COWS 
Project goals and Objectives
The COWS
project will investigate the potential of coherent
detection as a means to improve coverage, mobility
and data rate of indoor optical wireless systems.
The poor link budget of conventional intensity
modulation/direct detection (IM/DD) optical wireless
systems necessitates the existence of a
line-of-sight path between the transmitter and the
receiver. This renders the link susceptible to
obstacles such as people moving around the room,
furniture, etc, reduces network coverage and limits
user mobility. Coherent detection provides an
inherent receiver gain and can therefore be used in
order to detect signals carried from optical beams
that do not originate rely on line‑ reflected from
the walls or the ceiling of the room (diffuse
propagation regime).
Due to the
similarity of the coherent optical channel and the
radio channel and more specifically the fact the
optical phase information is retained, coherent
detection can also open up intriguing possibilities
for increasing the link capacity using
multiple-input-multiple-output (MIMO) techniques.
This is in contrast to IM/DD systems, where the
phase information is lost and space time block
coding schemes (STBC) are not very effective. COWS
will investigate the applicability of MIMO
techniques in coherent optical wireless links as a
means to a) improve the capacity and b) improve the
coverage of the system.
Besides
the link budget, multipath dispersion is a major
concern in the diffuse optical channel and can limit
the achievable data rate due to intersymbol
interference. Orthogonal frequency domain
multiplexing (OFDM) can compensate the impulse
response and increase the transmission bit rate.
COWS will investigate the performance limitations
imposed by the frequency selective channel and
consider the enhancement brought about through the
use of OFDM or other channel equalization
techniques.
The
objectives of the project are summarized below:
Infrared
optical wireless may provide one of the cornerstones
of future local area networks since they combine a
vast unregulated emission spectrum, multi
gigabit-per-second data rates, zero radio
interference and negligible biological tissue
interaction. COWS attempts to add another two
appealing features and namely enhanced coverage and
mobility both achieved through the internal gain of
the coherent receiver which can be used to detect
signals even in the absence of a line-of-path. These
extra features will enable optical wireless to
become an ultra broadband successor of WiFi and a
key technology for high speed local area networks,
facilitating the penetration of broadband e-services
and applications. COWS technology will provide
fiber-like data rates in indoor environments without
the need for installing new data cables. This is a
crucial factor in home network installations,
especially in existing dwellings where users like to
avoid such costly and annoying tasks as much as
possible. In addition the increased receiver
sensitivity can be used to relax the transmission
power, thereby leading to less power consumption by
the optoelectronic circuits.
Short
reach communications is another driver for COWS.
Optical wireless can provide a wireless alternative
to high speed wireline interfaces such as universal
serial bus (USB) or fiber channel. One can envisage
such systems providing point-to-point or
point-to-multipoint connections and connecting
several devices such as printers, laptop and
television sets in a room. COWS can serve as a
primary or backup connection in server racks and
storage area networks. It can also provide a
backbone for small to medium size computer clusters.
On a
regional scale, COWS will allow the strengthening of
the existing research work carried out at Harokopio
University of Athens in the field of optical
wireless technology and networks. The project will
allow the hiring of experienced post‑doc researchers
and the support of PhD students in order promote
excellence in an emerging and rapidly progressing
field of science and technology. Synergies with
other optical wireless groups both in Greece and
abroad will be sought. The COWS project is a first
class opportunity to strengthen existing research
collaborations and know-how exchange with top
research groups such as the Fraunhofer Heinrich
Hertz Institute, France Telecom, University of
Oxford, etc. On the other hand, COWS will act as a
vehicle to solidify the Greek optical wireless
research community which is considerable which
includes the Aristotle University of Thessaloniki,
the University of Athens and the University of
Peloponnese.
Existing Knowledge - Project innovation and
originality
Optical
Wireless Communications: Need and state-of-the-art
The
proposed project deals with two of the most
challenging contemporary research areas, namely
broadband and wireless communication technologies.
Very Fast Internet has been identified as the one of
the 8 pillars for the Digital Europe 2020, reported
by European Commission. Services ranging from high
definition television to videoconferencing it is
planned to use rely on much faster internet access
than it is currently available in Europe. In order
to compete with world leaders like South Korea and
Japan, Europe aims to download rates of 30 Mbps for
all of its citizens and at least 50% internet
connections at least 100 Mbps for all European
households by 2020. In this direction, among the EC
planned actions is the adaption of an EU broadband
communication, funding for high-speed broadband, and
facilitation of broadband investment by the member
states. In order to enter this new fast internet era
broadband and wireless communication are key
enabling technologies. It comes as no surprise that
during the first assembly for the digital agenda
2020 [1],
two of the workshops that took place were dedicated
to such technologies. More precisely, the “Spectrum
for wireless innovation in Europe”
[2]
and the “Financing and facilitating broadband
projects” workshops have gained the attendees
attention
[3].
Figure
1
shows the evolution of data transmission
technologies, either wireless or wireline in the
access and home network. It is easily seen that
gigabit-per-second can be supported up to the
customer premises using optical fiber-to-the-home
(FTTH) access technologies like gigabit passive
optical networks (GPON) [4],[5].
The figure also suggests that radio-based wireless
technologies such as 802.11n and ultra wide band
(UWB) cannot achieve gigabit-per-second connectivity
at the customer premises. Various alternatives to
WiFi are currently being sought. Wired alternatives
include multimode or plastic fiber systems, power
line communications and unshielded twisted pair
(UTP) Ethernet. Wireline alternatives can lead to
the well known cable “spaghetti” (mixing of cables)
which must be avoided in domestic environmenents for
both practical and aesthetic reasons. They also
provided limited user mobility compared to wireless
alternatives. The fiber-based alternatives provide
very high data rates [6]
but need to be installed from scratch in existing
dwellings, implying increased cost and user
discomfort. This is also true for UTP-based Ethernet
which operate at lower data rates and smaller
distances. Power line communications
[7]
use the existing power cables but are generally
limited to data rates below 1Gb/s. Wireless
technologies currently being pursued include 60GHz
millimeter-wave [8]
and terahertz systems [9].
60GHz systems take advantage of the 57-64GHz
unlicenced RF spectrum . Low power 60GHz
transceivers have been implemented operating at
3.5Gb/s on a standard CMOS platform
[8].
Similar data rates can be achieved at the THz
spectrum using a resonant tunneling diode
transmitter
[9].
Both systems rely on a line-of-sight path between
the transmitter and the receiver, unlike
conventional 802.11 networks.
Another
alternative is to use optical wireless technology,
which in essence means transmitting light signals
without any waveguiding medium such as an optical
fiber. Taking advantage of the significant technical
know-how in optical transceivers, line-of-sight
indoor optical wireless systems operating at the
infrared regime can support 10Gb/s data rates
[11].
Since the infrared communication spectrum is
practically unlimited (0.7μm-1.6μm), it is possible
to envision a multitude of optical wireless systems
operating in the same room transmitting at a
different wavelength. Besides the vast available
spectrum and the large data rates, optical wireless
systems exhibit zero interference with existing
radio systems and provided some eye safety
precautions are met, optical wireless are known to
have negligible biological interaction, unlike THz
or 60GHz radiation where this still remains a much
debated subject. Outdoor optical wireless links,
a.k.a. free space optical systems have already been
commercialized and can support multigigabit
connectivity. In indoor environments, optical
wireless systems are also being considered for high
bandwidth connectivity.
Figure 1:
The evolution of access technologies (source: Orange
Labs white paper)
Optical
wireless systems usually consist of transceivers
operating at the infrared (IR) part of the spectrum.
Figure
2
illustrates four different variations of the two
basic transmitter/receiver arrangements.
Figure 2(a)
depicts a truly diffuse system, where no direct
line-of-sight path exists and communication is
achieved through beam reflections at the various
surfaces of the room (walls, ceiling, etc). Figure
2(b)
depicts a quasi diffuse arrangement. The transmitter
points towards the ceiling and the terminals are
illuminated through reflection of the beam through
the ceiling or some other reflecting surface.
Figure
2(c)
shows a wide line-of-sight system (WLOS), where the
transmitter beam is wide enough to cover a
relatively large portion of the room. Figure
2(d)
is a narrow line of sight (NLOS) system, where the
transmitter beam is narrow and covers only a small
portion of the room.
Figure
2:
Various IR transmitter/receiver configurations
Diffuse
and quasi diffuse systems have a stringent link
budget since light reaches the receiver through
(possibly) multiple reflections which significantly
attenuate the signal strength. In addition, there
can be more than one optical paths, giving rise to
multipath dispersion, corresponding to a frequency
selective channel. At high data rates, this implies
intersymbol interference (ISI) which can
significantly degrade the achievable bit error rate
(BER) at the receiver side [4]. Quasi diffuse links
are usually less impaired by ISI than diffuse links.
Multipath dispersion can be present in WLOS as well.
The fact that the beam is wide enough may result in
portions of the signal arriving after a single or
multiple reflections at the various surfaces of the
room. In NLOS systems, multipath fading is
negligible and the link budget poses less stringent
restrictions on transmitter optical power. LOS
configurations are susceptible to blocking from
various obstacles which can accidentally interrupt
the line-of-sight path. Diffuse links, are more
robust, since they do not rely on line-of-sight.
The first
IR system was demonstrated by Gfeller and Bapst
[10]
and was a diffuse link operating at 1Mbit/s. A
diffuse system operating at 50Mbit/s was later
demonstrated using IM/DD
[11].
In later demonstrations and because of the multipath
dispersion and the poor link budget which limit the
system performance, attention has recently been
shifted towards line-of-sight systems. Recently, the
ICT-OMEGA project demonstrated a multiple element
transceiver operating at 1.25Gbit/s
[12]
using low-cost commercial of the shelf components.
Recently 12.5Gb/s connectivity
has been demonstrated with the signal being fed by a
remote central office node (~6Km distance) without
any optoelectronic conversion at the room hotspot
[13].
Coherent
detection in optical communications
From the
perspective of coherent optical detection in optical
communications, the first wave of the intense
interest appeared in the 1980s and early 1990s, when
the coherent detection was viewed as a promising
technique to improve the receiver sensitivity
[14].
The earlier incarnations of coherent receivers were
complicated, and typically included high speed
analog electronics for demodulation from the carrier
or for optical phase locking, and also active
optical components for polarization control.
However, the ensuing invention of the erbium doped
fiber amplifier (EDFA) reduced research on coherent
communication to peripheral interest. Recent years
have seen a revival of coherent detection schemes in
fiber-based systems where the bandwidth constraints
of optical amplifier and ultimately the fiber
itself, it is important to maximize the spectral
efficiency but at the same time keep the transmitted
power low enough in order to prevent signal-to-noise
ratio degradation due to fiber nonlinearity
[15].
In optical wireless communications where small size,
low cost components are needed, the use of optical
amplifiers such as EDFAs or semiconductor optical
amplifiers (SOA) is prohibited. In recent
implementations of coherent receiver, the
functionalities of phase locking and polarization
control are performed in the digital domain using
digital signal processing (DSP). Provided that the
DSP hardware and also the necessary optical passive
components can be produced in volume at low cost,
the new version of the coherent receiver will be
cost-effective compared to direct detection, and is
likely to be widely deployed.
To
compensate for the effect of the frequency selective
channel, equalization methods such as OFDM can be
applied. OFDM has emerged as the leading
physical-layer interface in wireless communications
in the past decade. It is a special form of a
broader class of multicarrier modulation where a
data stream is carried with many lower-rate
subcarrier tones
[16].
OFDM has been widely studied in mobile
communications to combat hostile frequency-selective
fading and has been incorporated into wireless
network standards (802.11a/g WiFi, HiperLAN2, 802.16
WiMAX) and digital audio and video broadcasting (DAB
and DVB-T) in Europe, Asia, Australia, and other
parts of the world. The synergies between coherent
optical communications and OFDM are twofold. The
coherent system brings OFDM a much needed linearity
in RF-to-optical (RTO) upconversion and
optical-to-RF (OTR) downconversion. OFDM brings
coherent system computation efficiency and ease of
channel and phase estimation. The complementary
metal–oxide semiconductor (CMOS)
application-specific integrated circuit (ASIC) chips
recently demonstrated for single carrier coherent
systems
[17]
signify that the current silicon speed can support
40 Gbit/s OFDM transmission systems. Because of its
superior scalability with the bit rate of the
transmission systems, coherent OFDM (CO-OFDM) is
well-positioned to be an attractive choice of
modulation format for the next generation of 100
Gbit/s wireline transmission.
The COWS
approach: Bringing coherent detection in the optical
wireless realm.
The above
points illustrate the potential advantages of
CO-OFDM in diffuse optical wireless links. The
inherent internal gain of the coherent receiver
offers a promising alternative for improving the
link budget. OFDM can be used to provide channel
equalization and mitigate the frequency selective
nature of the diffuse optical wireless channel.
Figure 3
shows the COWS link which takes advantage of both
technologies in order to combine the optical
wireless high data rates with the mobility and room
coverage of a Wi‑fi network.
As seen in
the figure, the input data pass through a serial to
parallel converter and then to a subcarrier symbol
mapper, which maps the incoming bits to transmitted
symbols. The symbol stream the pass though an
inverse discrete Fourier transform (IDF) which
estimates the OFDM waveform samples. A guard
interval (GI) is inserted in order to ensure that
subsequent symbols do not interfere. The
digital-to-analog converters (D/A) converters
synthesize the real (Re) and the imaginary (Im) part
of the ODFM waveform which is fed to the
Mach-Zehnder modulators (MZM) which imprint the
waveform in the carrier emitted by the transmitter
laser diode. At the receiver the optical signal is
mixed with the signal of a local oscillator laser
diode at the four photodiodes which are
differentially placed in two group, one detecting
the real and the other the imaginary part of the
signal. These signals are fed at the
analog-to-digital (A/D) converters and pass through
the discrete Fourier transform (DFT) module and
after removing the guard interval, the subcarrier
symbols and the corresponding output are estimated.
Figure 3:
The COWS link. S/P: serial-to-parallel, GI: guard
interval, D/A: digital-to-analog converter, DFT:
discrete Fourier transform, LPF: low pass filter,
MZM: Mach Zenhder Modulator, PD: photodiode, LD:
laser diode.
COWS
Contribution to the current state-of-the-art.
The
objective of COWS is to demonstrate the merits of
optical coherent detection as a cornerstone of
future wireless networks. More specifically, the
project will advance the current state of the art in
the following ways:
(a)
Provide a proof-of-concept on the applicability of
coherent detection in diffuse optical wireless
systems in order to provide gigabit-per-second
connectivity without the need of a direct optical
path. This will constitute a reliable choice for
next generation home networks, combining the high
data rates of optical technologies and WiFi-like
flexibility and mobility. The proof-of-concept
entails experimental demonstration of several key
aspects of the system. It should be noted that
coherent detection has only been considered
theoretically in line-of-sight indoor optical
wireless systems assuming a flat channel [18].
As a consequence, COWS is expected to advance the
state-of-the-art of optical wireless both on a
theoretical and a numerical basis and in addition
provide experimental validation.
(b)
Investigate suitable techniques for compensating
multipath-induced distortion. Coherent optical
frequency domain multiplexing (CO-OFDM) is a
suitable choice for such applications since they can
be implemented using DSP electronics. Other choices
to be considered are distributed feedback
equalization (DFE), linear equalization (LE) and
frequency domain equalization (FDE). Bit loading at
a subcarrier level can also increase the net
transmission rate. The objective of COWS is to
provide a systematic and thorough comparison of
these techniques and analyze the pros and cons of
each approach both on a performance and on an
implementation cost level. These issues have not
been addressed in the literature and are expected to
provide a clear advancement in the current state of
the art in the case of both diffuse and line of
sight systems.
(c)
Identify the best possible design for system
implementation. The choice of LO laser type is of
paramount importance since a low cost and low power
technology such as vertical-cavity surface-emitting
laser (VCSEL) should be preferred. The receiving
photodiode type is another important issue to
consider, since although emphasis should be placed
on low cost p-i-n diodes, silicon avalanche
photodetectors (APD) should also be considered as
they provide internal gain. The choice of the
transmission wavelength is also an important matter,
favoring the 800nm band where components are much
cheaper. In choosing the right wavelength, certain
eye-safety and skin-safety considerations should be
taken into account however. Finally, the DSP circuit
characteristics such as analog-to-digital sampling
rate must be chosen wisely so as to keep the cost as
low as possible. All these aspects have been
previously considered in fiber-based systems, but
given the cost sensitive nature of the application,
the findings cannot be applied directly and can only
serve as a starting guide.
(d) Study
the applicability of MIMO techniques in diffuse
CO-OFDM systems in an attempt to provide increased
capacity and coverage. Preliminary calculations show
that this is a promising approach in line-of-sight
coherent optical wireless links but the results need
to be extended to include the intricacies of the
diffuse wireless channel response and the use of
multiple subcarrier modulation schemes such as OFDM.
This is an exciting area of research and a terra
incognita since little is actually known for the
applicability of MIMO techniques is optical
wireless.
