Content uploaded by Ahmad Usman
Author content
All content in this area was uploaded by Ahmad Usman on Jul 05, 2014
Content may be subject to copyright.
Visible Light Communication using Wavelength
Division Multiplexing for Smart Spaces
Talha A. Khan*, Muhammad Tahir** and Ahmad Usman**
*LUMS School of Science & Engineering
**University of Engineering & Technology,Lahore
talha.khan@lums.edu.pk, mtahir@uet.edu.pk, ahmad
usman86@hotmail.com
Abstract—LED lights used mainly for illumination purpose,
can be simultaneously used for data transmission. In this pa-
per, a visible light communications system is proposed that
employs wavelength division multiplexing, to transmit multiple
data streams from different data sources simultaneously, over
the wireless channel. Successful implementation of a low cost
prototype system is reported. The effect of ambient noise on
system performance is studied. On the basis of empirical results,
performance evaluation of proposed solution is presented.
I. INTRODUCTION
LED-based lighting has evolved as the lighting technology
of choice due to high brightness, affordable cost, low power
consumption and minimal heat generation [1]. In Visible
Light Communication (VLC), LEDs used for illumination
purpose are simultaneously used for wireless data transmis-
sion. It offers numerous advantages such as high data rates,
unlicensed large bandwidth and better data security leading
to smart spaces [2]. VLC system exploits the unregulated
visible spectrum and poses no health hazards. Existing lighting
infrastructure also paves the way for novel VLC applications
in home as well as industry.
Different solutions have been proposed in literature for VLC
system architecture, its performance analysis, improved data
transmission rates and brightness control, to name a few. A
fundamental theoretical analysis of an indoor VLC system was
presented in [3]. The feasibility of different indoor as well as
outdoor VLC applications has since been presented [4]. We
first give a brief overview of recent research activity on differ-
ent VLC frontiers, starting with under-water communications.
It is not possible to realize underwater RF communications
due to high absorption of electromagnetic signals at radio fre-
quencies. However, VLC has enabled high data-rate (10Mbps),
moderate distance(100m), underwater communication [5] as
the visible spectrum is subjected to lesser attenuation. Thus,
VLC emerges as a suitable alternate where RF fails to deliver
due to bandwidth constraints or physical limitations. Similarly,
novel VLC applications for outdoor scenarios are also being
explored. Active research is underway in the realm of Intelli-
gent Transport System (ITS) employing visible light for traffic-
signal to vehicle communications [6].
Although VLC promises very high data transmission rates
but it holds true in principle only. In practical VLC systems,
data rate is typically limited to a few Mbps owing to a limited
modulation bandwidth of white (phosphor-based) LEDs [4].
Several pre-equalization techniques have been proposed in
order to realize higher modulation bandwidths in a (phosphor-
based) white light VLC system [7], [8]. Alternatively, white
light can be produced by mixing red, green and blue colors
in appropriate proportion. Unlike the phosphor-based white
LEDs, colored LEDs have a much larger modulation band-
width and can support relatively higher data-rates. Further-
more, it obviates the need of using pre-equalization filters
at the transmitter, as needed in conventional phosphor-based
white light VLC systems. It also provides superior illumination
compared to phosphor-based white LEDs while increasing the
energy efficiency at the same time [9].
The use of colored LEDs enables us to exploit the inher-
ent frequency gap between different colors for simultaneous
transmission of multiple signals. This paper is the first step
towards realization of a line-of-sight (LOS), white light VLC
system based on Wavelength Division Multiplexing (WDM),
where colored LEDs are used in place of white phosphor-
based LEDs. These colored LEDs are modulated individually
with an optical mixer at front end of the transmitter to produce
white light. On the receiver side, received signal is first de-
multiplexed by passing it through optical filters and then data
streams are recovered by the corresponding optical sensors.
This scheme not only allows high data-rate communication but
also ensures efficient usage of the visible spectrum. Diversity
and/or multiplexing gains can be leveraged by concurrent
transmission of multiple data streams. Moreover, this scheme
could also be incorporated in Visible Light Information Broad-
cast systems based on multi-color advertisement boards as they
usually consist of LEDs of different colors. Different colors
can be reserved for different kind of information, rendering
the entire application more useful.
The rest of the paper is organized as follows. In Section
II, we describe channel model of a generalized VLC system.
Section III highlights the VLC system design aspects including
modulation techniques and transmitter/receiver front-ends. The
implementation of the proposed VLC system is discussed in
Section IV. Performance evaluation of developed platform is
presented in Section V. Section VI concludes the paper.
II. C
HANNEL MODEL
An indoor channel model for optical wireless communica-
tion (OWC) system was presented in [10]. Although it was
proposed for an Infra-red (IR) based OWC system, it has
widely been adopted for a VLC channel. However, unlike
an IR source that can be approximated as a monochromatic
emitter, a white light LED source is inherently wideband ( 380-
780nm). It calls for the inclusion of wavelength-dependent
channel parameters while modeling VLC systems. For in-
stance, as the reflectance characteristics of indoor materials
are wavelength dependent, they should not be assumed to have
a constant value. Recently, [11] presented a modified VLC
channel model taking into account the wideband nature of the
white light VLC sources. It turns out that the delay spread and
the reflection power for the diffuse signal is very small in a
VLC system compared with an IR system.
We now present the channel model for a general indoor VLC
system. The received signal consists of the line-of-sight (LOS)
signal which directly reaches the receiver, and the diffuse
component comprising of transmitted signals that arrive at the
receiver after multiple reflections. Thus the impulse response
ℎ(𝑡) for VLC channel is expressed as:
ℎ(𝑡)=ℎ
𝐿𝑂𝑆
(𝑡)+ℎ
𝑑𝑖𝑓𝑓𝑢𝑠𝑒
(𝑡) (1)
where ℎ
𝑑𝑖𝑓𝑓𝑢𝑠𝑒
is the impulse response for diffuse light
component, while ℎ
𝐿𝑂𝑆
(𝑡) denotes the impulse response of
the line-of-sight light component [12]. The expression for
ℎ
𝐿𝑂𝑆
is ℎ
𝐿𝑂𝑆
=
𝐼
𝑖=1
𝑃
𝑖
𝛿(𝑡 − 𝜏
𝑖
), where 𝑃
𝑖
denotes the
emitted optical power of the 𝑖
𝑡ℎ
ray while the propagation
delay for the 𝑖
𝑡ℎ
ray is given by 𝜏
𝑖
=
⃗𝑟
𝑡
− ⃗𝑟
𝑟
𝑐
. Here ⃗𝑟
𝑡
, ⃗𝑟
𝑟
denote the position vectors for the transmitter and receiver
respectively, and c is the speed of light. Similarly, contribution
of diffuse signal is
ℎ
𝑑𝑖𝑓𝑓𝑢𝑠𝑒
=
𝐽
𝑗=1
[
𝐾
𝑘=1
𝛼
𝑗,𝑘
]𝑃
𝑗
𝛿(𝑡 −
𝑊
𝑗
𝑐
),
where 𝛼
𝑗,𝑘
denotes the path loss for the 𝑗
𝑡ℎ
ray after 𝑘
reflections, 𝑊
𝑗
/𝑐 is the propagation delay for the 𝑗
𝑡ℎ
ray and
𝐾 is the maximal ray reflection number. Thus the impulse
response of the single-source VLC channel with multiple
reflections is given by
ℎ(𝑡)=
𝐼
𝑖=1
𝑃
𝑖
𝛿(𝑡 − 𝜏
𝑖
)+
𝐽
𝑗=1
[
𝐾
𝑘=1
𝛼
𝑗,𝑘
]𝑃
𝑗
𝛿(𝑡 −
𝑊
𝑗
𝑐
). (2)
Similarly, we can obtain the impulse response of a multiple-
source VLC channel with multiple reflections using
ℎ(𝑡)=
𝑁
𝑛=1
𝐼
𝑖=1
𝑃
𝑛,𝑖
𝛿(𝑡 − 𝜏
𝑛,𝑖
)+
𝑁
𝑛=1
𝐽
𝑗=1
[
𝐾
𝑘=1
𝛼
𝑛,𝑗,𝑘
]𝑃
𝑛,𝑗
𝛿(𝑡 −
𝑊
𝑛,𝑗
𝑐
), (3)
where 𝑁 is the number of LED sources. Monte-Carlo sim-
ulations for the above channel model reveal that a furnished
room generally acts as a linear time invariant low pass filter
for the transmitted optical signals. However, for directed single
source LOS communication, effect of shadowing and multiple
reflections can be ignored. Thus, channel impulse response can
be approximated by a scaled delayed delta function.
We now consider the proposed WDM based VLC system
to derive an expression for signal-to-interference ratio (SIR).
It can be verified that for a short range LOS VLC system, the
photometric path loss (𝑝
𝑙
) equals the free space power path
loss, for Lambertian sources and is given by [13]
𝑝
𝑙
≈
(𝑚 +1)𝐴
𝑟
cos(𝛼)cos
𝑚
(𝛽)
2𝜋𝐷
2
, (4)
where 𝑚 =
− ln(2)
ln(cos(𝜓))
(𝜓 is the semi-angle of the LED at half
illumination) is the order of Lambertian emission, 𝐴
𝑟
is the
receiver area, 𝐷 is the transmitter-receiver distance and 𝛼, 𝛽
are the angles from transmitter-receiver axis to receiver normal
and transmitter normal respectively. Since the commonly used
LEDs usually do not have any beam-shaping fixture in front,
they can be approximated as Lambertian sources. We now
consider an LOS, VLC system where 𝑁 transmitters of
distinct wavelengths are being used for simultaneous data
transmission. We can express the signal to interference (𝑆𝐼𝑅)
ratio as
𝑆𝐼𝑅 =
(𝑚
𝑠
+1)𝐴
𝑟
cos(𝛼
𝑠
)cos
𝑚
𝑠
(𝛽
𝑠
)
2𝜋𝐷
2
𝑠
𝐹
𝑠
(𝜆, 𝜃)
𝑁−1
𝑖=1
(𝑚
𝑖
+1)𝐴
𝑟
cos(𝛼
𝑖
)cos
𝑚
𝑖
(𝛽
𝑖
)
2𝜋𝐷
2
𝑖
𝐹
𝑖
(𝜆, 𝜃)
, (5)
where 𝑠, 𝑖 indicate desired and interfering VLC transmitters
respectively, 𝐹
𝑥
(𝜆, 𝜃), 𝑥 ∈{𝑠,𝑖}, is the average attenuation
experienced by visible light signal from a given transmitter
as it passes through the wavelength-filter and 𝜆 and 𝜃 are
respectively the wavelength and angle of incidence. For the
filter response independent of 𝜃, 𝐹
𝑥
for transmitter 𝑥 can
be calculated by averaging the filter response over the band
of interest as 𝐹
𝑥
=
1
𝜆
ℎ
𝑥
−𝜆
𝑙
𝑥
𝜆
ℎ
𝑥
𝜆
𝑙
𝑥
𝑓(𝜆)𝑑𝜆, with 𝑓 (𝜆) be-
ing the normalized response(transmission) of the filter, and
𝜆
ℎ
𝑥
,𝜆
𝑙
𝑥
denote the upper and lower wavelength bounds of
the transmitter. It could be verified that for symmetrically
placed identical transmitters equidistant from the receiver
(𝑖.𝑒.𝑚
𝑖
= 𝑚
𝑠
,𝐷
𝑖
= 𝐷
𝑠
, ∀𝑖, 𝑖 =1, ..., 𝑁 − 1), 𝑆𝐼𝑅 depends
on the ratio
𝐹
𝑠
∑
𝑖
𝐹
𝑖
(𝜆)
, where denominator equals the sum of
the average attenuation experienced by the interfering data
streams.
III. S
YSTEM DESIGN
In a VLC system, Intensity Modulation (IM) is used in
conjunction with Direct Detection (DD) for data modulation
and demodulation at the transmitter and receiver. As the main
purpose of LEDs is to provide illumination, several techniques
are used to provide dimming/brightness control including
Pulse-Position Modulation (PPM), Pulse-Width Modulation
(PWM), Pulse-Amplitude Modulation (PAM) and Bit-Angle
Modulation (BAM).
A. Transmitter and Receiver Front Ends
The transmitter front end of a VLC system generally
consists of an LED source, driver and modulation circuitry
and optical components for beamshaping and diffusion. Two
types of LED based white light sources are commercially
available namely the three-chip LED and the single-chip LED.
In a three-chip or Red-Green-Blue (RGB) LED, high intensity
white light is generated by mixing red, green and blue lights
in appropriate proportion. Alternatively, white light could also
be produced using a blue LED coated with yellow phosphor-
which gives us the one-chip LED. From an illumination
perspective, phosphor-based LEDs enjoy widespread popu-
larity due to lower cost and simpler manufacturing process.
However, it has a major drawback when it comes to data
communications- a limited modulation bandwidth(∼3 MHz
[14]) which results in low data rates. It can be attributed to
the long response time of the phosphor coating which limits
the otherwise high modulation bandwidth of the underlying
blue emitter. Pre-equalization techniques have to be employed
at the transmitter to increase the modulation bandwidth.
Another way of getting around this bottleneck is to employ
multichip LEDs instead, which promise higher data rates. Ad-
vances in solid-state lighting indicate that low cost multichip
devices will be available in future [15], rendering them even
more desirable than their one-chip counterparts. Furthermore,
they also lend themselves as potential candidates for WDM
due to presence of red, green and blue emitters at the core.
The VLC receiver front-end generally consists of a concen-
trator lens, an optical filter, and a photodetector, followed by a
pre-amplifier, a post-equalizer and an electrical filter. Mostly,
a photodiode is used for direct detection owing to its low cost
and high reception bandwidth. As the received optical power
is proportional to the light collection area, it is desirable to
employ an optical concentrator to increase the effective area
1
.
In addition, it also compensates for the high spatial attenuation
due to the beam divergence from the LEDs.
B. Wavelength Division Multiplexing
WDM has long been used for multiplexing optical carrier
signals onto the same optical fiber strand by using a different
wavelength for each signal. As it has proved beneficial for
fiber-optic communications, it is expected that WDM will
also bring about multiple benefits in future VLC systems
[10]. Using WDM in VLC systems results in an efficient
utilization of the unregulated visible spectrum while offering
multiplexing gain at the same time. It can also be exploited
to provide frequency diversity in application scenarios where
the white light VLC link is prone to frequency selective noise
(due to unmodulated colored lighting devices or impure white
lighting devices that result in frequency-selective noise). A
viable solution would be to use WDM for robust white light
visible light communication (different wavelengths will not be
affected equally by noise at a given location).
IV. I
MPLEMENTATION
We now present the proposed VLC system which uses
WDM to transmit multiple data streams simultaneously over
1
It is expensive to directly increase the photodiode light collection area and
it also tends to decrease the receiver bandwidth while increasing the receiver
noise [10].
Data from
Laptop
USB to Serial
Bridge Controller
Level Conversion (RS
232 to TTL)
Visible Light
Modulation
(OOK)
Data from
Laptop
LEDs
Visible Light
Data from
Laptop
Data
Optical Filter
Photo
diode
Visible Light
Demodulation
Level Conversion (TTL
to RS 232)
Serial to USB
Bridge Controller
Data to
Laptop
Fig. 1. System architecture block diagram.
the wireless channel. Distinct wavelengths are used for trans-
mitting each data stream by using LED sources of different
colors. An optical mixer can then be used at the transmitter to
produce white light by mixing the output of colored sources.
In order to de-multiplex the received data streams, wavelength-
specific filters are used.
In the prototype implementation, we transmitted three
different audio-streams simultaneously using 115𝐾𝑏𝑝𝑠 data
rate(limited due to RS232 interface) for each, over a short
distance in free space. The receiving laptop played the live
audio-stream which was selected by using the appropriate
wavelength-filter. It should be noted that single low-cost minia-
ture LEDs were used, rather than the more expensive high
power ones, which also accounts for the lower transmission
distance achieved (50 cm). Red, green and blue LEDs were
used as visible light data transmitters, while providing illumi-
nation at the same time. A computer program was used to send
audio data to the serial port. A USB-to-serial converter cable
was used to connect the laptop via USB port to an RS232-to-
TTL level converter IC (MAX232). The TTL level output of
this IC was then fed to the LED driving circuitry, consisting
of MOSFETs (IRF 520). For ease of implementation, OOK-
NRZ was then used to modulate the LEDs. At the receiving
end, the desired optical signal was first selected through
wavelength filter and then received using a digital optical
receiver (TORX173)
2
. The receiver output was conditioned
using a level converter IC interfaced to the laptop where the
received signal was demodulated to recover the transmitted
data.
In order to de-multiplex the received data streams, filters
were needed at the receiver to provide wavelength-selectivity.
Instead of using expensive commercially available optical
filters, several off-the-shelf materials were tested to serve the
purpose. After a rigorous search, extremely low-cost bandpass
filters were improvised using colored plastic sheets, without
compromising the design efficiency. Each filter allowed only
a specific band of wavelengths, corresponding to a particular
color (red, green or blue), to pass through it while significantly
2
Although, this module is designed for fiber optic communication but it
can also be used for free space communications over a short range.
TABLE I
I
NSERTION LOSS FOR OPTICAL FILTERS.
Insertion Loss in dBs Red Light Blue Light Green Light
RED Filter -1.6 -43.7 -49.7
BLUE Filter -38.4 -3.6 -14
GREEN Filter -20.9 -12.6 -3.4
attenuating the undesired wavelengths which was sufficient
to provide wavelength-selectivity. In this way, red, green
or blue lights (data streams) were selected by placing the
corresponding (red, green or blue) filter at the receiver.
In order to validate our claims, we calculated the insertion
loss of the filters for each of the three light sources as shown
in Table I. Using a broadband white light LED panel as the
source, we also characterized the spectral performance of the
three optical filters. To quantify the filter performance, we
evaluated typical parameters used for filter characterization,
which include Center Wavelength (CWL), Full-Width at Half
Maximum (FWHM) and Peak Transmission (T) and the results
for these parameters are tabulated in Table II.
TABLE II
F
ILTER PERFORMANCE CHARACTERIZATION.
PARAMETERS RED BLUE GREEN
Center Wavelength (CWL) 605nm 529nm 563nm
Full Width at Half Max. (FWHM) 50nm 38nm 96nm
Peak Transmission (T) 588nm 535.5nm 572nm
These results reveal that the green filter has double the value
of FWHM compared with the red and blue filters. One may
expect it to perform a bit poorly as it might allow red or
blue light to pass. However, the green LED operates at around
565 nm, which is almost the same as the peak transmission
wavelength 563 nm of the green filter. Secondly, the operating
wavelengths of red (660 nm) and blue (450 nm) LEDs lie
beyond the half-maximum points (at 615 nm and 520 nm) of
the green filter. As a result, the possibility of interference from
red and blue data streams is minimal as both wavelengths are
sufficiently attenuated, relative to green. Thus, the undesirably
large value of FWHM for green filter does not, in effect,
compromise filter performance. In short, the performance of
the improvised wavelength-filters was noteworthy considering
the total cost of approximately 50 cents.
V. P
ERFORMANCE EVALUATION
In an indoor VLC system, some of the common sources
responsible for generating ambient noise include fluorescent
and incandescent lights, unmodulated LED lights and solar
radiations diffusing in through the windows. In addition, the
photo-detector shot noise (induced by the signal and the
ambient light) and the pre-amplifier noise are also present [3].
In an indoor environment, all the lights might not be used
for data communication. Thus, a situation may arise where
the (unmodulated) LED sources (used only for illumination)
assume the role of being a major noise contributor. With
this motivation, we also evaluated the performance of the
designed VLC system in the presence of noise sources using
the developed hardware test-bed shown in Fig. 2.
Transmitter
Module
Receiver
Module
Level Shifter
Optical
Filter
Optical
Receiver
Transmitting
LEDs
LED MOSFET
Drivers
Level Shifter
Fig. 2. Developed hardware platform for experimental evaluation.
BER Calculation
The experiments were conducted by operating the designed
system in a moderately lit room with sunlight entering through
four symmetrically distributed window panes. The transmitter
and receiver were placed in line at a fixed distance from
each other. Red and green LEDs were used at the transmitter
and two data streams were transmitted simultaneously. On
the receiver side red filter was applied and average BER was
measured for different transmission distances by moving the
transmitter (containing both transmitting LEDs) towards the
receiver along fixed axis. For each position, the direction of
the transmitter was adjusted so as to have maximum signal
strength at the receiver. It is observed that the interference
due to green LED did not result in any degradation of the
BER.
1) BER vs Distance: A new noise source, was introduced in
the system by placing an unmodulated red LED in the vicinity
of the transmitter with the receiver lying in its field of view.
For a fixed transmission distance, the position of noise source
was changed relative to the receiver and its effect on BER was
observed. The empirical data was used to plot the variations
in BER with distance of noise source from the receiver for
different transmission distances and is shown in Fig 3. As
the distance between noise source and receiver is reduced,
the average BER increases due to a decrease in the SNR
at the receiver. This can be attributed to an increase in the
noise power at the receiver for the same transmission power
as the relative positions of the transmitter and receiver do not
change. If the experiment is repeated for a greater transmission
distance, a similar trend is obtained. However, there is an
increase in BER for the same noise source positions. This is
because the received signal power reduces due to an increase in
transmission distance while the noise power remains the same,
leading to a reduction in the SNR at the receiver. Furthermore,
the curve shows a sharp transition at some points. It can be
accredited to the receiver characteristics. The receiver has a
26 28 30 32 34 36
0
0.2
0.4
0.6
0.8
1
Distance between noise source and receiver
(
cm
)
Average BER (Bit Error Rate)
transmission distance = 15 cm
transmission distance = 30 cm
Fig. 3. Average BER as a function of distance between noise source and
receiver.
minimum receivable power of −27 dBm while a maximum
receivable power of −14.5 dBm. As soon as the received
power levels cross these thresholds, the receiver malfunctions,
leading to a sudden transition in BER.
2) BER vs Noise Intensity: The second experiment was
conducted by keeping the noise source at a fixed distance from
the receiver. The transmitter and receiver were also placed
a fixed distance apart. The noise intensity was changed by
changing the current supplied to the red LED acting as the
noise source. For a given value of current, average BER was
computed as done in the first experiment. The variation in
BER vs noise intensity is shown in Fig 4. As the intensity of
1 1.1 1.2 1.3 1.4 1.5 1.6 1.7 1.8
0
0.5
1
Current of noise source (A)
Average BER
0 1 2 3 4 5 6
0
0.5
1
Normalized intensity of noise source (dB)
Average BER
BER vs current
BER vs noise intensity
Fig. 4. Average BER as a function of noise intensity.
the noise source is increased, the average BER increases due
to a decrease in the SNR at the receiver; the noise power at
the receiver increases while the signal power remains the same
because both the transmission power and the relative positions
of the transmitter and receiver are kept constant.
VI. C
ONCLUSION AND FUTURE WORK
The use of wavelength division multiplexing was suc-
cessfully demonstrated by using colored LEDs for trans-
ferring multiple data streams simultaneously in an indoor
VLC system. Owing to the inherent frequency gap between
the wavelengths used, data streams were transmitted with
negligible interference. Extremely low-cost bandpass filters
were improvised using off-the-shelf colored plastic sheets.
The performance of the designed system was evaluated using
empirical data. The proposed scheme has the potential to
provide robust high data-rate communication in the presence
of frequency selective noise/interference sources compared to
traditional VLC systems. In future, we want to extend our
work to VLC systems involving plural multi-color LEDs, with
a focus on optimum mixer design.
A
CKNOWLEDGEMENT
This work is supported in part by Higher Education Com-
mission of Pakistan and the graduate research program of UET
Lahore.
R
EFERENCES
[1] E. Schubert and J. Kim, “Solid-state light sources getting smart,”
Science, vol. 308, no. 5726, p. 1274, 2005.
[2] M. Nakagawa, “Visible light communications,” in Conference on Lasers
and Electro-Optics/Quantum Electronics and Laser Science Conference
and Photonic Applications Systems Technologies, 2007.
[3] T. Komine and M. Nakagawa, “Fundamental analysis for visible-light
communication system using LED lights,” IEEE transactions on Con-
sumer Electronics, vol. 50, no. 1, p. 100, 2004.
[4] S. Haruyama, “Visible light communications,” in Optical Communica-
tion (ECOC), 2010 36th European Conference and Exhibition on, pp. 1–
22, IEEE.
[5] N. Farr, A. Bowen, J. Ware, C. Pontbriand, and M. Tivey, “An integrated,
underwater optical/acoustic communications system,” in OCEANS 2010
IEEE-Sydney, pp. 1–6, IEEE, 2010.
[6] S. Arai, S. Mase, T. Yamazato, T. Endo, T. Fujii, M. Tanimoto,
K. Kidono, Y. Kimura, and Y. Ninomiya, “Experimental on hierarchical
transmission scheme for visible light communication using led traffic
light and high-speed camera,” in Vehicular Technology Conference,
2007. VTC-2007 Fall. 2007 IEEE 66th, pp. 2174–2178, IEEE, 2007.
[7] H. Le Minh, D. O’Brien, G. Faulkner, L. Zeng, K. Lee, D. Jung,
and Y. Oh, “High-speed visible light communications using multiple-
resonant equalization,” Photonics Technology Letters, IEEE, vol. 20,
no. 14, pp. 1243–1245, 2008.
[8] A. B. Siddique and M. Tahir, “Joint brightness control and data trans-
mission for visible light communication systems based on white leds,” in
Consumer Communications and Networking Conference (CCNC), 2011
IEEE, pp. 1026 –1030, jan. 2011.
[9] M. Crawford, “Leds for solid-state lighting: Performance challenges and
recent advances,” Selected Topics in Quantum Electronics, IEEE Journal
of, vol. 15, no. 4, pp. 1028–1040, 2009.
[10] J. Kahn and J. Barry, “Wireless infrared communications,” Proceedings
of the IEEE, vol. 85, no. 2, pp. 265–298, 1997.
[11] K. Lee, H. Park, and J. Barry, “Indoor channel characteristics for visible
light communications,” Communications Letters, IEEE, vol. 15, no. 2,
pp. 217–219, 2011.
[12] D. OBrien, L. Zeng, H. Le-Minh, G. Faulkner, S. Randel, J. Walewski,
J. Borges, K. Langer, J. Grubor, K. Lee, et al., “Visible-light communica-
tions,” Short-Range Wireless Communications: Emerging Technologies
and Applications. R. Kraemer and M. Katz, eds. New York, NY: Wiley,
2009.
[13] K. Cui, G. Chen, Z. Xu, and R. 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.
[14] L. Zeng, D. O’Brien, H. Minh, G. Faulkner, K. Lee, D. Jung, Y. Oh, and
E. Won, “High data rate multiple input multiple output (mimo) optical
wireless communications using white led lighting,” Selected Areas in
Communications, IEEE Journal on, vol. 27, no. 9, pp. 1654–1662, 2009.
[15] W. Jeung, S. Shin, S. Hong, S. Choi, S. Yi, Y. Yoon, H. Kim, S. Lee,
and K. Park, “Silicon-based, multi-chip led package,” in Electronic
Components and Technology Conference, 2007. ECTC’07. Proceedings.
57th, pp. 722–727, IEEE, 2007.