See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/312133486

Overview of Wireless Optical Communication Systems

Chapter · January 2017

DOI: 10.1007/978-81-322-3691-7_1

CITATION
S READS

14 20,762

3 authors:

Hemani Kaushal
Vk Jain
University of North Florida
Indian Institute of Technology Delhi
70 PUBLICATIONS 4,126
CITATIONS 167 PUBLICATIONS 1,728 CITATIONS

SEE PROFILE SEE PROFILE

Subrat Kar
Indian Institute of Technology Delhi
256 PUBLICATIONS 2,413 CITATIONS

SEE PROFILE

All content following this page was uploaded by Hemani Kaushal on 12 February 2018.

The user has requested enhancement of the downloaded file.

Chapter 1
Overview of Wireless Optical Communication
Systems

1.1 Introduction

WOC communication is considered as the next frontier for high-speed broadband

connection due to its unique features: extremely high bandwidth, ease of deploy-
ment, tariff-free bandwidth allocation, low power (~1/2 of radio-frequency (RF)
systems), less mass (~1/2 of RF systems), small size ( 1/10 the diameter of
RF antenna), and improved channel security. It has emerged a good commercial
alternative to existing radio-frequency communication as it supports larger data
rates and provides high gain due to its narrow beam divergence. It is capable
of transmitting data up to 10 Gbps and voice and video communication through
the atmosphere/free space. WOC have two broad categories, namely, indoor and
outdoor wireless optical communications. Indoor WOC is classified into four
generic system configurations, i.e., directed line-of-sight (LOS), non-directed
LOS, diffused, and quasi diffused. Outdoor wireless optical communication is also
termed as free-space optical (FSO) communication. The FSO communication
systems are also classified into terrestrial and space systems. Figure 1.1 shows the
classification of WOC systems.
Over the last few years, massive expansion in WOC technology has been
observed due to huge advances in optoelectronic components and tremendous
growth in the market offering wireless optical devices. It seems to be one of the
promising technologies for addressing the problem of huge bandwidth
requirements and “last mile bottleneck.” There are many commercial applications
of WOC technology which includes ground-to-LEO, LEO-to-GEO/LEO-to-
ground, GEO- to-ground, LEO/GEO-to-aircraft, deep space probes, ground
stations, unmanned aerial vehicles (UAVs), high-altitude platforms (HAPs), etc.
[1–4]. It also finds applications in the area of remote sensing, radio astronomy,
space radio commu- nication, military, etc. When WOC technology is used over
very short distances, it is termed as FSO interconnects (FSOI), and it finds
applications in chip-to- chip or board-to-board interconnections. FSOI has
gained popularity these days

© Springer (India) Pvt. Ltd. 2017 1

H. Kaushal et al., Free Space Optical Communication, Optical Networks,
DOI 10.1007/978-81-322-3691-7_1
2 1 Overview of Wireless Optical Communication Systems

Fig. 1.1 Classification of wireless optical communication systems

as it potentially addresses complex communication requirement in optoelectronic

devices. This technology offers the potential to build interconnection networks
with higher speed, lower power dissipation, and more compact packages than
possible with electronic very large-scale integration (VLSI) technology. However,
the cost of optoelectronic devices, their integration, and overall packaging makes
FSOI a costly affair. A throughput upto 1 Tbps per printed circuit board (PCB) has
been experimentally demonstrated in [5] using 1000 channels per PCB with 1 mm
optical beam array at 1 Gbps per channel.
Based on their transmission range, WOC can be classified into five broad
categories (refer Fig. 1.2):
(i) Ultrashort-range WOC – used in chip-to-chip communication or all optical
lab-on-a-chip system.
(ii) Short-range WOC – used in wireless body area networks (WBANs) or
wireless personal area networks (WPANs).
(iii) Medium-range WOC – used in indoor IR or visible light communication
(VLC) for wireless local area networks (WLANs) and inter-vehicular and
vehicle-to-infrastructure communications.
(iv) Long-range WOC – used in terrestrial communication between two buildings
or metro area extensions.
(v) Ultra-long-range WOC – used in ground-to-satellite/satellite-to-ground or
inter-satellite link or deep space missions.
Commercially available FSO equipment provide much higher data rates
ranging from 10 Mbps to 10 Gbps [10, 11]. Many optical companies like
LightPointe in San Diego, fSONA in Canada, CableFree Wireless Excellence in
UK, AirFiber in
1.1 Introduction 3

Fig. 1.2 Applications of WOCs: (a) chip-to-chip communication, (b) wireless body area network,
(c) indoor IR or visible light communication, (d) inter-building communication, and (e) deep
space missions [6–9]

California, etc. provide a wide range of wireless optical routers, optical wireless
bridges, hybrid wireless bridges, switches, etc. that can support enterprise connec-
tivity, last mile access, and HDTV broadcast link with almost 100 % reliability in
adverse weather conditions.
4 1 Overview of Wireless Optical Communication Systems

1.1.1 History

The first experiment of transmitting signal over the atmosphere was conducted
by Alexander Graham Bell in 1880. He used sunlight as a carrier to transmit
voice signal over a distance of about a few feet. However, the experiment was not
successful due to inconsistent nature of the carrier. Later, in the 1960s, Theodore
H. Maiman discovered the first working laser at Hughes Research Laboratories,
Malibu, California. From this point onward, the fortune of FSO has changed. Vari-
ous experiments were conducted in military and space laboratories to demonstrate
FSO link. In the 1970s, the Air Force sponsored a program known as Space Flight
Test System (SFTS) to establish satellite to ground link at Air Force ground
station, New Mexico. The program was later renamed as Airborne Flight Test
System. This program achieved its first success in the 1980s where a data rate of 1
Gbps was demonstrated from aircraft to ground station. After this, a flurry of
demonstrations were recorded during the 1980s and 1990s. They include Laser
Cross-Link Sub- system (LCS), Boost Surveillance and Tracking System (BSTS),
Follow-On Early Warning System (BSTS), and many more [12]. A full duplex
ground to space laser link known as Ground/Orbiter Laser Communication
Demonstration was first established in 1995–1996 by National Aeronautics and
Space Administration (NASA) in conjunction with Jet Propulsion Laboratory
(JPL). In addition to this, various demonstrations were carried out for deep space
and inter-satellite missions such as Mars Laser Communication Demonstration
(MLCD) and Space Inter- satellite Link Experiment (SILEX) [12], respectively.
Very large-scale development is carried out by NASA in the USA, Indian
Space Research Organisation (ISRO) in India, European Space Agency (ESA) in
Europe, and National Space Development Agency (NSDA) in Japan.
Demonstrations have established a full duplex FSO link with high data rates
between various onboard space stations and ground stations, inter-satellite, etc.
with improved reliability and 100 % availability. Besides FSO uplink/downlink,
extensive research is carried out for FSO terrestrial links, i.e., link between two
buildings to establish local area network segment that will provide last mile
connectivity to the users (Fig. 1.3).
FSO communication is well suited for densely populated urban areas where
digging of roads is cumbersome. Terrestrial FSO links can be used either for short
range (few meters) or long range (tens of km). Short-range links provide high-
speed connectivity to end users by interconnecting local area network segments
that are housed in building separated within the campus or different building of
the company. Long-range FSO communication links extend up to existing
metropolitan area fiber rings or to connect new networks. These links do not
reach the end user but they extend their services to core infrastructure. FSO
communication system can also be deployed within a building, and it is termed as
indoor wireless optical communication (WOC) system. This short-range indoor
WOC system is a futuristic technology and is gaining attention these days with
the advancement of technology involving portable devices, e.g., laptops, personal
digital assistants, portable telephonic devices, etc.
1.1 Introduction 5

Fig. 1.3 FSO terrestrial link [13]

1.1.2 Indoor Wireless Optical Communication

Indoor wireless optical communication links provide a flexible interconnection

within a building where setting up a physical wired connection is cumbersome.
It consists of lasers or light-emitting diodes as transmitter and photodetectors as
the receiver. These devices along with their drive circuits are much cheaper as
compared to radio-frequency equipment or existing copper cables. Further, indoor
WOC is inherently secure technology since the optical signals do not penetrate
walls unlike electromagnetic waves which can cause interference and thus
provides a high degree of security against eavesdropping. These optical waves
are either in the visible light spectrum or in the IR spectrum which is able to
provide very large (THz) bandwidth. Since these devices consume very little
power, they are also suitable for mobile terminal systems. Besides many
advantages, indoor optical wireless system is influenced by various impairments
that impact the performance of the communication system. Some of the factors that
lead to these impairments are
(i) limiting speed of optoelectronic devices; (ii) large path loss; (iii) noisy indoor
environment due to incandescent, fluorescent lighting or sunlight that contributes
to noise in the detector; (iv) multipath dispersion; and (v) interference due to
artificial noise sources. The range of the system is restricted as the average
transmitted power is limited due to eye safety regulations [14].
6 1 Overview of Wireless Optical Communication Systems

The most commonly used optical sources in IR transmitters for indoor WOC
are light-emitting diodes (LEDs) and laser diodes (LDs). LEDs are preferred over
LDs as they are cheap and have broader modulation bandwidth and linear
characteristics in the operating region. Since LEDs are nondirectional optical
sources, their output power is not very high. Therefore, to compensate for lower
power levels, an array of LEDs can be used. However, LEDs cannot work at high
data rates beyond 100 Mbps. Laser diodes can be used at data rates of the order of
few Gbps. Due to eye safety regulations, laser diode cannot be used directly for
indoor WOC as they are highly directional and can cause optical damage.

1.1.2.1 Types of Link Configurations

The classification of an indoor optical link depends upon two major factors: (i)
transmitter beam angle, i.e., degree of directionality, and (ii) the detector’ s field-
of-view (FOV), i.e., whether the view of the receiver is wide or narrow. Based on
this, there are mainly four types of link configuration, i.e., directed line-of-sight
(LOS), non-directed LOS, diffused, and multi-beam quasi diffused links.
(i) Directed LOS Link: In this type of link, the beam angle of transmitter
as well as FOV of receiver are very narrow. The transmitter and receiver
are directed toward each other. This configuration is good for point-to-point
link establishment for indoor optical communication. The advantages and
disadvantages of directed LOS link are as follows:
Advantages:
• Improved power efficiency as path loss is minimum
• Reduced multipath distortion
• Larger rejection of ambient background light
• Improved link budget
Disadvantages:
• Links are highly susceptible to blocking (or shadowing), and therefore,
they cannot provide mobility in a typical indoor environment.
• Reduced flexibility as it does not support point-to-multipoint broadcast
links.
• Tight alignment between transmitter and receiver is required making it
less convenient for certain applications.
Figure 1.4 shows the pictorial representation of the link. This configuration
has been used for many years at low data rate for electronic appliances
using remote control applications like television or audio equipments. It
provides point-to-point connectivity between portable electronic devices such
as laptops, mobile equipments, PDAs, etc. Depending upon the degree of
directionality, there is another version of LOS link called hybrid LOS. In
this case, transmitter and receiver are facing each other, but the divergence
1.1 Introduction 7

Fig. 1.4 Directed LOS link [14]

TX

RX

angle of transmitter is much larger than FOV of receiver. This configuration

provides larger coverage area than directed LOS, but at the cost of reduced
power efficiency, and it also suffers from blocking problems.
(ii) Non-directed LOS Link: In this type of link, the beam angle of transmitter
and detector FOV is wide enough to ensure the coverage throughout the
indoor environment. Such links do not require tight pointing and alignment
as compared to directed LOS. However, in this case, the received irradiance
is reduced for a given link distance and transmitted power. This link is suited
for point-to-multipoint broadcast applications since it provides the desired
high degree of mobility. In case of larger room dimensions, the entire room
can be divided into multiple optical cells, and each cell is controlled by a
separate transmitter with controlled beam divergence. The advantages and
disadvantages of this link configuration are as follows:
Advantages:
• Allow high user mobility
• Increased robustness against shadowing
• Alleviate the need of pointing
• Well suited for point-to-multipoint broadcast applications
Disadvantages:
• The received signal suffers from multipath distortion as the beam gets
reflected from walls or other objects in the room due to wider beam
divergence
• Less power efficient
This link provides connectivity up to tens of meters and supports data rate up
to 10 Mbps which can be shared among users operating within the cell.
Figure 1.5 depicts a non-directed LOS link.
(iii) Diffused Link: In this type of link, the transmitter is facing the ceiling/roof,
and it emits a wide beam of IR energy toward the ceiling. The IR signal
after it undergoes multiple reflections from walls or room objects is collected
by a receiver placed on the ground with wider FOV. Its advantages and
disadvantages are as follows:
8 1 Overview of Wireless Optical Communication Systems

Fig. 1.5 Multi-beam

non-directed LOS link [14] TX

RX

Fig. 1.6 Diffused link [14]

TX RX

Advantages:
• There is no requirement of alignment between transmitter and receiver as
the optical signal is uniformly spread within the room by making use of
reflective properties of walls and ceilings.
• This link is most robust and flexible as it is less prone to blocking and
shadowing.
Disadvantages:
• Severe multipath distortion
• High optical path loss typically 50–70 dB for a link range of 5 m [15].
Diffuse link is the link configuration of choice for IEEE 802.11 infrared
physical layer standard. It can support data rates up to 50 Mbps and is
typically as shown in Fig. 1.6.
(iv) Multi-Beam Quasi Diffused Link: In this type of link, a single wide beam
diffuse transmitter is replaced by multi-beam transmitter, also known as
quasi diffused transmitter. The multiple narrow beams are pointed outward
in different directions. These optical signals are collected by angle diversity
receiver [16, 17] placed on the ground. Angle diversity of the receiver can
be achieved in two ways: First is to employ multiple non-imaging receiving
elements oriented in different directions, and each element is having its own
lensing arrangement/concentrators. The purpose of the lensing arrangement
1.1 Introduction 9

or the concentrators is to improve the collection efficiency by transforming the

light rays incident over a large area into a set of rays that emerge from a
smaller area. This allows the usage of smaller photodetectors with lesser cost
and improved sensitivity. However, this approach is not a good choice as it
will make the receiver configuration very bulky. Second is by using the
improved version of angle diversity, also called “fly-eye receiver” [18]. It
consists of imaging optical concentrator with a segmented photodetector
array placed at its focal plane. In both the cases, the photocurrent generated
by the individual receiver is amplified and processed using various
combining techniques. Various advantages and disadvantages of this link are
[17]:
Advantages:
• Provides high optical gain over wide FOV
• Reduced effect of ambient light source
• Reduced multipath distortion and co-channel interference
• Immunity against blockage near receiver
• Reduced path loss
Disadvantages:
• Complex to implement
• Costly
In Fig. 1.7 quasi diffused links with multiple and single lens arrangement are
depicted. Experimental studies have shown that this link can support data rate
up to 100 Mbps.
Various organizations have implemented indoor WOC systems. The chronology
of indoor optical wireless research is given in Table 1.1.

a b

Fig. 1.7 Multi-beam quasi diffused links. (a) Receiver with multiple lens arrangement. (b)
Receiver with single lens arrangement
10 1 Overview of Wireless Optical Communication Systems

Table 1.1 Chronology of indoor optical wireless research [19]

Date Organization Configuration Bit rate Characteristics
1979/1981 IBM Diffuse 64–125 kbps 100 mW, 950 nm, BPSK
1983 Fujitsu LOS 19.2 kbps 15 mW, 880 nm, FSK
1985 Hitachi Hybrid 0.25–1 Mbps 300 mW, FSK
1985 Fujitsu Hybrid 48 kbps 880 nm, BPSK
1987 Bell Labs Directed LOS 45 Mbps 1 mW, 800 nm, OOK
1988 Matsushita Hybrid 19.2 kbps 880 nm, FSK
1992 MPR Teltech Ltd Non-directed 230 kbps DPSK, 800/950 nm
1994 Berkeley Diffused 50 Mbps 475 mW, 806 nm, OOK

Table 1.2 Comparison of indoor WOC and Wi-Fi systems

S.No. Property Wi-Fi radio IR/VLC Implication for IR/VLC
1 Spectrum licensing Yes No Approval not required
World wide compatibility
2 Penetration through walls Yes No Inherently secure
Carrier reuse in adjacent rooms
3 Multipath fading Yes No Simple link design
4 Multipath dispersion Yes Yes Problematic at high data rates
5 Dominant noise Other users Background Short range
light

Most of the indoor wireless optical links operate in 780–950 nm range. For this
reason, indoor wireless systems is also called IR systems. For directed and non-
directed LOS, typically one LED is used that emits an average power of tens of
mW. In case of diffused configuration, an array of LEDs oriented in different
directions are used so as to provide flexibility in coverage area. These LEDs
transmit power in the range of 100–500 mW. The comparison of indoor WOC and
Wi-Fi systems is given in Table 1.2.

1.1.3 Outdoor/Free-Space Optical Communication

Free-space optical communication requires line-of-sight connection between trans-

mitter and receiver for propagation of information from one point to another.
Here, the information signal from the source is modulated on the optical carrier,
and this modulated signal is then allowed to propagate through the atmospheric
channel or free space, rather than guided optical fibers, toward the receiver.
Ground-to-satellite (optical uplink) and satellite-to-ground (optical downlink)
involve propagation of optical beam through the atmosphere as well as in free
space. Therefore, these links are a combination of terrestrial and space links.
Figure 1.8 illustrates the application areas of FSO links.
1.1 Introduction 11

Fig. 1.8 Applications of FSO communication links

Fig. 1.9 Block diagram of FSO communication link

The basic block diagram of an FSO link is shown in Fig. 1.9. Like any other
communication technologies, FSO communication link comprises of three basic
subsystems, viz., transmitter, channel, and receiver [20].
(i) Transmitter: Its primary function is to modulate the message signal onto the
optical carrier which is then propagated through the atmosphere to the
receiver. The essential components of the transmitter are (a) the modulator,
(b) the driver circuit for the optical source to stabilize the optical radiations
against temperature fluctuations, and (c) the collimator or the telescope that
collects, collimates, and directs the optical radiations toward the receiver.
The most widely used modulation is the intensity modulation (IM) in which
the source data is modulated on the irradiance/intensity of the optical carrier.
This can be
 12 1 Overview of Wireless Optical Communication Systems

achieved by varying the driving current of the optical source directly with the
message signal to be transmitted or by using an external modulator.
(ii) Channel: Since the FSO communication channel has the atmosphere as its
propagating medium, it is influenced by unpredictable environmental factors
like cloud, snow, fog, rain, etc. These factors do not have fixed
characteristics and cause attenuation and deterioration of the received signal.
The channel is one of the limiting factors in the performance of FSO system.
(iii) Receiver: Its primary function is to recover the transmitted data from the
incident optical radiation. It consists of a receiver telescope, optical filter,
photodetector, and demodulator. The receiver telescope collects and focuses
the incoming optical radiation onto the photodetector. The optical filter
reduces the level of background radiation and directs the signal on the
photodetector that converts the incident optical signal into an electrical
signal.

1.2 Comparison of FSO and Radio-Frequency

Communication Systems

FSO communication system offers several advantages over the RF system. The major
difference between FSO and RF communications arises from the large difference
in the wavelength. Under clear weather conditions (visibility >10 miles), the
atmospheric transmission window is in the near IR region and lies between 700
and 1600 nm. The transmission window for RF lies between 30 mm and 3 m.
Therefore, RF wavelength is a thousand of times larger than the optical
wavelength. This high ratio of wavelengths leads to some interesting differences
between the two systems as given below:
(i) Huge modulation bandwidth: It is a well-known fact that an increase in
carrier frequency increases the information-carrying capacity of a
communication sys- tem. In RF and microwave communication systems, the
allowable bandwidth can be up to 20 % of the carrier frequency. In optical
communication, even if the bandwidth is taken to be ≈ 1 % of carrier frequency
( 1016 Hz), the allowable bandwidth will be 100 THz. This makes the usable
bandwidth at an optical frequency in the order of THz which is almost 10 5
times that of a typical RF carrier.
(ii) Narrow beam divergence: The beam divergence is~ h/DR, where h is the
carrier wavelength and DR the aperture diameter. Thus, the beam spread
offered by the optical carrier is narrower than that of RF carrier. For example,
the laser beam divergence= at h 1550 nm and aperture diameter DR 10 cm
come out to be 0.34 µrad. On the other hand, radio-frequency signal say at X
band will produce a very large beam divergence, e.g., at 10 GHz = i.e., h 3
cm, and aperture diameter
= DR 1 m yields beam divergence to be 67.2 mrad.
Much smaller beam divergence at optical frequency leads to increase in the
intensity of signal at the receiver for a given transmitted power. Figure 1.10
shows the
1.2 Comparison of FSO and Radio-Frequency Communication Systems 13

Fig. 1.10 Comparison of

optical and RF beam
divergence from Mars
toward Earth [21]

comparison of beam divergence for optical and RF signals when sent back from
Mars toward Earth [21].
(iii) Less power and mass requirement: For a given transmitter power level, the
optical intensity is more at the receiver due to its narrow beam divergence.
Thus, a smaller wavelength of optical carrier permits the FSO designer to
come up with a system that has smaller antenna than RF system to achieve
the same gain (as antenna gain is inversely proportional to the square of
operating wavelength). The typical size for the optical system is 0.3 vs. 1.5 m
for the spacecraft antenna [22].
(iv) High directivity: Since the optical wavelength is very small, a very high
directivity is obtained with small-sized antenna. Antenna directivity is
closely related to its gain. The advantage of optical carrier over RF carrier
can be seen from the ratio of antenna directivity as given below

2
Gainoptical 4v/ϴ div(optical)
= , (1.1)
GainRF 2
4v/ϴ div(RF)

where ϴdiv(optical) and ϴdiv(RF) are the optical and RF beam divergence, respec-
tively, and are proportional to h/DR. For system using optical carrier with
aperture diameter DR 10 cm and h 1550 nm gives ϴdiv(optical) 40 µrad.
At beam divergence of=40 µrad, the antenna = gain, Gainoptical, is approximately
100 dB. In order to achieve the same ≈ gain in RF system using X band at
h 3 cm, the size of aperture diameter DR becomes extremely large and
=
unpractical to implement.
(v) Unlicensed spectrum: In RF system, interference from adjacent carrier is the
major problem due to spectrum congestion. This requires the need of
spectrum licensing by regulatory authorities. But optical system is free from
spectrum licensing till now. This reduces the initial setup cost and
development time.
(vi) Security: It is difficult to detect the transmuted optical beam as compared to
RF signal because of its narrow beam divergence. In order to detect the
transmitted
 14 1 Overview of Wireless Optical Communication Systems

optical signal, one has to be physically very close ( 0.1≤miles) to the beam spot
diameter. Studies have shown that the optical signal would drop to 140 dB
from its peak transmission power at a distance of 10 miles. However, RF
signal has much wider region of listening. In this case, signal can easily be
picked up roughly at a distance of 40 miles and is down about 40 dB at
approximately 100 miles.
In addition to the above advantages, the secondary advantages of FSO communica-
tion system are: (a) it is beneficial in the cases where fiber optic cables cannot be
used, (b) easily expandable and reduces the size of network segments, and (c) light
weight and compact.
Besides these advantages, an FSO communication system has got some disad-
vantages as well. It requires tight alignment and pointing due to its narrow beam
divergence. Since light cannot penetrate walls, hills, buildings, etc., a clear line-of-
sight is required between the transmitter and receiver. Also, unlike RF systems,
an FSO system is highly prone to atmospheric conditions that can degrade the
performance of system drastically. Another limiting factor is the position of the
sun relative to the laser transmitter and receiver. In a particular alignment, solar
background radiations will increase leading to poor system performance.

1.3 Choice of Wavelength in FSO Communication System

Wavelength selection in FSO communication system is a very important design

parameter as it affects link performance and detector sensitivity of the system.
Since antenna gain is inversely proportional to operating wavelength, it is more
beneficial to operate at lower wavelengths. However, higher wavelengths provide
better link quality and lower pointing-induced signal fade [23]. A careful
optimization of operating wavelength in the design of FSO link will help in
achieving better performance. The choice of wavelength strongly depends on
atmospheric effects, attenuation, and background noise power. Further, the
availability of transmitter and receiver components, eye safety regulations, and
cost deeply impact the selection of wavelength in FSO design process.
The International Commission on Illumination [24] has classified optical radia-
tions into three categories: IR-A (700–1400 nm), IR-B (1400–3000 nm), and IR-C
(3000 nm–1 mm). It can be subclassified into:
(i) near-infrared (NIR) ranging from 750 to 1450 nm which is a low attenuation
window and mainly used for fiber optics,
(ii) short-infrared (SIR) ranging from 1400 to 3000 nm out of which 1530–
1560 nm is a dominant spectral range for long-distance communication,
(iii) mid-infrared (MIR) ranging from 3000 to 8000 nm which is used in military
applications for guiding missiles,
(iv) long-infrared (LIR) ranging from 8000 nm to 15 µm which is used in thermal
imaging, and
(v) far-infrared (FIR) which is ranging from 15 µm to 1 mm.
 1.4 Range Equation for FSO Link 15

Almost all commercially available FSO systems are using NIR and SIR wavelength
range since these wavelengths are also used in fiber-optic communication, and
their components are readily available in the market.
The wavelength selection for FSO communication has to be eye and skin safe
as certain wavelengths between 400 and 1500 nm can cause potential eye hazards
or damage to the retina [25]. Table 1.3 summarizes various wavelengths used in
practical FSO communication for space applications.

1.4 Range Equation for FSO Link

The link performance analysis begins with the basic component values and system
parameters which are assumed to be known and are fixed in advance. For
example, the laser with the fixed output power as a function of operating
wavelength, transmitter, receiver telescope sizes, etc. directly affect the link
performance. Therefore, the link performance analysis can be performed if all the
components and operational parameters are precisely specified. The three basic
steps for evaluating the performance of an optical link are as follows:
(i) Determine the number of detected signal photons at the detector taking into
account the various losses in the transmitter, channel, and receiver.
(ii) Determine the number of detected background noise photons generated at the
detector.
(iii) Compare the number of detected signal photons with the number of detected
noise photons.
At the transmitter side, the optical source emits optical power with a varying
degree of focusing often described by its emission angle. The total power (in
watts) emitted from a uniform source characterized by brightness function B
(watts/steradian·area), surface area As, and emission angle ▲s is given by [43]

PT = BAs▲s. (1.2)

For symmetrical radiating sources, the solid emission angle ▲s can be related to the
planar emission angle ϴ s (refer Fig. 1.11a) by

▲s = 2v [1 — cos (ϴs/2)] . (1.3)

For any Lambertian source that emits power uniformly in forward direction has
ϴ| | ≤v/2 implying ϴ s =
v. It gives ▲s = and therefore transmitted power,
2v
= s. The light can also be collected and refocused by means of beam- forming
PT 2vBA
optics as shown in Fig. 1.11b. The light from the source beam is focused to a spot
with the help of converging lens, and the diverging lens expands the beam to
planar beam diameter DR given by
16 1 Overview of Wireless Optical Communication Systems

Table 1.3 Wavelengths used in practical FSO communication system

Mission Laser Wavelength Other parameters Application
Semi-conductor AlGaAs laser 830 nm 60 mW, 25 cm Inter-satellite
Inter-satellite Link diode telescope size, comm.
Experi- 50 Mbps, 6 µrad
ment(SILEX) [26] divergence, direct
detection
Ground/Orbiter Argon-ion Uplink: 13 W, 0.6 m and Ground-to-satellite
Lasercomm laser/GaAs 514.5 nm 1.2 m tx. and link
Demonstration laser Downlink: rx.telescopes size,
(GOLD) [27] 830 nm respectively;
1.024 Mbps,
20 µrad divergence
RF Optical System Diode 1064 nm 6 W, 0.135 m and Deep space
Study for Aurora pumped 10 m tx. and rx. missions
(ROSA) [28] Nd:YVO4 telescopes size,
laser respectively;
320 kbps;
Deep Space Optical Master 1058 nm 1 W, 10–20 Mbps Inter-satellite/Deep
Link oscillator space missions
Communications power
Experiment amplifier
(DOLCE) [29] (MOPA)
Mars Orbiter Laser Diode 1064 nm 32.4 W, 420 µrad Altimetry
Altimeter (MOLA) pumped Q divergence, 10 Hz
[30] switched pulse rate, 618 bps;
Cr:Nd:YAG 850 µrad receiver
field-of-view
(FOV)
General Atomics Nd:YAG 1064 nm 2.6 Gbps Remotely Piloted
Aeronautical Aircraft (RPA) to
Systems (GA-ASI) LEO
& TESAT [31]
Altair Laser diode 1550 nm 200 mW, 2.5 Gbps, UAV-to-ground link
UAV-to-ground 19.5 µrad jitter
Lasercomm error 10 cm and 1 m
Demonstration [32] uplink and
downlink
telescopes size,
respectively
Mars Polar Lander AlGaAs laser 880 nm 400 nJ energy in Spectroscopy
[33] diode 100 nsec pulses,
2.5 kHz rate,
128 kbps
Cloud-Aerosol Nd:YAG 532 nm/1064 n1m15 mJ energy, Altimetry
Lidar and IR 20 Hz rate, 24 ns
Pathfinder Satellite pulse
Observation
(CALIPSO) [34]
(continued)
1.4 Range Equation for FSO Link 17

Table 1.3 (continued)

Mission Laser Wavelength Other parameters Application
KIrari’s Optical AlGaAs laser 847/810 nm 50 Mbps, 40 cm and Satellite-to-ground
Downlink to diode 4m tx. and rx. downlink
Oberpfaffenhofen telescopes size,
(KIODO) [35] respectively; 5 µrad
divergence
Airborne Laser Lumics fiber 800 nm 300 mW, 50 Mbps Aircraft and GEO
Optical Link laser diode satellite link
(LOLA) [36]
Tropospheric Nd:YAG 1064 nm 360 W, 5 cm Interferometry
Emission telescope size,
Spectrometer (TES) 6.2 Mbps
[37]
Galileo Optical Nd:YAG 532 nm 250 mJ, 12 ns Deep Space
Experiment pulse width, 110 Missions
(GOPEX) [38] µrad divergence,
0.6 m primary and
0.2 m secondary
transmitter
telescope size,
12.19 h 12.19 mm
CCD array receiver
Engineering Test AlGaAs laser Uplink: 13.8 mW, Bi-directional
Satellite VI diode 510 nm 1.024 Mbps ground-to-satellite
(ETS-VI) [39] (downlink) Downlink: bidirectional link, link
Argon laser 830 nm direct detection,
(uplink) 7.5 cm spacecraft
telescope size,
1.5 m Earth station
telescope
Optical Inter-orbit Laser Diode 819 nm 200 mW, Bi-directional
Communications 2.048 Mbps, direct Inter-orbit link
Engineering Test detection, 25
Satellite (OICETS) cm telescope
[40] size
Solid State Laser Diode 1064 nm 1 W, 650 Mbps GEO-GEO link
Communications in pumped return channel and
Space (SOLACOS) Nd:YAG 10 Mbps forward
[41] channel, 15
cm telescope
size,
coherent reception
Short Range Diode 1064 nm 40 W, 1.2 Gbps, Inter-satellite link
Optical pumped 4 cm telescope size,
Intersatellite Link Nd:YAG BPSK homodyne
(SROIL) [42] detection
18 1 Overview of Wireless Optical Communication Systems

a b

Fig. 1.11 Demonstration of optical emission from light source. (a) Light emission from Lamber-
tian source [43]. (b) Light emission using beam forming optics [43]

DT2
" #1/2
2
hR
DR = DT 1 + , (1.4)

where h is the operating wavelength, DT the transmitter lens diameter, and R the
distance from the lens or link range.
8
< Near field,hR 2 < 1, DR = DT
D
For T (1.5)
: Far hR = DT
D2 > 1, DR
field, hR
T

First condition implies that emerging light is collimated with diameter equal to
transmitter lens diameter. The second condition means that emerging light
diverges with distance from the source. The planar beam angle, ϴ b of the
diverging light source, also called diffraction limited transmitter beam angle is
approximately given by [43]

ϴ b Š DR
. (1.6)
R
Substituting the value of DR for far field case, we get

h
ϴb =
D . (1.7)
T
The two-dimensional solid angle can be approximately related to planar beam angle
by

v b 2
▲b = 2v [1 — cos (ϴ
b
/2)] Š 4 ϴ . (1.8)
1.4 Range Equation for FSO Link 19

Transmitter gain GT from Eqs. (1.7) and (1.8) is given by

4v
G ≈ 4DT . (1.9)
=T ▲b 2

h
After propagating through link distance R, the field intensity of the beam will be
GT PT
I= . (1.10)
4vR2

A normal receiving area A within the beam collects the field power
G T PT
P A. (1.11)
R 2
= 4vR
Let us define the receiver gain, GR, in terms of A

G A
4v => A h2GR
R . (1.12)
= = 4v
h2
Therefore, from Eqs. (1.11) and (1.12),

2
PR = PT GT h
G R. (1.13)
4vR

When the other loss factors are incorporated, the above equation becomes

PR = PT (GT щT (GRщRщh) , (1.14)

щTP) 4vR
h
2
where
PR : signal power at the input of photodetector
PT : transmitter power
щT and щR : efficiencies of transmitter and receiver
h optics, respectively
GT : gain of the transmitting antenna GT ≈ (4DT /h)2
h i
GR : gain of the receiving antenna GR ≈ (4DR/h)2
щTP i
: transmitter pointing loss factor
h
4vR : space loss factor, where R is the link
distance
щ h : narrowband filter transmission factor
From the above equation, it is seen that the receive signal power can be
increased by one or more of the following options:
(i) Increasing transmit power: The most simplest way to improve the receive
signal power is to increase the transmit power since the receive power scales
 20 1 Overview of Wireless Optical Communication Systems

linearly with the transmit power. However, increasing the transmit power
implies the increase in the overall system power consumption, and this can
lead to issues like safety, thermal management, etc.
(ii) Increasing transmit aperture: The transmit aperture size and beamwidth are
inversely proportional to each other. Therefore, increasing the transmit
aperture size will effectively reduce the transmitter beamwidth and hence
deliver the signal with more intensity. However, it will lead to tight
acquisition, pointing, and tracking requirement. Further, transmit aperture
cannot be increased indefinitely as it will increase the overall mass of the
terminal and that will increase the cost of the system.
(iii) Increasing receiver aperture: The receive signal power scales directly with
the receive aperture area. However, the amount of background noise
collected by the receiver will also increase with the increase in the receiver
aperture area. It implies that the effective performance improvement does not
always scale linearly with the receiver aperture area.
(iv) Reducing pointing loss: Reducing the transmitter and receiver pointing loss
will improve the overall signal power level and will also reduce the pointing-
induced signal power fluctuations.
(v) Improving overall efficiency: The overall efficiency can be improved by
improving щT , щR, and щh through appropriate optics and filter design.

1.5 Technologies Used in FSO

Technologies used in FSO system are almost similar to conventional RF system.

Most of the techniques are adapted directly from RF systems. In the following,
various detection and modulation schemes used in FSO system are discussed. The
modulation of optical carrier differs from the modulation of RF carrier because of
the characteristics and limitations of the devices used for carrying the modulation
process. The optical modulation can be carried out in two ways: internal or
external as shown in Fig. 1.12. An internal modulator is one where the
characteristics of the source are directly varied in accordance with the information
signal to produce the modulated optical signal. Intensity modulation can be
performed by varying the bias current. Frequency or phase modulation can be
obtained by changing the cavity length of the laser. Pulse modulation can be
achieved by varying the driving current above and below the threshold. These
modulations are limited to the linear range of power characteristics of the source.
In case of external modulators, an external device is used which varies the
characteristics of the carrier in accordance with the modulating signal. These
systems are capable of utilizing full power of the source. However, external
modulators limit the modulation range and require relatively high drive current.
At optical frequencies, these modulators operate directly on carrier intensity
(amplitude square of electric field) rather than the amplitude of the carrier. Other
1.5 Technologies Used in FSO 21

a b

Fig. 1.12 Optical modulators. (a) Internal modulator. (b) External modulator

ways to modulate optical carrier are by using its phase or polarization. However,
due to simplicity reasons, the most popular modulation scheme used in FSO
system is intensity modulation: baseband or subcarrier. The modulation schemes
can be classified into two categories: baseband intensity modulation and subcarrier
intensity modulation. The most common method used for the detection of the
optical signal is direct detection. When the intensity-modulated signal is detected
by a direct detection receiver, the scheme is known as intensity-modulated/direct
detection (IM/DD) and is commonly used in FSO systems. The other approach of
detecting the modulated optical signal is coherent detection. It makes use of local
oscillator to down convert the optical carrier to baseband (homodyne detection) or
to RF intermediate frequency (heterodyne detection). This RF signal is
subsequently demodulated to baseband via conventional RF demodulation
process.

1.5.1 Direct Detection System

In direct detection technique, the received optical signal is passed through optical
band-pass filter to restrict the background radiation. It is then allowed to fall on
the photodetector which produces the output electrical signal proportional to the
instantaneous intensity of the received optical signal. It may be regarded as linear
intensity to current convertor or quadratic (square law) converter of optical
electric field to detector current. The photodetector is followed by an electrical
low-pass filter (LPF) with bandwidth sufficient enough to pass the information
signal.
The signal-to-noise ratio (SNR) of direct detection receiver can be obtained by
using noise models for a particular detector, i.e., PIN or avalanche photodetector
(APD). With the received power given in Eq. (1.14) and detector noise sources,
the SNR expressions are obtained. The SNR for PIN photodetector is given by

(R0PR)2 L
SNR = , (1.15)
2qB (R0 R + R0 B + Id ) + 4K
B
P TB/R
P
22 1 Overview of Wireless Optical Communication Systems

Fig. 1.13 Block diagram of direct detection receiver

where R0 is the detector responsivity and is given by

щq
R0 = . (1.16)
hv
In the above equation, щ is the detector quantum efficiency, q = 1.602 10—19
J is the electronic charge, h= 6.623 10—34 Js the Planck’s constant, and v the
operating frequency. Other parameters in Eq. (1.15) are B the receiver bandwidth,
Id the dark current, KB the Boltzmann’s constant, T the absolute temperature, RL the
equivalent load resistance, and PB the background noise power. When APD is
used, the dark current and shot noise are increased by the multiplication process;
however, the thermal noise remains unaffected. Therefore, if the photocurrent is
increased by a factor of M, avalanche multiplication factor, then the total shot
noise is also increased by the same factor. The direct detection SNR for APD
photodetector is given by

(MR0PR)2
SNR = (1.17)
[2qB (R0 R + R0 PB + dbI ) M2F ds ] + 4KB TB/R L
+I
P

In Eq. (1.17), F is the excess noise factor arising due to random nature of
multiplication factor, Idb the bulk dark current, and Ids the surface dark current.
The block diagram of direct detection receiver is given in Fig. 1.13. Since the
photodetector response is insensitive to the frequency, phase, or polarization of the
carrier, this type of receiver is useful only for intensity-modulated signals.

1.5.1.1 Baseband Modulation

In baseband modulation, the information signal directly modulates the LED/laser

drive current and hence the optical carrier. This signal, often called baseband-
modulated signal, is then transmitted through the atmospheric channel. At the
receiver side, the information is recovered from the baseband-modulated signal
using direct detection technique. The modulation schemes that come under this
category include on-off keying (OOK) and digital pulse-position modulation
(PPM). Other pulse modulation schemes like digital pulse interval modulation
(DPIM), pulse amplitude and position modulation (PAPM), and differential
amplitude pulse interval modulation (DAPIM) are upcoming modulation schemes,
but they have not yet achieved enough popularity as compared to OOK and PPM.
Most of the work on FSO systems has been carried out using OOK modulation
scheme because it is simple and easy to implement [44].
1.5 Technologies Used in FSO 23

In OOK, the transmission of binary data is represented by the presence or

absence of light pulse, i.e., if the information bit is 1, laser is turned on for the
duration Tb, and if it is 0, nothing is transmitted. OOK system requires adaptive
threshold in order to deal with the fluctuating attenuation of the atmosphere. In
OOK with nonreturn- to-zero (NRZ-OOK) signaling, the bit one is simply
represented by an optical pulse that occupies the entire bit duration while the bit
zero is represented by the absence of an optical pulse. In OOK with return-to-zero
(RZ-OOK) signaling, a one bit is represented by the presence of an optical pulse
that occupies part of the bit duration.
For long-distance communication, M-PPM scheme is most widely used
because of its high peak-to-average power ratio that improves its power efficiency.
Also, unlike OOK it does not require adaptive threshold. In M-PPM scheme, each
symbol period is divided into M time slots each of duration Ts seconds, and the
information is placed in one of the M time slots to represent a data = word.
Here, M 2n where n is the number of information bits. Therefore, each PPM
symbol is mapped directly to n bit sequence and thus allows log 2M bits within
each PPM symbol. The bit-to-symbol mapping can be viewed as one-to-one
assignment of symbols to each of n consecutive information bits. The PPM
scheme is preferred for long- distance communication except when the transmitter
power is peak power limited or system is bandwidth limited. Figures 1.14 and 1.15
show OOK and 8-PPM scheme, respectively, for the transmission of random bit
sequence say 110010. It is clear that PPM scheme requires more = bandwidth (= 1/Ts
where Ts Tb log2 M/M) than the OOK-modulated signal bandwidth (= 1/Tb ) and
complex transceiver design due to tight synchronization requirements.

1.5.1.2 Statistical Model for Direct Detection

In case of direct detection receiver, the energy of the received signal is considered
but not its phase. For every transmitted message, say X, a number of photons fall
on the detector. The absorption of these photons results in an output Y from the
detector that can be passed on to the demodulator and the decoder. Several direct
detection statistical models are mentioned below which define the conditional
probability density function of the output Y given X.

Fig. 1.14 OOK modulation scheme for the transmission of message 110010

Fig. 1.15 8-PPM scheme with eight slots for the transmission of message 110010
24 1 Overview of Wireless Optical Communication Systems

(i) The Poisson channel model for ideal photodetectors

The Poisson distribution in the detection of photons [43] can be expressed as
Kkexp (—K0)
0
fY/X (k/0) = k = 0, 1, 2, . . .
k
K exp!(—K )
k
1 1
fY/X (k/1) k = 0, 1, 2, . . . , (1.18)
= k!

where K0 and K1 are the average number of photons detected when X 0 and
X = 1, respectively. =
(ii) The McIntyre-Conradi model for APDs
The average number of photons absorbed over the active surface of an APD
with optical power P (t) in T seconds [43] can be expressed as

щ ZT
K hv
= P (t) dt, (1.19)
0

where h is the Planck’s constant, v the optical frequency, and щ the quantum
efficiency of the photodetector, defined as the ratio of average number of
photons absorbed by the APD (each absorbed photon produces an electron-
hole pair) to the average number of incident photons. The actual number of
absorbed photons is a Poisson-distributed random variable with mean K
(where K K0 Kb for bit “0,” i.e., the sum of background photon
= or K K1 Ks Kb for bit “1,” i.e., the sum of actual and background
count,
= count).
photon = In an APD, the density of the output electrical signal in
response+ to the absorbed photons is modeled accurately by McIntyre-Conradi
distribution [45]. The conditional probability density of obtaining k0
electrons in response
to n0 absorbed photons is given by

k0
f 0
k /n 0
= n0T 1 — keff + 1
Y N
/
k0 (k0 — n0)!T keffk0
+
1 — keff
n0 + 1
n0 + keffk0
1 + keff (M — (1 — keff) (M — k —
0
· 1) 1 — keff n 0 ,
1)
M · M (1.20)

where M is the average gain of the APD and keff the ionization ratio ranging
from 0 < keff < 1. Taking an average of Eq. (1.20) over the number of
absorbed photons n0 gives
1.5 Technologies Used in FSO 25

k0
0
fY X0
k = f Kn
Y/N
nD1 k0/n0
exp (—K) , k0 ≥ 1. (1.21)
n0!
It should be noted that the limit of summation is up to k0 instead of infinity as
the number of absorbed photons can never be more than the released
electrons. Therefore, for k0 N where N is the set of natural numbers, the
conditional probability
e density function of received photons is given by [45,
46]

f
Y/X k0/x
k
n0 + keffk0
0 n0T 1 — keff 0
1 + keff (M — Kn exp (—K )
k0 +
1) 1

x x
= 1 — keff M
keffk0
k0 (k0 n0)!T
X
— 1 — keff
|n0| n0!
n 0

D1

k—n
(1 — keff) (M — 1)
, (1.22)
·
M
where K0 = Kb is the average number of photons detected when x 0 and
K1 =Ks +Kb is the average number of photons detected when x 1. Both
bulk and = surface dark currents in APD will add to the background noise
photon count Kb.
(iii) Additive white Gaussian noise approximation
The additive white Gaussian noise model is commonly used for direct
detection receiver. In this case, the conditional probability density function of
received photons is given by
1 h i
exp — (y — μ )2 /2ơ 2 , (1.23)
fY/X (y/x) = p2vơ x x
2 x

where x e 0,{1 . The parameters μ x and ơ 2x are the mean and variance,
respectively, when X = x. This model is often used with APD and the mean
and variance in Eq. (1.23) are given as [47]

μ0 = MKb + IsTs/q,
μ1 = M (Ks + Kb) + IsTs/q,

0 b IsTs 2K2BTTs s
ơ 2 = M2FK + q + q RL 2BT , (1.24)
26 1 Overview of Wireless Optical Communication Systems

1 s b IsTs 2K2BTTs s
ơ 2 = M2F (K + K ) + q + q RL 2BT .
1.5 Technologies Used in FSO 27

In the above equations, M is the average APD gain, F the excess noise
factor of APD, Is the surface dark current, Ts the slot width, B=
( 1/2Ts) the
electrical bandwidth of receiver, and Ks and Kb the average number of signal
and background noise photons, respectively.

1.5.1.3 Subcarrier Modulation

In a subcarrier intensity modulation (SIM) scheme [48], the radio-frequency (RF)

electrical subcarrier signal is pre-modulated with the information signal. The
electrical subcarrier can be modulated using any modulation scheme like binary
phase-shift keying (BPSK), quadrature phase-shift keying (QPSK), quadrature
amplitude modulation (QAM), amplitude modulation (AM), frequency modulation
(FM), etc. This pre-modulated signal is used to modulate the intensity of the
optical carrier. At the receiver, the signal is recovered using direct detection as in
IM/DD system. It does not require adaptive threshold unlike OOK scheme, and it
is more bandwidth efficient than PPM scheme. Optical SIM inherits the benefits
from more mature RF system; therefore, it makes the implementation process
simpler. The SIM technique allows simultaneous transmission of several
information signals over the optical link. The subcarrier multiplexing can be
achieved by combining differ- ent modulated electrical subcarrier signals using
frequency-division multiplexing (FDM), which is then used to modulate the
intensity of a continuous laser source that serves as the optical carrier. Figure 1.16
shows the principle of SIM optical system for FSO link. The disadvantage of this
multiplexing scheme is tight synchronization and design complexity at the
receiver side.
Both baseband and SIM signals can be demodulated using direct detection/non-
coherent detection techniques which are low cost, less complex, and widely used
in FSO communication system. Direct detection technique can also be used with
analog modulation of optical carrier. However, it is not widely used as it puts
linearity constraints on the laser source and modulation technique that are difficult
to achieve with present state of art. The most commonly used modulation schemes
in FSO systems are shown in Fig. 1.17. The choice of a suitable modulation
scheme

Fig. 1.16 Block diagram of SIM for FSO link

28 1 Overview of Wireless Optical Communication Systems

Fig. 1.17 Modulation schemes in FSO system

Fig. 1.18 Block diagram of coherent optical communication system

requires the trade-off between power efficiency, bandwidth requirement for a

given data rate, and implementation complexity.

1.5.2 Coherent Detection

In coherent detection receiver, incoming signal is mixed with a locally generated

coherent carrier signal from a local oscillator (LO). This mixing of incoming
weak optical signal and strong LO signal at the photodetector provides signal
amplification and converts the optical signal into electrical signal. The strong field
of LO raises the signal level well above the noise level of the electronics circuit.
Thus, the sensitivity of the coherent receiver is limited by the shot noise of the LO
signal. Furthermore, because of the spatial mixing process, the coherent receiver
is sensitive to signal and background noise only that falls within the same spatial
temporal mode of the LO. This allows coherent detection optical receiver to
operate in very strong background noise without significant degradation in the
performance. The basic block diagram of coherent receiver is shown in Fig. 1.18.
Depending upon the frequency of the local oscillator ωL and the frequency of
the incoming signal ωS, coherent detection can be categorized as heterodyne
detection or homodyne detection. If ωL is offset from ωs by an intermediate
frequency
1.5 Technologies Used in FSO 29

(IF) ωIF, then it is called heterodyne detection, i.e., ω=

L ωS ωIF. In case of
homodyne detection, there is no offset between ω L and ωS, i.e., ωIF = 0 implying
ωL =ωS. In both heterodyne and homodyne receivers, the photodetector current Ip
is proportional to the optical intensity and is given as

Ip / (eR + eL)2 , (1.25)

where eR and eL are incoming received signal and local oscillator electric fields,
respectively. The above equation can, therefore, be written as

Ip / [ERcos (ωSt + øS) + ELcos (ωLt + øL)]2 , (1.26)

where ER and EL are the peak incoming received and LO signals, respectively,
and øs and øL are the phase of transmitted and LO signals, respectively. Solving
Eq. (1.26) and removing higher frequency terms which are beyond the detector
response give

1 2 1
Ip / E + E2 + 2ER EL cos (ωL t — Sω t + ø) , (1.27)
2 R 2 L

where ø =øS øL. Since the signal power is proportional to the square of the
electrical field, the above equation can be written as
p
Ip / PR + PL + 2 PRPLcos (ωLt — ωSt + ø) . (1.28)

In the above equation, PR and PL are the optical power levels of incoming signal
and local oscillator signal, respectively. Photocurrent in relation to incident power
PR is governed by Ip = щqPR/hv. Hence, the above equation can be written as

щq h p i
Ip = PR + PL + 2 P R P
L cos (ωL t —S ω t + ø) , (1.29)
h
v
where щ is the quantum efficiency of photodetector, h the Plank’ s constant, and v
the optical frequency. Generally, the local oscillator signal power is much higher
than the incoming signal power, and therefore, the first term in the above equation
can be neglected. After that the signal component of the photodetector current is
given as

щq h p i
Ip = 2 PR PL cos (ωL t — Sω t + ø) . (1.30)
h
v
For heterodyne detection, ωs ≠ ωL and therefore the above equation can be written
as
щq h p i
Ip = 2 PR PL cos (ωIF t + ø) . (1.31)
h
v
30 1 Overview of Wireless Optical Communication Systems

It is clear from this equation that photodetector current is centered around an IF.
This IF is stabilized by incorporating the local oscillator laser in a frequency
control loop. In case of homodyne detection, ωL = ωs and therefore Eq. (1.31)
reduces to

2щq p p
Ip = PR PL cosø = 2R0 PR P
L cosø. (1.32)
h
v
In this case, output from the photodetector is in the baseband form, and local
oscillator laser needs to be phase locked to the incoming optical signal. It is
clear that the signal photocurrent pin both homodyne and heterodyne receivers is
effectively amplified by a factor 2 PRPL . This amplification factor has the effect
of increasing the incoming optical signal level without affecting the preamplifier
noise or photodetector dark current noise. This makes coherent receiver to provide
higher receiver sensitivity.
Various noise contributors in coherent detection are signal shot noise, back-
ground shot noise, LO shot noise, signal-background beat noise, LO-background
beat noise, background-background beat noise, and thermal noise. When the local
oscillator signal power is much greater than the incoming signal power, then the
dominant source of noise is due to local oscillator shot noise, and its mean square
noise power is given by
8
2qR0PLB for
INL2 = ˆ PIN (1.33)
ˆ
:
<
2qR0PLBM2F for APD.

The SNR in this case (assuming no phase difference between source and local
oscillator signal, i.e., ø = 0) is given by

SNR = I2P 2R0PR

. (1.34)
2qR0PLBF qBF

The value of F is unity in case of PIN photodetector. It has been seen that coherent
detection system provides larger link margin (approx 7–10 dB) over direct
detection system. Coherent system can employ any modulation scheme like OOK,
FSK, PSK, PPM, etc. Due to complexity and high cost of coherent receiver
design, it is rarely used in FSO systems. It becomes cost-effective at high data
rates and may find applications in the future.

1.5.3 Optical Orthogonal Frequency-Division Multiplexing

Optical orthogonal frequency-division multiplexing (OFDM) [49] belongs to the

category of multi-carrier modulation (MCM) where the data information is carried
1.5 Technologies Used in FSO 31

Fig. 1.19 Block diagram of OFDM based FSO system

over many lower rate subcarriers. OFDM when implemented with wireless optical
system gives a very cost-effective solution for improving its performance. OFDM
allows the high data rate to be divided into multiple low data rates, and they are
transmitted in parallel form. The main objective of using this MCM scheme is to
lower the symbol rate and provide high tolerance to deep fades that deteriorate the
performance of FSO communication system. OFDM-based FSO system will
exploit the advantages of both OFDM and FSO to become a good candidate for
“last mile” solution for broadband connectivity. Such systems are capable of
providing high spectral efficiency and increased robustness against intensity
fluctuations caused by turbulent atmosphere. OFDM-based FSO system can also
be used in some coding techniques. Low-density parity check (LDPC)-coded
OFDM is able to give performance much better than LDPC-coded OOK over
fluctuating atmospheric environment both in terms of coding gain and spectral
efficiency. However, due to sensitivity of OFDM scheme to phase noise and
relatively large peak-to-average power ratio, the design of OFDM for FSO system
has to be implemented very carefully. Figure 1.19 shows the block diagram of
OFDM-based FSO system.
In OFDM system, the input from the source is baseband modulated using any
of the modulation scheme like PSK, QAM, etc. It is also known as mapping
and the mapped signal is converted from serial to parallel form. This will allow
high data rate stream to split into multiple low data rate narrowband subcarriers.
These narrowband subcarriers experience lesser distortion than high data rates and
require no equalization. Inverse fast Fourier transform (IFFT) and cyclic prefix
(CP) operations are performed on the low data rate narrowband subcarrier to
generate OFDM signals. It is then followed by digital-to-analog and parallel-to-
serial convertors. This OFDM signal now modulates the laser diode, and it is
then allowed to propagate through FSO channel. At the receiver side, after the
signal is being detected by photodetector, the reverse process is carried out to
recover the information signal. Since OFDM uses FFT algorithms for modulation
and demodulation, such system does not require equalization. OFDM employed in
32 1 Overview of Wireless Optical Communication Systems

Table 1.4 Comparison of RF and optical OFDM systems

Type Mathematical model Speed
Wireless Time domain multiple discrete Can be fast for mobile environment
OFDM Rayleigh fading
Optical OFDM Continuous frequency domain Medium
dispersion

optical system is a little bit different as compared to RF-OFDM system. Table 1.4
gives the comparison of RF and optical OFDM systems.
In order to improve the power efficiency of OFDM-based FSO, three variants of
OFDM schemes are used:
(i) Biased-OFDM single side-band scheme: This scheme is based on intensity
modulation and is also known as “Biased-OFDM” (B-OFDM) scheme. In
this case, the transmitted signal is given by

S (t) = SOFDM (t) + D, (1.35)

where D is the bias component. Since IM/DD does not support bipolar
signals, the bias component D has to be sufficiently large so that when it is
added to SOFDM (t) , it results in a nonnegative component. The main
disadvantage of B-OFDM scheme is the poor efficiency.
(ii) Clipped-OFDM single side-band scheme: It is based on single side-band
transmission with clipping of the negative portion of OFDM signal after bias
addition. The conversion of DSB to SSB can be made in two ways: (a) by the
use of Hilbert transformation of inphase signal as the quadrature signal in the
electrical domain or (b) by the use of optical filter. By choosing the optimum
bias value, the power efficiency of clipped OFDM (C-OFDM) can be
improved as compared to B-OFDM.
(iii) Unclipped-OFDM single side-band scheme: This scheme employs LiNb O 3
Mach-Zehnder modulator (MZM) to improve its power efficiency. To avoid
distortion due to clipping, the information signal is transmitted by modulating
the electric field so that the negative part of OFDM is given to the photode-
tector. The distortion introduced by the photodetector is removed by proper
filtering and in recovered signal distortion is insignificant. Unclipped OFDM
(U-OFDM) is less power efficient than C-OFDM, but is significantly more
power efficient than B-OFDM.
The optical OFDM signal at the receiver can be detected by using non-coherent or
coherent technique. The main features of non-coherent and coherent OFDM
systems are as follows:
(i) Non-coherent OFDM: It uses direct detection at the receiver but it can further
be classified into two categories according to how the OFDM signal is
generated: (a) linearly mapped direct detection optical OFDM (DDO-OFDM
or
 1.5 Technologies Used in FSO 33

LM-DDO-OFDM) where the optical OFDM spectrum is a replica of baseband

OFDM and (b) nonlinearly mapped DDO-OFDM (or NLM-DDO-OFDM)
where the optical OFDM spectrum does not display a replica of baseband
OFDM.
(ii) Coherent OFDM: It achieves high spectral efficiency by overlapping
subcarrier spectrum while avoiding interference with the help of coherent
detection and signal set orthogonality. Coherent OFDM improves the
sensitivity of the receiver and increases the robustness against polarization
dispersion. The synergies between coherent optical communication and
OFDM are twofold. OFDM brings computation efficiency of coherent
systems along with ease of channel and phase estimation. Also, it brings
linearity in RF to optical (RTO) up conversion and optical to RF (OTR) down
conversion. However, it increases the cost and design complexity of the
system.

1.6 Eye Safety and Regulations

While designing an FSO link, the designer has to ensure that the chosen operating
wavelength has to be eye and skin safe. This means that the laser should not
pose any kind of danger to the people who might encounter the communication
beam. Figure 1.20 clearly shows the region where different wavelengths of light
get absorbed in the human eye.
Microwaves and gamma rays are absorbed by the human eye and can cause
high degree of damage to lenses and retina. Near-ultraviolet (UV) wavelengths are
absorbed in the lens making them cloudy (cataract) which leads to dim vision or

Fig. 1.20 Pictorial representation of light absorption in the eye for different wavelengths [50]
1.6
34 Eye Safety and Regulations 33
1 Overview of Wireless Optical Communication Systems

Fig. 1.21 Absorption of light

vs. wavelength
10 100
0
80

(%)
Transmission
8
0
60

Absorption
6 40
0

(%)
20
4
0
400 600 800 1000 1200 1400
2 Wavelength (nm)
0

blurring. In far UV and IR regions, wavelengths are absorbed in the cornea and
produce an effect called photokeratitis which can lead to pain/watering in the eye
or pigmentation in the cornea. Visible and near IR region (wavelengths used in
FSO communication) has the highest potential to damage the retina of the eye, and
it can lead to permanent loss of vision that cannot be healed by any surgery. The
range of wavelengths between 400 and 1400 nm can cause potential eye hazards
or even skin damage [25]. The impact of laser injury is more significant in case of
eyes than the skin as the outer layer of the eye, i.e., the cornea, acts as a band-pass
filter to the wavelengths. Therefore, the cornea will be transparent to these
wavelengths, and the energy emitted by the light sources will get focused on the
retina and may cause damage to the eye due to increase in concentration of the
optical energy. However, the light below 400 nm and above 1400 nm is
absorbed by the cornea and does not reach the retina. The absorption coefficient
of the cornea is more for higher wavelengths than (>1400 nm) than for the shorter
wavelengths as can be observed from the Fig. 1.21.
Lasers can cause damage to our skin by causing thermal burns or
photochemical reaction. The penetration of laser beam inside the human body
depends upon the choice of operating wavelength. UV rays are absorbed by the
outer layer of the skin and cause skin cancer or premature aging of the skin.
Exposure of high-intensity beam for very long period can cause thermal burns or
skin rashes. IR radiation can penetrate deep into skin leading to thermal burns.
Therefore, it is very essential to regulate the laser power to ensure the safety of the
human eye and skin.
Various international standard bodies (such as American National Standards
Institute (ANSI) Z136 in the USA, Australian/New Zealand (AS/NZ) 2211
Standard in Australia, and International Electrotechnical Commission (IEC) 60825
interna- tionally) provide safety guidelines of laser beam depending upon their
wavelength and power. Laser Institute of America (LIA) is an organization
that promotes the safe use of lasers and provides laser safety information. These
standards are global benchmark for laser safety, and they are used as guidelines
for various manufacturers. Every organization has its own way of laser
classification, and accordingly, safety precautions and administrative control
measures have to be taken. The classification of the laser is based on whether or
not the maximum permissible emission (MPE) is longer or shorter than the human
aversion response.
34 1 Overview of Wireless Optical Communication Systems

MPE is a quantity that specifies a certain level up to which the unprotected human
eye can be exposed to laser beam without any hazardous effect to the eye or
skin. Aversion response is the autonomic response (within 0.25 s) of the blinking
eye when it moves away from the bright source of light. Another quantity that
determines classification of lasers is accessible emission limit (AEL) which is the
mathematical product of two terms, i.e., MPE limit and limiting area (LA) factor.
Therefore, AEL =MPE LA.
Based upon MPE and AEL calculations, lasers are broadly classified into four
groups, i.e., Class 1 through Class 4. The lower classifications (Class 1 and 2)
have minimum power and therefore do not require protective eye wear. This class
has extended MPE measurements as the human eye will avert from the bright light
long before the beam can injure the unprotected eye. The higher classification
(Class 3R, 3B, and 4) has high power levels; therefore, proper eye safety
precautions have to be taken during laser operations. In this class MPE is shorter
than aversion response. Table 1.5 gives the comparison of laser classification
according to IEC and ANSI standards.
Table 1.6 presents the AEL for two most commonly used wavelengths in FSO
communication systems. It is evident from the table that for Class 1 and 2, lasers
operating at 1550 nm are almost 50 times more powerful than lasers at shorter
operating wavelength, i.e., 850 nm. Also, the combination of low attenuation, high
component availability, and eye safety at 1550 nm wavelength makes it a
preferred choice for FSO communication. Lasers operating at 1550 nm
wavelength when used with erbium-doped fiber amplifier (EDFA) technology are
capable of providing high data rates (>2.5 Gbps) and high power.
Class 1 and Class 1M lasers are preferred for terrestrial FSO communication
links as their radiations are safe under all circumstances. IEC60825-12 [51] covers
the safety standard for free-space optical communication links, and it lists out the
requirements like power, aperture size, distance, and power density for 850 and
1550 nm wavelengths which are presented in Table 1.7. Higher classes of lasers
are also used for FSO communication links for long distance communication or
deep space missions. To ensure the safety of these systems, they are installed on
higher platforms like rooftops or towers to prevent any kind of human injury.
It is to be noted that high-powered pulsed lasers are more dangerous than lower
power continuous lasers. However, lower power laser beams can also be
hazardous when given long-term exposure.

1.7 Applications of FSO Communication Systems

Applications of FSO communication systems range from short range (<1 km) to
long range and space applications. It provides broadband solution (high data rates
without cabling) for connecting end users to the backbone. Short-range application
provides last mile access by connecting various towers, buildings, etc. in urban
areas
1.7 Applications of FSO Communication Systems 35

Table 1.5 Laser classification according to IEC and ANSI standards

Classification IEC 60825 ANSI-Z136.1
Class 1 Very low power lasers and are safe under reasonably foreseeable conditions of
operation. This class is exempted from all beam-hazard control measures. It includes
optical instruments for intrabeam viewing.
Class1M Low power lasers operating between 302.5 N/A
nm and 4000 nm wavelengths and are safe
under reasonably foreseeable conditions
except when used with optical instruments
such as collecting lens, binoculars, telescope,
etc. These lasers produce either collimated
beams with large beam size or highly
divergent beams.
Class 2 Low power laser operating between 400 nm to 700 nm (visible range). This laser class
can be continuous wave (CW) or repetitively pulsed lasers. It is safe to use if it emits
energy below Class 1 AEL for emission duration of less than 0.25 sec (i.e., the time
period of the human eye aversion response). It have an average radiant power of 1mW
or less.
Class 2M Low power laser operating between 400 nm N/A
to 700 nm (visible range). It can cause
optical hazards when viewed with optical
instruments such as collection lens,
telescope, etc. Any emissions outside this
wavelength region must be below the Class
1M AEL.
Class 3R Average power lasers operating between N/A
302.5 nm and 106 nm. The accessible
emission limit is within 5 times the Class 2
AEL for visible range wavelengths and
within 5 times the Class 1 AEL for
wavelengths outside this region. It is unsafe
to view the beam directly with diameter >
7mm.

Class 3A N/A Average power lasers operating between

302.5 nm and 106 nm. The accessible
emission limit is within 5 times the
Class 2 AEL for visible range
wavelengths and between 1 and 5 times
the Class 1 AEL for wavelengths
outside this region. It is unsafe to view
the beam directly with diameter > 7mm.

Class 3B Average power lasers that cannot emit an average radiant power greater than 0.5 Watts
for an exposure time equal to or greater than 0.25 seconds. It is unsafe to view the beam
directly but are normally safe when view diffused reflections
Class 4 High power lasers and are very dangerous both under intrabeam and diffuse reflection
viewing conditions. They may also cause skin injuries and are potential fire hazards.

where digging of cables is a difficult task. It includes point-to-point or point-to-

multipoint links or broadband links. Various applications of FSO systems are
listed below:
(i) Enterprise connectivity: FSO link can easily be deployed to connect various
tower/building enabling local area connectivity. It can also be extended to
connect metropolitan area fiber rings, connect new networks, and provide
high- speed network expansion.
36 1 Overview of Wireless Optical Communication Systems

Table 1.6 Accessible emission limits for 850 and 1550 nm according to IEC standard
Average output optical power (mW)
Laser classification
850 nm 1550 nm
1 <0.22 <10
2 Used only for 400–700 nm and has same AEL as Class 1
3R 0.22–2.2 10–50
3B 2.2–500 50–500
4 >500 >500

Table 1.7 Various requirements of Class 1 and 1M lasers for 850 and 1550 nm [52]
Classification Power (mW)Aperture size (mm) Distance (m) Power density (mW/cm2)
850 nm Wavelength
Class 1 0.78 7 14 2.03
50 2000 0.04
Class 1M 0.78 7 100 2.03
7 14 1299.08
500
50 2000 25.48
1550 nm Wavelength
Class 1 7 14 26
10
25 2000 2.04
Class 1M 10 3.5 100 103.99
7 14 1299.88
500
25 2000 101.91

(ii) Fiber backup: In case of optical fiber link failure, FSO link can be deployed
as a backup link to ensure availability of the system.
(iii) Point-to-point links: It coves inter- (LEO-LEO) and intra-orbital (LEO-GEO)
links and satellite-to-ground/ground-to-satellite link. This type of link
requires good pointing and tracking system. Here, the output power of the
transmitter, power consumption, size, mass, and deployment cost increase
with the link range.
(iv) Point-to-multipoint links: Multi-platform multi-static sensing, interoperable
satellite communications, and shared spaceborne processing are unique net-
work application of FSO system.
(v) Hybrid wireless connection/network redundancy: FSO communication is
prone to weather conditions like fog, heavy snow, etc. In order to obtain 100
% availability of the network, FSO links can be combined with microwave
links that operate at high frequencies (in GHz range) and offer comparable
data rates.
(vi) Disaster recovery: FSO communication system provides high-capacity scal-
able link in case of collapse of existing communication network.
(vii) Backhaul for cellular networks: With the advent of 3G/4G cellular communi-
cation, there is a growing challenge to increase the backhaul capacity
between
 Bibliography 37

the cell towers to cope up with the increase in demand of broadband mobile
customers. The viable backhaul options for 4G network are to deploy fiber-
optic cables or to install FSO connection between towers. Deploying fiber
cables is time-consuming and an expensive task. So FSO communication
system plays an important role in providing backhaul capacity for cellular
networks.

1.8 Summary

This chapter begins with the discussion of various types of optical wireless
communication ranging from indoor IR to outdoor FSO communication. However,
the chapter mainly focuses on outdoor terrestrial FSO communication link. It
presents the advantages of optical carrier over RF carrier where the transmission
rate can exceed 10 Gbps and can find its application in enterprise connectivity,
video surveillance and monitoring, disaster recovery, backhaul for cellular
systems, etc. Various technologies used in FSO communication system, i.e., direct
detection, coherent detection, and OFDM, are presented in this chapter. The
choice of operating wavelength based on absorption losses and component
availability in market is also discussed.

Bibliography

1. R.F. Lucy, K. Lang, Optical communications experiments at 6328 Å and 10.6 µ. Appl. Opt.
7(10), 1965–1970 (1968)
2. M.S. Lipsett, C. McIntyre, R. Liu, Space instrumentation for laser communications. IEEE J.
Quantum Electron. 5(6), 348–349 (1969)
3. I. Arruego, H. Guerrero, S. Rodriguez, J. Martinez-Oter et al., OWLS: a ten-year history in
optical wireless links for intra-satellite communications. IEEE J. Sel. Areas Commun. 27(9),
1599–1611 (2009)
4. S. Kazemlou, S. Hranilovic, S. Kumar, All-optical multihop free-space optical
communication systems. J. Lightwave Technol. 29(18), 2663–2669 (2011)
5. K. Hirabayashi, T. Yamamoto, S. Hino, Optical backplane with free-space optical
interconnec- tions using tunable beam deflectors and a mirror for bookshelf-assembled
terabit per second class asynchronous transfer mode switch. Opt. Eng. 37, 1332–1342 (2004)
6. N. Savage, Linking with light. IEEE Spectr. (2002). [Weblink: http://spectrum.ieee.org/
semiconductors/optoelectronics/linking-with-light]
7. G. Forrester, Free space optics, in Digital Air Wireless. [Weblink: http://www.
digitalairwireless.com/wireless-blog/2013-07/free-space-optics.html]
8. http://andy96877.blogspot.com/p/visible-light-communication-vlc-isdata.html. Visible light
communication- VLC & Pure VLCTM. [Weblink: http://andy96877.blogspot.com/p/visible-
light-communication-vlc-is-data.html]
9. Weblink: http://lasercommunications.weebly.com/
10. Weblink: http://artolink.com
11. Weblink: http://www.fsona.com
View publication stats

38 1 Overview of Wireless Optical Communication Systems