmission filter and g(α) the the
concentration gain. drepresents the minimum distance between the LED and the PD. It is to be noted
that g(LOS)is null for ϕ > Φm[8]. In [16], the VLC channel is detailed with more distribution options
and different situations is evaluated to characterise the transmission environment. A difference between
the direct LOS and the non-direct LOS is underlined. The channel is modelled as proposed in [13], taking
into account the direct link between the transmitter and the receiver, including the reflection paths as
presented in [17], [18].
•NLOS VLC link
In NLOS VLC system, the light rays from the LED reach the PD after single or multiple reflections, this
is due to an obstacle between the sender and the receiver. In a typical NLOS link between sender and
receiver, the channel impulse response is seen as an infinite sum of light rays after many reflections [19],
[20], and can be expressed by
HNLOS =
∞
X
k=0
h(k),(5)
where h(k)is the impulse response of rays undergoing the k(th)path. But this equation can be rearranged
by subdividing the indoor environment into a finite number of portions. The transmission is characterised
in this case by a transmission equation using the LOS transfer matrix multiplied by a coefficient ρ
characterising the NLOS link [16]. For a NLOS link, (1) becomes
ri=HN LOS si+ωi=ρHLOS si+ωi.(6)
2.2.2 Multi-channel VLC systems
Multi-carrier communication systems can be implemented over the VLC channel by using more than one
colour LED to inject the message signal to the channel. In this situation, we have finite numbers nand
zof LEDs and PDs used as antenna and detectors respectively. ncan be divided by the number mof
groups of LEDs to obtain the number of LEDs per group. The transfer matrix in multi-carrier VLC
systems is given by
Hmulti =
h1,1h1,2... h1,n
h2,1h2,2... h2,n
. . ... .
. . ... .
. . ... .
hz,1hz,2... hz ,n
,(7)
where the entries hi,i represent the front-end gain between the ith LED and the corresponding PD, and
hi,j represent the crosstalk gain between the ith LED and the jth PD. If there is no crosstalk, Hmulti
becomes a diagonal matrix with hi,i entries. The channel capacity Cmulti in multi-carrier VLC is given
by
Cmulti = ΓCCI SO (8)
where Γ = min(n, z) and CC ISO is given in (2). RGB-LEDs being the most used multi-wavelength LEDs,
the channel transfer matrix H3×3in the case of a single RGB-LED transmitter is given by
H3×3=
hrr hrg hrb
hgr hgg hgb
hbr hbg hbb
,(9)
where the diagonal entries (hrr,hbb , and hg g) represent the LOS link between a single LED and its
corresponding PD, and the rest of entries (hrg ,hrb ,hgr ,hgb ,hbr ,hbg) represents cross-talks between
channels.
5
2.3 The VLC receiver
The main element in the VLC receiver is the photo-detector used to collect the light radiation [14]. Two
main types of photodetectors are used in VLC receivers: the photo-diode and the phototransistors. The
digital camera, consisting of an array of photo transistor is a good device for receiving VLC signal in
smart devices such as smart phones and laptops [21]. As described in [13], a complete receiver system
made of components such as the concentrator, the optical filter, the amplifier and the equaliser, necessary
to capture the maximum light needed to convert the received signal into message. The rays pass through
the concentrator and the optical filter before they reach the proper detector core. The architecture of a
VLC receiver is presented in Fig. 5.
Figure 5: Architecture of a VLC receiver
2.3.1 Optical received power
Considering the instantaneous power of the LED, pLED(t), the average optical power produced by a
single LED is given by [16], [19]
PLED = lim
T→∞
1
2T[ZT
−T
pLED (t)dt].(10)
The received power PPD is calculated using the transmission gain (g). In the case of a LOS VLC
communication system, PP D is given by [21]
PP D =PLE D ×g(LOS),(11)
where g(LOS)is the transmission gain defined in (4). This power depends on the wavelength as depicted
in [13]. But three other factors are also taken into account: the filter gain Tf(ϕ), the concentration gains
g(α) and the Lambertian distribution order (ξ) (see equation 4).
2.3.2 Distribution of Light from LEDs
With the rapid development of solid state lighting technologies, LEDs are designed to generate 10-120
lumen each with very good efficiency. Indoor illumination demands about 400 to 1200 luxes in a single
room. A single LED is not enough, meaning an array of LEDs is required, which is an advantage for the
uniformity of the illumination required for a comfortable visual impression. The ideal LED optical model
is a perfect Lambertian, meaning that the intensity of the propagation is proportional to the cosine of
the viewing angle. But in the real world, some LEDs can be represented by an imperfect Lambertian.
We distinguish the far field illumination areas, in which the LED illumination range is about 5 times
larger than its maximum size. The scattering produced by many light sources increases the number
of paths in the VLC channel. As the number of packages (pa) increases, the number (β) of resolvable
paths increases according to the relation β=αpa, where αis the number of LEDs per group. Hence,
increasing the number of LEDs enhances the uniformity of the illumination an makes the implementation
more complex in indoor environment. In this case, an efficient equalization technique is required to
overcome the effects of delay spread owing to the multipath effects. Generally, they are all manufactured
with the Lambertian emission principle. The link transmitter-receiver in VLC is then based on this
principle. Fig. 6 shows the distribution of light from LED. A Lambertian distribution is characterised by
6
its Lambertian order (ξ). This distribution confers to the VLC channel a multipath environment. The
Lambertian order ξis given by [17]
ξ=−ln2
ln(Θ0.5),(12)
where Θ0.5represents the semi-angle corresponding to half the received optical power (Pr/2). An impor-
tant characteristic of this distribution is the radiant intensity κ(ϕ) of the Lambertian transmitter. It is
used to define the channel gain given in (4). κ(ϕ) is given for a SISO channels by [8]
κ(ϕ) = ξ+ 1
2πcosξ(ϕ).(13)
In [22], a multiple inputs - single output system (MISO) version of the radiant intesity of the Lambertian
transmitter related to the number nof inputs is studied, taking into account the number of transmitting
antennas. This extends the VLC link proposed for a single user to a multi-user scenario.
Figure 6: Scattering of the light from LEDs (Lambertian distribution); (A−A′) represents the Lamber-
tian surface in the profile view
2.3.3 Other parameters affecting the received signal
Figure 7: Factors influencing the VLC transmission link
In the design of the light distribution by LEDs, some other parameters to be thoroughly analysed are:
the luminous spatial intensity distribution, the optical power and the spectral density. In addition to
the influence of the type of distribution, many parameters are likely to affect the communication link.
These are: the illumination related to the current intensity flowing in the LEDs, the number of sources
7
in the LEDs array (uniform illumination and spacing between sources), the external sources of light
(interference), the order of the Lamtertian emission ξ, and the nature of the wall surface (multi path
scenario and resolvable path) [18]. Hence, the bit error rate (BER), the symbol error rate (SER) and the
signal to noise ratio (SNR) are functions of these parameters. This has previously been analysed by the
authors and summarised in Fig. 7.
2.4 Noise and SNR in VLC systems
2.4.1 Noise
In VLC systems, the noise sources include sunlight, incandescent and fluorescent light in indoor or outdoor
environments [9]. But in addition to the noise present in the environment, the receiver in its operation
produces noise. This is due to the impact of the photons on the surface of the receiver. Two main
types of noise source are inventoried: shot and thermal noise. Thermal noise is an energy equilibrium
fluctuation phenomenon and shot noise produced by current fluctuations, refers to the random nature of
photon absorption and electron hole recombination. It can be modelled using Poisson distribution and
it is white noise [23]. Additionally to the current generated from the photons of the light, there is a
component called dark current, which influences the variance of the shot noise (σ2
shot =σ2
light +σ2
dark ).
It is proved that the variance of the shot noise has two components related to the direct and the ambient
light. This is detailed in [24] where the variance of shot noise is defined as a function of the electronic
charge (q), the received average optical power (Pr), the noise bandwidth (B), the background current
(Ibg), the responsivity Rrof the photo receiver and a factor related to the noise bandwidth (ζ). The
thermal noise is also defined as depending on factors such as the Boltzmann constant (K), the absolute
temperature (Tk), the detector area (A), the open loop voltage gain (G) and the field effect transistor
channel noise factor (η). But the fluctuations in a current across a semi-conductor due to shot noise can
be distributed using Fourier transform as [25]
In(t) = X
k=1
ik(t), k ∈N(14)
Practically, it is very difficult to separate shot noise and thermal noise. But the variance σ2
nof the total
noise is equal to the sum of the variances (σ2
shot) of the shot noise and (σ2
th) of the thermal noise (15).
σ2
n=σ2
shot +σ2
th (15)
σ2
shot and σ2
th are defined in [24], [26] as
σ2
shot =B[2qIbg ζ+ 2qPrγRr] (16)
and
σ2
th =8πKTk
GηAζB2+16πKTkΓ
gm
η2A2ζthB3.(17)
2.4.2 Signal to noise ratio
The SNR is a parameter defining the strength of the signal carrying the information and comparing it
to that of unwanted signal. In VLC system, since the noise vector appears as a composition of the shot
and the thermal noise, in most research, it is assumed that the total noise is dominated by the white
Gaussian component [24]. Hence, the SNR is defined by [9]
SN R =R2
rP2
r
σ2
shot +σ2
th
,(18)
But considering the case of a system with inter-symbol interference (ISI) from multi-path propagation,
the received noise power due to ISI will affect the SNR. Under the same condition for which (18) is given,
the SNR in ISI systems is also given by [27]
SN R =R2
rP2
r
σ2
shot +σ2
th +R2
rP2
r,ISI
.(19)
8
2.4.3 Optical interference noise
One cannot study noise over the VLC channel without mentioning the optical interference that occurs
on the optical link between the LEDs and the PD. It is to be noted that in the environment of the trans-
mission, there is a background noise sourced by the sunlight. This background noise can be accompanied
by the incandescent and the fluorescent noise. This will induce a background current in the PD. But
that background current was taken into account when formulating the variance of the shot noise given
in (16).
3 Standardisation of VLC
To regulate transmission in VLC technology, the Institute of Electrical and Electronics Engineers (IEEE)
working group IEEE 802.15.7, proposes schemes and techniques. The IEEE 802.15.7’ standard devises
the physical layer (PHY) of VLC technology in three parts: PHY I, PHY II and PHY III [3], [28]. Specific
modulation schemes and coding techniques are dedicated to each of these layers. PHY I operates from
11.67 kb/s to 266.6 kb/s, PHY II operates from 1.25 Mb/s to 96 Mb/s, and PHY III operates from 12 Mb/s
to 96 Mb/s [28]. PHY III, dedicated to multiple optical sources using CSK, was developed and presented
in [29]. PHY I and PHY II use schemes such as OOK and VPPM. Other standardisation organisations
do exist in VLC. In Japan, the visible light communication consortium (VLCC) provides a collaborative
platform for researchers, universities and industries, for improving the VLC technology. The VLCC
membership includes the following: Nippon Electric Company (NEC) corporation, Panasonic, Toshiba
corporation, Samsung Electronics, Casio Computer, Nakagawa Laboratories, and Sharp corporation.
The activities of the VLCC consortium are to develop standards for VLC technology. VLCC proposes
to use VPPM to implement communication systems in VLC technology. In Europe, the Wireless World
Research Forum (WWRF) also works on VLC technology. Its working group 5 is in charge of investigating
the VLC environment. Other organisations such as the Telecommunications Technology Association in
South Korea also looks at VLC technology.
4 Modulations methods
The deployment of a modulation technique depends on the application of the system to be designed.
Throughput, received signal quality, and the required channel capacity are the main parameters influ-
encing the choice of a modulation technique. Various modulation techniques are available to be deployed
over the VLC channel [30], [31], [32]. They are organised regarding the number of channels (carri-
ers). For example, CSK works with more than one carrier. OFDM and sub-carrier index modulation
(SIM-OFDM) can work with single or multiple LEDs [33], [34]. Dual-header pulse interval modulation
(DH-PIM) and sub-carrier PSK intensity modulation are also suitable modulation techniques for VLC
system implementation. From a standardization point of view, VLCC recommends the sub-carrier pulse
position modulation (SC-PPM) because of its robustness in avoiding DC [35]. OOK, VPPM CSK are
proposed in IEEE 802.15.7 [30],[28].
In this section, some modulation schemes are briefly studied, including OOK, VPPM. Some complex
modulations (CSK, OFDM, Spatial diversity) are also mentioned.
4.1 Pulse Position Modulation (PPM)
PPM is a scheme in which the position of the pulse, relative to its unmodulated time of occurrence,
is varied in accordance to the message signal. It was designed to reduce the energy waste occurring
during transmission in pulse duration modulation (PDM) or pulse width modulation (PWM). In these
modulation schemes (PDM), (PWM) and PPM, samples of message signals are used to vary the duration
of the individual pulse.
4.1.1 VPPM
In the literature [13], many variants of the PPM are mentioned, including level pulse-position modulation
(LPPM), differential PPM (DPPM), and variable PPM (VPPM). VPPM is needed in VLC technology
9
because it simultaneously supports illumination, dimming control and communication. In a multipath
dispersion channel, the VPPM signal can be written as [36]
lvppm(t) =
+∞
X
i=−∞
sppm(t−iTs),(20)
where Tsis the total duration of a symbol transmission including the LOS duration and the duration of
the extra time caused by the NLOS path. sppm(t) is defined as the modulated signal depending on the
bit value {0,1}and is used for controlling the LED illumination. It is expressed as
sppm(t) = fi(t)pEb̺, (21)
where i={0,1}and fi(t) is the basic function providing dimming control of the LEDs. ̺represents the
dimming level and is defined in a percentage between 0 and 100 (0 ≤̺≤100 %). To insure constant
balanced lighting between ones and zeros, a 50 percent duty cycle is applied for both ones and zeros
[30]. Let l′(t) be the light emitted by the LEDs, l′(t) depends on the electrical signal lvppm (t). l′(t) is
subject to the convolution by the optical response of the channel ho(t). Then the transmission equation
is expressed as
re(t) = l′(t)∗ho(t) + ωe(t),(22)
where l′(t) is related to lvppm(t) by l′(t) = Υlvppm(t), Υ being the total electrical to optical conversion
factor. ωerepresents the electrical AWGN at the output of the photodetector and rerepresents the
electrical received signal from the photodetector. he(t) is rearranged from (5) taking into account the
effects of multipath dispersion. he(t) is given by
he(t) =
M−1
X
i=1
∆ie(−τi(t−di)),(23)
where Mis the number of single clusters, direpresents the delay in the ith path, and ∆iand τiare the
channel gain and the time constant of the ith cluster respectively. In terms of performance of the VPPM,
the correlation factor between AWGN noise over the bits 0 and 1 requires careful handling because it
affects the error probability of the received symbol.
4.2 On-Off Keying (OOK)
OOK is a very low complexity modulation scheme. It is a special case of amplitude shift keying (ASK)
employing two voltage levels where the second one is zero. The carrier signal in OOK is given by [37],
[38]:
look(t) =
+∞
X
i=−∞
v[i]sook(t−iTb) (24)
In case of consecutive ones or zeros, OOK suffers from unbalanced intensity provided to the LED. Two
methods are available to solve the unbalance. The first method consists of redefining the voltage level
by keeping them identical and playing with the duration of the pulse. This method gives constant data
rate as the light changes, even though in the case of colour LEDs, the shifting of the colours is expected
[30]. It also wastes energy since the control is now based on the variation of the width of the pulse. The
second technique is base on introducing a compensation time into the wave form. In this case, the data
rate is automatically reduced because of dimming.
4.3 Frequency Shift Keying (FSK)
FSK-VLC is non-standardised scheme. This is due to the non-negative aspect of the VLC technology.
The introduction of a DC offset (see Fig. 2) automatically solves the problem. However, FSK-VLC is
still to be properly investigated. In FSK, symbols are mapped to frequencies. The authors initiated the
analysis of the FSK transmission principle for VLC technology (the analysis is still under-way). They
look at the conditions to perform a communication under a constant average transmit optical power.
10
The intention being to keep the illumination constant during transmission. By definition, the average
transmitted electric power over a symbol period Tsis given by:
Pavg =1
TsZTs
0
p(t)dt =DPmax + (1 −D)Pmin (25)
where Dis the duty-cycle, of the PWM, Pmax and Pmin correspond to the power during on and of f
times respectively. Then, the average value of the power over the ith symbol depends on the duty cycle
D, and on Pmax and Pmin, not on the ith frequency fi. Transmitting the symbols Siand S(i+1) at the
same average power (see Fig. 8) brings up two scenarios:
Figure 8: LED’s control signal in frequency shift keying (FSK) for VLC systems
•Di=D(i+1)
In the case of identical duty cycles between the signals over the ith and the (i+ 1)th symbols, the signals
si(t) and si+1(t) must be defined with the same maximum and minimum values (Pi,max =P(i+1),max
and Pi,min =P(i+1),min).
•Di6=D(i+1)
In this case, the loss on the duty cycle during the transmission of the ith symbol will be compensated
by changing the maximum and/or the minimum power during the transmission of the (i+ 1) symbol as
shown in Fig. 8. The equations for compensation are given in (29) and (30).
Pi,max =Di
Di+1
P(i+1),max (26)
Pi,min =1−Di
1−Di+1
P(i+1),min (27)
4.4 Complex modulations
4.4.1 Colour Shift Keying
CSK is a scheme mapping data stream symbols to colours. The colours are produced by the tritimu-
lus principle using the three basic colours red-green-blue. Consequently, RGB-LEDs are used in CSK
modulation.
•CSK constellation design
Let Si={S1, S2, ..., SN}and Ci={C1, C2, ..., CN}be the incoming symbol set and the colour constellation
respectively, Nbeing the constellation size. Siis mapped into Ciand the light colour produced by the
RGB-LEDs changes by sequence of the data symbols. Cifits in a triangle delimited by the primary
colours red, green and blue (see Fig 9). Some design examples for CSK constellation are provided in
the literature in [39], [40] and [41]. Each Ck,k={1,2, ..., N }represents one point on the constellation
triangle shown in Fig 9. Ckis represented in the RGB colour space by its coordinates Rk,Gkand Bk.
Rk,Gkand Bk, the sum of which is 1, give the current intensities to be apply to the red, green and blue
LEDs for producing the colour Ck[39] [40] and [41].
11
Figure 9: RGB plan and constellation triangle
Nevertheless, Ckcan also be represented in the chromaticity plan by its xy coordinates given by
(xk=Rkxr+Gkxg+Bkxb(a),
yk=Rkyr+Gkyg+Bkyb(a),(28)
where xr,xg,xband yr,yg,ybare the xy coordinates of the primary colours red, green and blue. Fig. 10
shows the CSK transmission system using a single RGB-LED. Its channel matrix is given in (9) and the
transmission is governed by
rk= (H3×3)sk+ω, (29)
where rk=[R′
kG′
kB′
k]T,sk=[RkGkBk]Tand ω=[ωrωgωb]Trespectively.
Figure 10: A colour shift keying (CSK) transmission system
•Design constraints
The CSK constellation design has two main constraints. The first constraint is related to the lighting
while the second touches on the communication efficiency. The average power over the Ntransmitted
symbols and the average colour must remain constant. This has been extensively studied in [1] and [41].
The average power and colour are given by
Pavg =1
N
N−1
X
i=0
pi(30)
and
Cavg =
N
X
i=1
γiSi.(31)
12
where pirepresents the power allocated to the ith symbol and γithe weight of the the ith symbol. The
communication efficiency is determined in the design objective: under the conditions of average power
and colour, the system should maximise the minimum Euclidean distance between Siand Sk(Si,Sk
∈Si). This is achieved by applying the maximum a posteriori probability (MAP) detection rule (̟=
arg max probability {received|sent}). When the transmission is dominated by AWGN, the optimum
receiver has to minimize the squared Euclidean distance metric between two points of the constellation.
Hence, the objective is given by
̟=min{kH3×3(Si−Sk)k2}(32)
•Average energy
The average energy Eavg is an important characteristic of the constellation Si. It is generally given by
Eavg (Si) = 1
N
N−1
X
i=0
kSik2,(33)
where Siis mapped to Ci, and Ciis set of 3-tuple signal point having the sub-sets Si,r ={S1,r, S2,r , . . . , SN,r },
Sig,Sg ,r ={S1,g , S2,g, . . . , SN,g }, and Si,b ={S1,b, S2,b , . . . , SN,b }as sets of coordinates on the three axis
u,vand wrespectively (see Fir 9).
4.4.2 Orthogonal Frequency Division Multiplexing
OFDM is a multi-symbol modulation. It is one of those schemes where the form of the carrier wave must
be sinusoidal. Over the VLC channel, OFDM is performed using a single LED and a single photodetector
as shown in (Fig. 11). In the transmitter, the original data is converted to symbol data using inverse fast
Fourier transform (IFFT). The symbol data of all sub-carriers are superimposed to produce the baseband
signal expressed by:
lofdm =
N−1
X
i=0
Biej(2πi k
N).(34)
lofdm(t) is then used to modulate the LED. Considering the square matrix a2of the LED array in multiple
LED transmission, and taking into account the multipath effect and the delay of the ith propagation path
∆t, the output of the PD can be given by [42], [43]:
r(t) =
a2
X
j=0 X
i
Ωji νi(t−∆t) + ω(t),(35)
where Ωji represents a weighting factor characterised by the number of LEDs per array and the number
of arrays, and ν(t) is the amplified signal from the PD. The reverse operation in the receiver uses fast
Fourier transform (FFT) to recover the transmitted data expressed by:
Bi=
N−1
X
i=0
lofdme−j(2πi k
N).(36)
Figure 11: Basic principle of an orthogonal frequency division multiplexing (OFDM) for VLC systems
•Sub-carrier index OFDM (SIM-OFDM)
13
Sub-carrier index modulation OFDM exploits the sub-carrier orthogonality to add a new dimension to
the complex two dimensions signal plan used in the normal OFDM [33], [42] and [43]. SIM-OFDM may
increase power per sub-carrier, and will also increase the throughput and give better spectral efficiency
in respect to the normal OFDM.
4.4.3 Multiple inputs multiple outputs
The use of multiple LEDs or multiple arrays of LEDs offers the potential for parallel transmissions across
the VLC channel. The channel matrix given in (7) is suitable for a MIMO channel. It considers nand
ztransmitting and receiving antennas respectively. The principle diagram of a MIMO-VLC system is
depicted in Fig. 12.
Figure 12: Basic principle of an n×zMIMO-VLC system
Indoor MIMO was reviewed in the literature and more information can be found in [44] and [45].
The incoming symbols are redistributed for a parallel transmission using the total number of LEDs or
group of LEDs. This technique is characterised by a very hight channel capacity. In fact the capacity of
a MIMO system is the capacity of a SISO system multiplied by the scalar number Γ (see (8)).
Figure 13: Basic principle diagram of a space-time-coding (STC) for VLC systems
In MIMO-VLC system implementation, it is very important to distribute the redundancy in space and
time in an appropriate manner in the transmitter. This is achieved by applying the space-time-coding
(STC) technique [44], [46]. The aim here is to seek better coding and diversity gains. STC provides
very good signal quality at the receiver, thus, good SNR. A basic representation of the STC and spatial
multiplexing transmitter is given in Fig. 13. The received symbols set is given by the relation in (1),
14
where the channel response His replace by the n×zimpulse response (Hmulti) given in (7). In fact,
Hmulti is seen as a composed matrix having three components: two unitary rotation matrices Uand V,
and a diagonal matrix Λ, hence, Hmulti is given by [8]
Hmulti =UΛV∗(37)
(1) becomes
R= (UΛV∗)S+W,(38)
where Sis a n-dimensional transmitted vector, Rand W, are z-dimensional receiving and noise vectors
respectively. A geometry rotation is applied to (38), Afterwards,the parameters S,Rand Ware given
by their representation S′,R′and W′respectively, such that
R′= ΛS′+W′,(39)
where S′=V∗S,R′=U∗Rand W′=U∗W. The corresponding analogue baseband signal yj(t)
measured at the jth PD can be expressed by [47]
yj(t) = 1
nT
nT
X
i=1
hji Si(t) + Wj(t),0≤t≤Ts,1≤j≤z, (40)
where Wjis the noise components at the jth PD.
•Optical spacial modulation (OSM)
OSM is a combination of the spacial multiplexing and the pulse position modulation. This technique is
studied and presented in [48]. In terms of performance, a 4×4 OSM achieves the same BER performance
as OOK, even though at the double data rate.
5 Challenges of VLC practical implementation
Implementing a VLC communication system requires in-depth knowledge of both communication and
electronic systems. The communication perspective demands a good definition of the modulation scheme
which depends on the application of the communication system and should meet the communication
objectives. This aspect also requires us to know if it is necessary or not to use a coding technique.
The practical implementation of the design requires one to have good knowledge of electronic circuit
design, in this case, accuracy of the design, quality of the schematic diagram and quality of the selected
components are very important. The responsivity of all the elements appearing in the design is to be
taken into consideration. The complexity of the design depends on the complexity of the modulation
technique. So, modulation schemes such as OOK are easier to implement when compared to complex
schemes such as CSK. In this aspect, the type of waveform is also very important. Schemes demanding
pulses provide less complexity in design compared to those demanding sine waves. This section of the
article details a few techniques that may be used in the practical implementation of a VLC system. The
aims are to comply with the illumination constraints of the VLC technology. Before we go in-depth into
the development, let us start by presenting the LED control and dimming.
5.1 LED control and dimming
LEDs are semi-conductors with two modes of operation: the forward and reverse biased modes. In
forward biased mode, the electrons are able to recombine with holes within the device, releasing energy
in the form of photons, producing the forward current IF.IFdefines the brightness of the LED and
therefore requires to be limited. The main elements used to control IFare: a series resistor and a current
source used to produce a stable current. Here, LEDs is a combination of a finite number αof identical
light emitting diodes in series and in βbranches. The current limiting resistor is then given by
R=Vs−∆v−αVF
βIF
.(41)
15
In practice, the value of VFfor on-shelf LEDs is between 1.2V and 2.5V for a current IFof about 60 mA.
The combination to be used depends on the number of lumens required in the room. A PWM control
method is used to control the duty cycle and the dimming of the LEDs. This module will be accompanied
by the communication module grouping the modulator and many other components. The output power
of the LED is given by Pd=VFIF.Pdis also given as a function of the duty cycle as expressed in (25).
The basic diagram of LED’s control is given in Fig. 14.
Figure 14: Basic diagram of LED’s control
5.1.1 Controlling the Brightness of LEDs
Dimming and illumination of LEDs are controlled by the brightness. Many methods can be used to control
the brightness of the LEDs. These include pulse frequency modulation (PFM), bit angle modulation
(BAM) and PWM. PWM is the most used method. It is based on the variation of the duty cycle of the
pulses, through varying the duration of the on-time ton. Fig. 15 shows brightness control using the PWM
method. The percentage of the duty cycle corresponds to the percentage of the brightness needed for a
specific illumination and a specific value of Pd.
Figure 15: Brightness control of LEDs using a pulse width modulation (PWM) signal
5.2 LED control for OOK systems
On-off-keying is possibly the simplest modulation that can be implemented over the VLC channel. After
the signal conditioning, channel coding, interleaving and bit mapping; the signal message generates a
PWM that will control the switching system. This is performed through a driver to match the voltage
and current level between the PWM source and the control pin of the switch. The baseband signal in
OOK is produced to control the LEDs, resulting in the carrier signal (Fig. 16).
16
Figure 16: Basic diagram for on-off keying (OOK) modulation and LED control
5.3 LED control for frequency shift keying (FSK) transmission systems
Fig. 17 shows an example of FSK modulation technique over VLC. It includes a module treating the
message to be sent. Channel coding and interleaving if necessary, and a function to map symbols to
different circuits is produced. The circuit corresponds to a specific PWM generator with its specific
frequency, a switching system and a driver. Then the LED will be lighting at variable frequencies
according to the input data stream. It is to be emphasised that, this system works with a DC offset
injected in the circuit.
Figure 17: Basic diagram for frequency shift keying (FSK) modulation and LED control
5.4 LED control for colour shift keying (CSK) systems
Figure 18: Principle diagram for colour shift keying (CSK) modulation and control of the RGB-LED
The complexity of the mapping module makes difficult the implementation of a CSK system. In CSK,
symbols are mapped in turn to colours. This can be done practically using the diagram presented in
Fig. 18. The symbol mapping processor executes an algorithm that is more complex than that used in
17
OOK. The RGB-LED circuit is made of three single colour LEDs in series with the current limiter. In this
modulation scheme, there is no flickering, meaning that there is no light interruption during transmission.
To achieve this, to each symbol is allocated three forward currents with different intensities in such a
way that the output of the RGB-LED produces a constant optical power. Considering that the three
intensities are IR,IGand IB, and the average power produced by a single LED is given in (11). Applying
(11) to each LED, an algorithm that meets the power envelop constraint of CSK can be developed.
5.5 LED control for VLC-OFDM systems
Implementing OFDM over the VLC channel is more complex when compared to other modulation
schemes. This is because the implementation needs to produce a signal similar to the graph shown
in Fig. 19. The OFDM signal is added to a DC power in such a way that the average value Vavg over the
symbol period remains constant. Vavg is then used to control the RGB-LED.
Figure 19: Typical wave form in VLC-OFDM
6 Applications of VLC
By modulating the fluorescent light, the VLC technology can transmit signals at 10kb/s. But LEDs can
provide up to 500 Mbps. From 1.2 Mbps, the system is said to be high data rate transmission [3]. The
potential applications of VLC technology can be grouped (see Table 1 and 2) in two main sets: Low data
VLC applications and applications for high data transmission.
6.1 Low data VLC
In general nowadays, low data rate communication systems are intended for control. VLC technology
can also be implemented in such applications. Table 1 presents a few of them.
Table 1: Application field of low data VLC technology
Application
fields
Comments
General position-
ing
VLC technology provides the possibility to uniquely identify each VLC transmitter
in an indoor environment
Vehicle and trans-
portation
VLC technology also allows inter-vehicle communication in using the front light
as transmitter and the back red light as receiver thus prevent accidents [49]. This
property is also used by engineers to design advanced traffic-light systems that
can, for example, give passage to a vehicle under control of the red light when
there is no vehicle travelling in the green light direction
Smart lighting The intensity and the switching of a light can always be controlled remotely thus
providing users comfort in utilising the light. This quality is advantageous because
of its energy saving property
18
6.2 High data VLC
The applications of high data rate VLC are beyong the scope of this paper. Table 2 highlights the most
promising applications of VLC that demand high data rate techniques.
Table 2: Application field of high data VLC technology
Application
fields
Comments
Mobile connectiv-
ity
Firstly, by enabling light fidelity (Li-Fi ), VLC can be used to relieve the wireless
radio frequency (RF) spectrum. The light bulbs become the base station for
millions of users. Secondly, it enables secure communication between two smart
devices at a very high speed data link.
Hospitals and
healthcare
RF is undesirable in some parts of hospitals, especially around scanners and in
operating theatres. VLC technology consequently is the promising technology
that can be used in hospitals owing to the fact that the VLC spectrum does not
interfere with the RF spectrum.
Aviation To reduce the risk of interference, LiFi may be used to replace WiFi in aircraft
cabins. Additionally, VLC can be used to provide air-passengers with continuous
communication services. This will impact the cost in aircraft manufacturer.
Underwater com-
munications
The propagation of RF signals is not efficient in underwater environments. This
is due to the quick attenuation of the transmitted signal [50]. The propagation of
visible light enhances the quality of underwater data transmission.
Defence, security,
and hazardous en-
vironments
The VLC signal cannot pass through walls. This technology is well suited for
application where security is of great importance. In hazardous environments
such as a petrol-chemical plan or a mining environment, there is a high risk of
explosion when the RF signal is used. RF signal can produce a spark of such
intensity that it could ignite a vapour air mix, which could lead to an explosion
[51]. VLC is an excellent technology that can be used in these settings
6.3 Other applications
There are probably many other applications of VLC not dealt with in this paper. The technology can
also be used in applications such as smart learning and teaching, smart grid-friendly appliances, asset
analyses, resource tracking, and smart networking, from an endless list.
7 Challenges of VLC commercialisation
The utilisation of LEDs for lighting provides many advantages such as in cost and comfort. Increasingly,
it is becoming unavoidable to light without using LEDs in light bulbs. In general, manufacturers exploit
the potential offered by the emerging technologies to meet the needs of the population. For a large scale
deployment of VLC technology, a number of challenges and issues relating to its commercialisation needs
to be addressed. Hence, highlights the regulation and market factors. Firstly, the VLC spectrum needs
to be organised. To date, the standardisation of VLC technology has received relatively little attention.
In IEEE, however, the working group IEEE 802.15.7 provides rules and regulations for VLC technology.
In Japan, the consortium VLCC has also developed a few working standards. The Japan Electronics and
Information Technology Industries Association (JEITA) has standards focusing on visible light identifi-
cation systems (see Section 3). Though these standards could be enough to allow the deployment of VLC
technology, there are still issues to be addressed when its comes to commercialising the technology. Other
communication technologies, such as wireless RF technology, could see VLC as a very strong competitor,
but this may not happen if the manufacturers are the same for both technologies. With new technologies,
new companies must come to the fore, but looking at the types of technology (short range versus long
range), the applications and advantages provided by each, VLC technology may never supersede RF
wireless technology. Many other barriers are emerging down the mass production of VLC transceivers.
19
Duplex transmission techniques are very difficult to implement, as driving LEDs at high speed demands
very accurate electronic calculations. The line of sight characteristic of light works best over very short
mobility distance. VLC has seamless interoperability with other networks. LEDs can not be used as
network product due to the disruptive nature of the lighting.
8 Advances in VLC technologies and field trials
Against the background knowledge that VLC technology is still in its infancy, we must highlight the
increasing volume work that has been done. Researchers, universities and companies are working to see
this technology deployed. In the United Kingdom, the first Li-Fi application is installed in a classroom
in the Business Academy Bexley, a school in South London [52]. The installed system provides access to
internet at 5 Mbps for both uplink and downlink. In Japan there have been many cases of field trials.
A 100 Mbps full-duplex multiple-access VLC system is in used. It is a multiple access system using a
carrier sense multiple access with collision detection (CSMA/CD) method to transmit at a distance of
3 meters [53]. Also developed in Japan is a high-speed parallel wireless VLC system using 2D image
sensor and LEDs on the Transmitter side. The transmitter is composed of an array of 8×8 LEDs, each
transmitting different data. The photodetector is composed of a high speed image sensor including an
array of photo transistors. Another Japanese achievements is a three-dimensional position measuring
system with an accurate position detection of a transmitter or a receiver. In this last example, the
LED lights send position data, and an image sensor of a robot receives the data as well as the image.
The signal processing is performed inside the robot to calculate its position. Some cases of hybrid
system including VLC and other communication technologies such as power line communication (PLC)
technology are reported [9]: this example presents the implementation of a hybrid system involving PLC
and VLC. Another example using spread FSK (S-FSK) modulation on the PLC side and OOK technique
on the VLC side is presented. In this case, the behaviour of the visible light channel corrupted by
background noise due to sunlight including the fluorescent light was investigated. Earlier in this paper,
the communication system is based on the spread orthogonal continuous phase binary frequency shift
keying (SOCPBFSK) and OOK. It is a bridge between PLC and VLC channels to relay communication
for low data rate purposes [38]. Some characteristics of the interface are presented and the error rate is
presented. It is reported from this work that the transmission over the VLC channel is affected by PLC
noise and the background light over the transmission environment.
9 Unsolved problems and future directions
Let us recall that an effective implementation of the VLC technology depends on the existence of appro-
priate regulations. From this angle, standards are lacking to regulate VLC. Likewise, VLC faces both
practical implementation and commercialisation challenges as discussed in Section 5 and Section 7 re-
spectively. Thus further research needs to be done both sectors. Most of this will be based on subjects
such as Li-Fi, MIMO, multi-user VLC, hybrid system combining PLC and VLC technology or RF and
the visible spectrum to name only few.
List of Abbreviations
AWGN: Additive white Gaussian noise
BAM: Bit angle modulation
BER: Bit error rate
CICTR: Center for information communication technology research
CSK: Colour shift keying
CSMA/CD: Carrier sense multiple access / Collision detection
DC: Direct current
20
DH: Dual-header
dLOS: Direct line-of-sight
DPPM: Differential pulse position modulation
(I)FFT: (Inverse) Fast Fourier transform
IEEE: Institute of Electrical and Electronics Engineers
ISI: Inter-symbol interference
JEITA: Japan Electronics and Information Technology Industries Association
LASER: Light amplification by stimulated emission of radiation
LED: Light emitting diode
LiFi: Light fidelity
(N)LOS: (Non) Line-of-sight
(L)(V)PPM: (Level)(Variable) Pulse position modulation
MAP: Maximum a posteriori probability
MIMO: Multiple input - multiple output
MISO: Multiple input - single output
ndLOS: Non-direct line-of-sight
NEC: Nippon Electric Company
OFDM: Orthogonal frequency division multiplexing
OOK: On-off keying
OSM: Optical spacial modulation
OWC: Optical wireless communications
PD: Photodetector
PDM: Pulse duration modulation
PFM: Pulse frequency modulation
PHY: Physical layer
PIM: Pulse interval modulation
PLC: Power line communications
PSK: Phase shift keying
PWM: Pulse width modulation
RF: Radio frequency
RGB: Red - green - blue
SDO: Standardisation organisation
SER: Symbol error rate
21
(S)FSK: (Spread) Frequency shift keying
SIM: Sub-carrier index modulation
SISO: Single input - single output
SNR: Signal-to-noise ratio
SOCPBFSK: Spread orthogonal continuous phase binary frequency shift keying
STC: Space-time-coding
VLC: Visible light communications
VLCC: Visible light communications consortium
WiFi: Wireless fidelity
WWRF: Wireless World Research Forum
References
[1] R. J. Drost and B. M. Sadler IEEE Transactions on Communications, vol. 62, pp. 1995–2005, 2014.
[2] K. Lee and H. Park Photonics Technology Letters, IEEE, vol. 23, pp. 1136–1138, 2011.
[3] IEEE IEEE P802.15.7/D8, April 2011, pp. 1–306, 2011.
[4] M. Nakagawa, “Visible light communications,” in Conference on Lasers and Electro-Optics/Quantum
Electronics and Laser Science Conference and Photonic Applications Systems Technologies, 2007.
[5] A. P. S. Louvros and D. Fuschelberger, “Vlc technology for indoor lte planning,” in System-Level
Design Methodologies for Telecommunication, pp. 21–41, Springer, 2014.
[6] M. Biagi, A. Vegni, and T. D. C. Little, “Lat indoor mimo-vlc localize, access and transmit,” in
Optical Wireless Communications (IWOW), 2012 International Workshop on, pp. 1–3, Oct 2012.
[7] Y. Hong, J. Chen, Z. Wang, and C. Yu Photonics Journal, IEEE, vol. 5, pp. 7800211–7800211, 2013.
[8] P. M. Butala, H. Elgala, and T. D. Little, “Svd-vlc: A novel capacity maximizing vlc mimo system
architecture under illumination constraints,” in Globecom Workshops (GC Wkshps), 2013 IEEE,
pp. 1087–1092, IEEE, 2013.
[9] T. Komine and M. Nakagawa Consumer Electronics, IEEE Transactions on, vol. 49, pp. 71–79,
2003.
[10] Z. Wu, J. Chau, and T. Little, “Modeling and designing of a new indoor free space visible light
communication system,” in Networks and Optical Communications (NOC), 2011 16th European
Conference on, pp. 72–75, IEEE, 2011.
[11] K. Lee, H. Park, and J. Barry Communications Letters, IEEE, vol. 15, pp. 217–219, February 2011.
[12] J. Barry, J. Kahn, W. Krause, E. Lee, and D. Messerschmitt Selected Areas in Communications,
IEEE Journal on, vol. 11, pp. 367–379, Apr 1993.
[13] M. Saadi, L. Wattisuttikulkij, Y. Zhao, and P. Sangwongngam International Journal of Electronics
& Informatics, vol. 2, 2013.
[14] S. il Choi, “Analysis of vlc channel based on the shapes of white-light led lighting,” in Ubiquitous
and Future Networks (ICUFN), 2012 Fourth International Conference on, pp. 1–5, July 2012.
22
[15] R. Roberts and Z. Xu, “Update on vlc link budget work,” in IEEE 802.15. 7 VLC Standard Meeting,
2009.
[16] M. Biagi, T. Borogovac, and T. D. Little Journal of Lightwave Technology, vol. 31, pp. 3676–3686,
2013.
[17] H. Yang, J. W. Bergmans, T. C. Schenk, J.-P. M. Linnartz, and R. Rietman Optics express, vol. 16,
pp. 21641–21646, 2008.
[18] J. Tan, K. Yang, M. Xia, and Y. Yang Opt. Appl, vol. 41, pp. 507–517, 2011.
[19] J. M. Kahn and J. R. Barry Proceedings of the IEEE, vol. 85, pp. 265–298, 1997.
[20] X. Zhang, K. Cui, M. Yao, H. Zhang, and Z. Xu, “Experimental characterization of indoor visible
light communication channels,” in Communication Systems, Networks & Digital Signal Processing
(CSNDSP), 2012 8th International Symposium on, pp. 1–5, IEEE, 2012.
[21] F. Porikli, “Inter-camera color calibration by correlation model function,” in Image Processing, 2003.
ICIP 2003. Proceedings. 2003 International Conference on, vol. 2, pp. II–133, IEEE, 2003.
[22] Z. Yu, R. J. Baxley, and G. T. Zhou, “Multi-user miso broadcasting for indoor visible light communi-
cation,” in Acoustics, Speech and Signal Processing (ICASSP), 2013 IEEE International Conference
on, pp. 4849–4853, IEEE, 2013.
[23] J. R. Janesick, “Photon transfer,” SPIE, 2007.
[24] K. Cui, G. Chen, Z. Xu, and R. D. Roberts, “Line-of-sight visible light communication system design
and demonstration,” in Communication Systems Networks and Digital Signal Processing (CSNDSP),
2010 7th International Symposium on, pp. 621–625, IEEE, 2010.
[25] M. Zhang and Z. Zhang, “Fractionally spaced equalization in visible light communication,” in Wire-
less Communications and Networking Conference (WCNC), 2013 IEEE, pp. 4282–4287, IEEE, 2013.
[26] P. S. Venkataram. Web information. Available at http://web.mit.edu/pshanth/www/
johnsonshot_psv1.pdf. Accessed: 2014-09-14.
[27] T.-H. Do and M. Yoo, “Received power and snr optimization for visible light communication system,”
in Ubiquitous and Future Networks (ICUFN), 2012 Fourth International Conference on, pp. 6–7, July
2012.
[28] R. D. Roberts, S. Rajagopal, and S.-K. Lim, “Ieee 802.15. 7 physical layer summary,” in GLOBE-
COM Workshops (GC Wkshps), 2011 IEEE, pp. 772–776, IEEE, 2011.
[29] A. Yokoi IEEE P802, vol. 15, 2008.
[30] S. Rajagopal, R. Roberts, and S.-K. Lim Communications Magazine, IEEE, vol. 50, pp. 72–82, 2012.
[31] K. Wong, T. O’Farrell, and M. Kiatweerasakul International Journal of Communication Systems,
vol. 13, pp. 551–576, 2000.
[32] S. Teramoto and T. Ohtsuki IEICE transactions on communications, vol. 87, pp. 3385–3388, 2004.
[33] R. Abu-Alhiga and H. Haas in Personal, Indoor and Mobile Radio Communications, 2009 IEEE
20th International Symposium on, pp. 177–181, IEEE, 2009.
[34] D. Tsonev, S. Sinanovic, and H. Haas, “Enhanced subcarrier index modulation (sim) ofdm,” in
GLOBECOM Workshops (GC Wkshps), 2011 IEEE, pp. 728–732, IEEE, 2011.
[35] L. Hanzo, M. M¨unster, B. Choi, and T. Keller, OFDM and MC-CDMA. Wiley, March, 2004.
[36] J.-H. Yoo and S.-Y. Jung EURASIP Journal on Wireless Communications and Networking, vol. 2013,
pp. 1–6, 2013.
23
