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

net/publication/312132783

FSO System Modules and Design Issues

Chapter · January 2017

DOI: 10.1007/978-81-322-3691-7_3

CITATIONS READS

0 985

3 authors:

Hemani Kaushal Vk Jain

University of North Florida Indian Institute of Technology Delhi
52 PUBLICATIONS   1,728 CITATIONS    166 PUBLICATIONS   1,165 CITATIONS   

SEE PROFILE SEE PROFILE

Subrat Kar
Indian Institute of Technology Delhi
219 PUBLICATIONS   1,444 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Optical communication View project

Study on physical layer impairments in Optical networks View project

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

The user has requested enhancement of the downloaded file.

Chapter 3
FSO System Modules and Design Issues

The basic functional components of FSO communication system consists of (i)

optical power source, transmitter; (ii) modulation and encoding of light, modulator
and encoder; (iii) acquisition, tracking, and pointing (ATP) system, fine pointing
mirrors and steering optics; (iv) background suppression, filter; (v) optical transmit
and receive aperture; and (vi) detector, demodulator, and decoder, receiver.
Various components of the ground-based transceiver, intervening optical channel,
and onboard optical flight transceiver are shown in Fig. 3.1. It consists of transmitter,
receiver, ATP system, and atmospheric channel. Before transmitting the mes-
sage/data signal, a beacon signal is transmitted from ground station to satellite, and
the link is established using ATP system. The uplink beacon signal is first acquired
by the distant target initiating the process of acquisition. During acquisition, one
station (say ground-based earth transceiver) interrogates the uncertainty area of
other station (say onboard satellite) either through beam decollimation or scanning
of the transmitter beam. Once the onboard satellite acquires the beacon from
ground-based earth transceiver, the transition from acquisition to tracking begins.
For narrow beam system, point ahead compensators are used to compensate for
the propagation delay over the link range. Initially for acquiring the signal, the
image of ground-based earth transceiver is captured on focal pixel array, and the
centroid is computed. The ground-based position vector is determined from the
computed centroid. The difference between the current onboard laser position and
the ground-based position with the addition of point ahead angle (in order to account
for two way travel of light between ground-based earth transceiver and onboard
satellite) is the position vector that drives the beam steering mirror. Once the link
is acquired and a line of sight connection is established between two stations, data
communication can start. Data transfer is performed using variety of coherent and
noncoherent modulation schemes and encoding techniques. For both beacon and
data transfer, the signal has to pass through atmospheric optical channel which
in addition to background noise introduces signal attenuation and scintillation.
Therefore, sufficient optical power, pointing accuracy, and directivity are needed to

© Springer (India) Pvt. Ltd. 2017 91

H. Kaushal et al., Free Space Optical Communication, Optical Networks,
DOI 10.1007/978-81-322-3691-7_3
92 3 FSO System Modules and Design Issues

Data/
Receiver
Beacon Beam Splitter
Telescope
Signal Laser Transmitter
Optics
Modulator Beam Steering/ Telescope
& Encoder Space Loss Detector &
Stabilization Optics Telescope Mount
Demodulator
Telescope Mount Atmospheric
Point Losses Focal
Ahead Inertial Plane Beacon Data
Compensators Reference Array
Background Detector Decoder
and Sensors
Noise
ATP System Beam Steering/
Stabilization
ATP System

Earth Transeiver/Ground Station Optical Channel On-board Satellite

Fig. 3.1 Schematic representation of various components for ground-to-satellite optical link

deliver the required signal for uplink beacon pointing or data communication. The
details of the various components used in FSO communication system are given in
the following sections.

3.1 Optical Transmitter

The optical transmitter including the choice of laser, concept of ATP system,
and various types of modulation schemes and coding techniques used in FSO
communication system are discussed in this section. Further, the details of commu-
nication and beacon detectors in FSO receiver are also discussed. The transmitter
converts the source information into optical signals which are transmitted to the
receiver through the atmosphere. The essential components of the transmitter are (i)
modulator, (ii) driver circuit for the optical source to stabilize the optical radiations
against temperature fluctuations, and (iii) collimator that collects, collimates, and
direct the optical signals toward the receiver via atmospheric channel. The optical
sources that are used for FSO transmission lie in the atmospheric transmission
window that is ranging from 700 to 10,000 nm wavelength. The wavelength range
from 780 to 1064 nm is most widely used as beacon operating wavelength. The
1550 nm wavelength is commonly used as data operating wavelength due to
following reasons:
(i) Reduced background noise and Rayleigh scattering: The absorption coefficient
of the Rayleigh scattering has functional dependence with the wavelength  as
4 : Consequently, there is almost negligible attenuation at higher operating
wavelengths as compared to those at the visible range.
(ii) High transmitter power: At 1550 nm a much higher power level (almost 50
times) than at lower wavelengths is available to overcome various losses due
to attenuation.
(iii) Eye-safe wavelength: The maximum permissible exposure (MPE) for eye is
much higher at 1550 nm wavelength than at 850 nm. This difference can be
3.1 Optical Transmitter 93

explained by the fact that at 850 nm, approximately 50 % of the signal can
reach the retina whereas at 1550 nm, the signal is almost completely absorbed
by cornea itself. And therefore the signal received at the retina is negligibly
small.
The component cost increases with the increase in the operating wavelength. For
good optical transmitter, the choice of laser power and wavelength has to be made
very carefully so that an appropriate optical power and transmit antenna gain can be
achieved in order to form a closed loop communication link. However, this is not the
only constraint for most laser sources. The selection of laser is influenced by several
other factors including efficiency, operational lifetime, and achievable diffraction-
limited output power and weight. A good source will have narrow, stable spectral
line width and nearly diffraction-limited single mode spatial profile. Some of the
key requirements which affect the choice of the transmitter laser for FSO-based
applications are given below:
(i) Pulse repetition frequency (PRF): The laser pulsing mechanism (e.g., Q-
switching, cavity dumping) determines the PRF of the laser. Q-switched
lasers using acousto-optic or electrooptic modulators have PRF less than
200 kHz. Cavity dumping lasers have PRF in the order of tens of megahertz.
PRF up to several gigahertz can be achieved with the power amplified lasers
used in conjunction with several stages of amplification.
(ii) Average output power: The laser should have sufficient average power for a
reliable communication link with adequate link margin. For any good laser,
it should provide pulse to pulse power stability and nearly constant average
power over different data rates. The peak power of any laser is given by the
product of energy per pulse and the pulse width. Solid-state lasers provide
large peak power at low PRF. However, the maximum peak power is limited
by the heat dissipation and laser safety norms.
(iii) Pulse width: Laser pulse width should be small to facilitate less background
noise in narrow temporal slots.
(iv) Pulse extinction ratio: The ratio of laser power in on-mode to that of in off-
mode is called pulse extinction ratio. The extinction ratio should be as large
as possible. If the laser emission is not switched to complete off-mode, it may
degrade the extinction ratio resulting in a lower link margin. Solid-state lasers
have modulation extinction ratio of 40 dB, whereas semiconductor lasers have
relatively poor extinction ratio of about 10 dB. Fiber lasers and amplifiers
have extinction ratios in the order of 30 dB.
(v) Output beam quality: The output of the laser should consist of single spatial
mode or at least have single null in the center of far field pattern. To avoid
undesired oscillations either within the laser or in the transmitted beam,
feedback isolation of the laser from the back reflected beam is required.
(vi) Beam pointing stability: For FSO-based applications, the pointing accuracy
on the order of micro-radian or better is desirable. Such an accuracy requires
the pointing stability of the laser to be maintained by the use of optomechan-
ical or spatial resonators within the laser.
94 3 FSO System Modules and Design Issues

(vii) Overall efficiency: In order to minimize the electrical power requirement, it

is desirable to have highest possible overall efficiency.
(viii) Mass and size: For any space-based applications, the mass and size of all
the components should be minimized to achieve low launch cost. It therefore
necessitates the use of optomechanical designs of laser resonator.
(ix) Operational lifetime: The lifetime of the active laser components (e.g., diode
laser, modulator and drivers, etc.) are expected to exceed the operational life-
time of the system. Redundancy of the active elements or block redundancy
of the laser will help in extending the operational lifetime. It should be noted
that the higher the pump power, the lower is the expected lifetime of the laser.
(x) Thermal control and management: An efficient thermal control is required so
that the dissipated heat does not affect the optical alignment integrity of the
system that would otherwise result in further loss.

3.1.1 Choice of Laser

The choice of laser is driven from the basic requirements, i.e., high electrical-
to-optical conversion efficiency, excellent beam quality, variable repetition rate,
stable operation over lifetime, quick start up operation, and high reliability. In
the early days, gas lasers such as carbon dioxide (CO2 ) laser were used because
they were stable and less sensitive to atmospheric effects. However, they are
not very popular in FSO-based applications because of their bulky size and
unreliability. Later, solid-state lasers became the choice of FSO-based applications
– the most common being the neodymium/yttrium aluminum garnet (Nd/YAG). The
fundamental lasing wavelength of Nd/YAG laser is at 1064 nm, and it can be doubled
to 532 nm using nonlinear crystals. Other sources emitting close to Nd/YAG device
include neodymiu/yttrium aluminum phosphate (Nd/YAP) and neodymium/yttrium
lithium fluoride (Nd/YLF). The solid-state devices have stable and narrow spectral
linewidth, and they can be designed in a pulse or continuous mode configuration.
They have very high peak power (in the order of kilowatt or more) and can be
operated with very narrow spectral line width (< in the order of ns).
Semiconductor laser diode, e.g., gallium arsenide (GaAs), gallium aluminum
arsenide (GaAlAs), indium gallium arsenide (InGaAs), and indium gallium arsenide
phosphate (InGaAsP), devices can be used for some specific FSO-based applica-
tions. Another kinds of semiconductor lasers are vertical cavity surface emitting
laser (VCSEL), Fabry Perot laser, and distributed feedback laser. The threshold
current requirement in VCSEL is quite low, and it allows high intrinsic modulation
bandwidths in these lasers. Fabry Perot and distributed feedback (DFB) lasers have
higher power density .100 mW=cm2 / and are compatible with EDFA. So these
lasers find wide applications in FSO system. Semiconductor lasers exhibit single
frequency and one spatial mode. They are also compact in size, light weight, and
easy to operate. However, their output power is quite low, and therefore they require
additional amplifier for long distance communication. These diodes require great
3.1 Optical Transmitter 95

care in drive electronics, otherwise they will get damaged easily. Due to reliability
issues, such diodes require alternate or redundant laser source to allow smooth
functioning of the system. Another kinds of lasers are erbium-doped fiber laser
operating in the range of 965 to 1550 nm and master oscillator power amplifier
(MOPA)-based lasers. Depending upon the amplifier architecture, these lasers can
generate broadband or narrow linewidth output beams. The MOPA laser allows the
amplification of oscillator through a suitable and efficient amplification medium.
The oscillator and amplifier can then be individually tailored for high speed and
more power, respectively. The factors that may limit the usefulness of MOPA laser
is high nonlinear gain and low damage threshold for high power pulses. Fiber-based
amplifiers allow tens of kilowatt of peak power and have the advantage of ease of
use, efficient coupling to fibers, and relatively low noise power. However, nonlinear
effects like stimulated Brillouin scattering, stimulated Raman scattering, self-phase
modulation, cross-phase modulation, and four-wave mixing result in a lower SNR
due to both signal reduction and introduction of additional noise.
Among existing lasers, MOPA and solid-state lasers satisfy the requirements
for space-based applications. Other lasers like semiconductor lasers and EDFA
are useful in multi-gigabit links for near Earth laser communication. However,
fiber amplifiers have less peak power. While choosing a laser, a complex trade-
off has to be made between laser power, spectral width, output wavelength, range,
optical background noise, data rate, and modulation schemes to be used. Table 3.1
summarizes the various types of lasers used for FSO applications. The choice
of operating wavelength for the laser has to be made vis-a-vis availability of

Table 3.1 Attributes for lasers used in FSO [6]

Laser type Material Wavelength (nm) Data rate Peak power
Solid-state pulsed Nd/YAG 1064 <10 Mbps Very high
Nd/YLF 1047 or 1053 10–100 W
ND/YAP 1080
Solid-state mode locked Nd/YAG 1064 Up to 1 Gbps High
Nd/YLF 1047 or 1053 10s of Watt
ND/YAP 1080
Solid-state CW Nd/YAG 1064 >Gbps 1–5 W
Nd/YLF 1047 or 1053
ND/YAP 1080
Semiconductor pulsed GaAlAs 780–890 1–2 Gbps 200 mW
InGaAs 890–980 1–2 Gbps 1 W (MOPA)
InGaAsP 1300 and Multi-Gbps <50 mW
VCSEL 1550 10 Gbps <30 mW
Fabry Perot 780–850 40 Gbps 200 mW
DFB 1300 and 1500
Doped fiber amplifier Erbium DFA 1550 10s of Gbps 1000s of Watt
96 3 FSO System Modules and Design Issues

components, attenuation, background noise power, and detector sensitivity at that

wavelength. Most of the FSO systems are designed to operate in the 780–850 and
1520–1600 nm spectral windows due to low attenuation in these regions. Out of
these, 1550 nm is most widely used for data transmission because of various factors,
viz., eye safety, reduced solar background, and scattering. Further, at 1550 nm, more
power can be transmitted to overcome attenuation due to fog, smog, clouds, etc.
However, it suffers from certain challenges due to tight alignment requirement and
higher component cost.

3.1.2 Modulators

Modulators help in converting low-frequency information signal onto optical carrier

for long distance communication. The parameters that characterize optical modula-
tors are bandwidth, insertion loss, modulation depth, drive power, and maximum
optical throughput power [4]. Modulation can be applied to any one parameter,
i.e., intensity, phase, frequency, or the polarization of the optical carrier. The
most commonly used modulation is intensity modulation (IM) in direct detection
where the intensity of the optical source is modulated in accordance with the
source information. This can be achieved either by varying the direct current
of the optical source or by making use of an external modulator. The use of
external modulator allows high data rate transmission. However, external modulator
has major disadvantages of nonlinear response, complexity, and cost. In case of
coherent detection, the commonly used modulation schemes are intensity and phase
modulation schemes.
A phase modulation produces frequency sidebands on a continuous wave
(CW) optical beam. The amplitude of the sidebands with respect to the carrier
is determined by the amplitude of the applied voltage and are given by Bessel
functions. Here, the signal is applied as a voltage across the electrodes placed on
top and bottom of the electrooptic crystal. This will develop electric field across
the crystal. Polarized light when allowed to pass through the crystal will experience
index of refraction change that will lead to change in the optical path length which
is proportional to applied electric field. Therefore, the phase of the optical beam
moving out of the crystal will experience phase change due to applied electric field.
Figure 3.2 shows electrooptic phase modulator. This kind of modulators uses lithium
niobate (LiNbO3 ) and magnesium oxide-doped lithium niobate (MgO/LiNbO3 )
which have high electrooptic coefficients that minimizes the required drive voltages
and have high quality crystal. Also, such modulators have low insertion loss
of 4.5 dB and large throughput power of more than 500 mW. The modulation
bandwidth of 2 and 3 GHz can be achieved with a simple electrode structure which
can be increased further by using more advanced electrode structures.
3.1 Optical Transmitter 97

Fig. 3.2 Schematic diagram

of phase modulator

Fig. 3.3 Integrated optic Electrodes

LiNbO3 phase modulator [9]
E Light

E RF Waveguide

Lithium Niobate

The phase-modulating electrooptic modulator can be efficiently used as an

amplitude modulator. Mach-Zehnder modulator is one of that kind. It consists of
Y-splitter which splits the incoming light into two branches. One of the branches
is having phase modulator. The output light from both the branches are combined
using another Y-adder. The phase of the incoming light is varied in accordance with
the modulating signal. The light from both the branches will then combine either
destructively or constructively depending upon the phase shift being introduced
in one of the branch. This therefore controls the amplitude or the intensity of the
incoming light resulting in amplitude-modulated output wave.
Integrated optical modulators (IOM) are also good choice for FSO communi-
cation and are available over broad range of wavelengths. These modulators are
made from LiNbO3 and are constructed using simple dielectric optical waveguide.
Figure 3.3 shows the concept of integrated phase modulator. The device consists of
waveguide on LiNbO3 chip and electrode system. In the presence of electric field,
the change in the travel time dt of light wave is given by

dt D dn  L=c; (3.1)

where dn is the absolute change in the index of refraction due to applied electric
field, L the interaction length, and c the speed of light in vacuum. This propagation
delay is equivalent to shift in phase of the output light which is given by

d D !  dt D !  dn  L=c; (3.2)

where ! is the angular frequency.

98 3 FSO System Modules and Design Issues

Fig. 3.4 Mach-Zehnder LiNbO3 chip

Optical wave guide Housing
amplitude modulator

Fiber Fiber

Y-junction Y-junction

Signal input Electrode system External termination

Similarly, integrated optics amplitude modulators constructed by patterning a

Mach-Zehnder interferometer on a LiNbO3 substrate is shown in Fig. 3.4. For an
ideal amplitude modulator, the total optical power at the output of the modulator is
given by

1
Po D Pi Œ1 C cos .d/ ; (3.3)
2
where Pi is the input optical power and d the phase difference between two paths.
Besides integrated optic phase and amplitude modulators, there are modulators with
combinations of both phase and amplitude modulators on the same chip. The design
and construction of these modulators are complex than simple IOM modulators;
however, their basic principle of operation is same as that of linear electrooptic
modulators.

3.1.2.1 Modulation Schemes

There are variety of modulation schemes that are suitable for FSO communication
system. Both binary and multi-level modulation signaling schemes can be used in
conventional FSO communication system. However, most popularly used modula-
tion scheme is based on binary signaling for which system design is simple and
inexpensive. In binary signaling scheme, the information is transmitted in each
symbol period by variation of two intensity levels. On-off keying (OOK) and binary
pulse-position modulation (PPM) are the most popularly used binary signaling
modulation schemes in FSO communication. These modulation schemes have been
discussed in Chap. 1.
The variants of PPM scheme which are becoming popular these days are
differential PPM (DPPM), differential amplitude PPM (DAPPM), overlapping PPM
(OPPM) [8], and combinatorial PPM (CPPM) [1]. All these modulation schemes
are obtained by simple modification of PPM to achieve improved power and
bandwidth efficiency. In DPPM, the empty slots following the pulse in PPM symbol
3.2 Optical Receiver 99

are removed therefore reducing the average symbol length and improving the
bandwidth efficiency. For long sequence of zeros, there could be a problem of
slot synchronization; however, this problem can be overcome by making use of
guard band/slot immediately after the pulse has been removed [3]. The DAPPM
is a combination of DPPM and pulse amplitude modulation (PAM). It is therefore a
multi-level modulation scheme where the symbol length varies from 1; 2; : : : ; M
and the pulse amplitude is selected from 1; 2; : : : ; A levels.
Before transmitting the optical-modulated signal, its power level has to be raised
in order to compensate the huge loss in the atmospheric channel. For this purpose,
optical amplifier can be used in FSO communication system as they enable the direct
amplification of light with a minimum of electronics. An optical post amplifier can
boost the output power of a transmitter by about 15–20 dB. Further, at front end
of the receiver, the received optical signal may be very weak. The usage of optical
preamplifier before the photodetector may improve the receiver sensitivity by about
10 dB. The optical post and preamplifiers are briefly discussed in Sect. 3.3.

3.2 Optical Receiver

The receiver helps to recover the transmitted data after propagating through
turbulent atmosphere. It consists of receiver telescope, filter, photodetector, signal
processing unit, and demodulator. The receiver telescope comprises of lenses that
focus the received optical signal onto the photodetector. The filter is used to reduce
the amount of background noise. The noise sources present at the receiver include
background, detector dark current, preamplifier, signal shot noise, and thermal
noise. The photodetector converts the received optical signal into electrical signal
which is passed on to the processing unit and then to the demodulator. In the
receiver, both PIN and APD can be used. In the FSO uplink, the received power
level is quite low due to large free-space loss. At this power level, an APD receiver
gives much better performance than the PIN receiver.
The choice of optical receiver depends upon various fundamental issues and
hardware parameters. Some of the important parameters are listed below:
(i) Modulation technique: The detection technique used at the receiver depends
upon the modulation format. Not every detection technique is suited for every
modulation format, e.g., direct detection receivers are insensitive to phase and
polarization information.
(ii) Hardware availability, reliability, and cost: Different types of receiver have
different hardware requirements which may or may not be readily available
at reasonable cost. For example, high gain Si-APD works efficiently only
at wavelength below 1000 nm. At higher wavelengths, other detectors like
InGaAs/InGaAsP can be used depending upon the requirement.
100 3 FSO System Modules and Design Issues

(iii) Receiver sensitivity: This is a very important parameter in all optical commu-
nication system including FSO communication system. It is measured in terms
of average received photons per bit and is given as

Pav
nav D ; (3.4)
hR

where h is the photon energy at the transmit operating wavelength ( D c=)

and Rb the bit rate/data rate. The receiver sensitivity depends mainly on photon
detection technique, modulation format, photodetector, and background noise.
These are briefly described below:
(a) Photon detection technique: As mentioned earlier, the detection technique
used in the optical receiver can be broadly classified into two types, viz.,
coherent detection and noncoherent direct detection. In coherent detection
receiver, incoming signal is mixed with the strong local oscillator (LO)
signal. This mixing of weak signal and strong LO signal at the front
end of the receiver provides linear 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.
(b) Modulation format: The type of modulation format used for FSO link
affects the sensitivity of the receiver. The coherent receiver, in particular
heterodyne type, can be used for all kinds of modulation schemes. On
the other hand, the homodyne receivers can be used only for intensity
and phase-modulation schemes. The direct detection receiver can be used
to detect only intensity-modulated signals, e.g., OOK or PPM. For long
distance FSO link, a major concern is the higher receiver sensitivity at
comparatively lower data rates. It is easily achievable with PPM scheme
as it provides high peak-to-average power ratio. However, the PPM scheme
is not bandwidth efficient. Therefore, except when the transmitter is peak
power limited or when the system is modulation bandwidth limited, most
of the FSO communication systems are based on PPM scheme [11].
(c) Photodetector and background noise: There are various types of noise
sources which contribute noise in the FSO receiver. These noise sources
include background noise, detector dark current noise, signal shot noise,
and thermal noise. The contributions of these noise sources in the final
output depend upon the optical design, receiver configuration, data band-
width, and type of FSO link. A brief description on these noise sources is
given below.
3.2 Optical Receiver 101

Fig. 3.5 Geometry of (a)

extended source when
FOV < S and (b) stellar or
point source when
FOV > S

Background noise: The noise that is detected by the photodetector due to the
surrounding environment is called background noise. The main sources of back-
ground noise are (a) diffused extended background noise from the atmosphere, (b)
background noise from the Sun and other stellar (point) objects, and (c) scattered
light collected by the receiver. Figure 3.5 shows the geometry of point and extended
sources relative to the receiver. The background power collected by the receiver due
to diffused, stellar, and scattered noise is given by
8
ˆ
<HB FOV LR AR filter
ˆ .Diffused or extended source/
PB D N L A  .Stellar or point source/ : (3.5)
ˆ B R R filter
:̂ I  L A  .Scattered noise source/
 FOV R R filter

In the above equation, HB and NB are the background radiance and irradiance energy
densities of large extended angular sources and point sources, respectively. The term
HB is expressed in units of W/m2 =sr/Å and NB in terms of W=m2 =Å. The parameters
FOV is the solid angle of receiver field of view (FOV), LR the transmission loss of
the receiver optics, AR the effective area of the receiver, and filter the bandwidth
of optical BPF in the receiver. In case of scattered noise source,  represents
the atmospheric attenuation coefficient and I the exo-atmospheric [2] (region of
space outside the earth’s atmosphere) solar constant (0.074 W/cm2 m). A strong
background source near the receiver FOV can lead to significant scattering. For
an optical receiver design with optics under direct exposure to sunlight, the major
background noise source is due to scattering. The contribution of celestial bodies
such as the Sun, the moon, the star, etc., are assumed to be almost negligible
except in deep space optical communication. With background noise power PBG ,
the background noise current variance in an APD is given as
2
BG D 2qR0 PBG M2 FB .A2 /: (3.6)
102 3 FSO System Modules and Design Issues

In the above equation, B is the signal bandwidth. The parameter F is the excess
noise factor which depends on the detector material and several other parameters.
The commonly used expression for determining F as a function of APD gain, M,
is given below

F .M/ D Mx ; (3.7)

where the value of x lies between 0 x  1:0 and is dependent upon the
material. The value of x is 0.3 for silicon, 0.7 for InGaAs, and 1.0 for Germanium
photodetector. For PIN photodetector, the values of M and F are unity. The other
parameters used in the equation are defined earlier.
Dark current noise: When there is no optical power incident on the photodetector,
a small reverse leakage current still flows through the device. This is called dark
current and contributes to total system noise. The detector dark current noise
variance in an APD receiver is given by
 
d2 D 2qM2 FIdb B C 2qIds B A2 : (3.8)

In the above equation, Idb and Ids are the bulk and surface dark currents, respec-
tively. The dark current noise is dependent upon the operating temperature of the
photodetector and its physical size. The dark currents can be reduced by cooling the
detector or by reducing the physical size of the detector. It may be mentioned that
avalanche multiplication is a bulk effect, and surface dark current is not affected by
the avalanche gain, and therefore it can be neglected. In PIN photodetector, the dark
current noise variance is given by
 2
d2 D 2qId B A : (3.9)

The typical dark current values for some most commonly used detectors are given
in Table 3.2
Signal shot noise: The number of photons emitted by an optical source at a given
time is not constant. Therefore, the detection of photon by a photodetector is a
discrete process (since the creations of electron-hole pairs result from the absorption
of a photons) and is governed by the statistics of the photon arrival. The statistics of
photons arriving at the detector follows the discrete probability distribution which is
independent of the number of photons previously detected and is given by Poisson

Table 3.2 Typical values of Material Dark current (nA)

dark current for various
materials PIN APD
Silicon 1–10 0.1–1
Germanium 50–500 50–500
InGaAs 0.5–2.0 10–50
3.2 Optical Receiver 103

distribution. The fluctuations of the photons at the detector result in quantum or

signal shot noise. The variance of signal shot noise current in an APD is expressed
in terms of average signal photocurrent Ip as
 2
s2 D 2qM2 FIP B A ; (3.10)

where Ip D R0 PR .
Thermal noise: This is the spontaneous fluctuations due to thermal interaction of
electrons in any receiver circuit that consists of resistors at specific temperature. The
thermal noise variance in the receiver due to resistor RL can be expressed as

2 4KB TB  
Th D A2 ; (3.11)
RL

where KB is Boltzmann’s constant, T the absolute temperature, and RL the equivalent

resistance of the circuit.

3.2.1 Types of Detectors

Based on the applications, the characteristics of the detector used are different and
they can be classified into two types, i.e., communication and beacon detectors.
The detectors used in FSO communication system for communication and beacon
purposes are listed in Table 3.3.
(i) Communication Detectors: In an FSO communication system, commonly
used detectors are PIN and APD. The APD provides moderate front end
gain (50 to 200), low noise, and good quantum efficiency, depending upon
the operating wavelength. The Si-APD has low noise, but suffers from low
quantum efficiency above 1000 nm. However, InGaAs- and InGaAsP-based

Table 3.3 Communication and beacon detectors in FSO link

Application Detector type Material used
Communication APD Silicon, InGaAs, InGaAsP
PIN Silicon, InGaAs, InGaAsP
CCD Silicon
PMT Solid-state silicon
photo-cathode
Beacon (acquisition and tracking) CCD Silicon
CID Silicon
QAPD Silicon
QPIN Silicon, InGaAs
104 3 FSO System Modules and Design Issues

APDs provide good quantum efficiency beyond 1000 nm. They are mostly
used in the 1300 and 1550 nm operating wavelength range. However, the
noise performance of these devices is very poor and as a result, Si devices
normally outperform the longer wavelength devices even with low quantum
efficiency. Besides APD, the PIN photodetector is also most widely used for
direct detection communication. The gain of PIN photodetector is unity, but
it provides good quantum efficiency. It is preferred for coherent applications
where the required gain is provided by the LO. PIN photodetector can be
fabricated on Si, InGaAs, and InGaAsP. At lower wavelengths, InGaAs and
InGaAsP PIN photodetectors are used as they provide high quantum efficiency.
Another type of detector used in early days was photomultiplier tubes (PMTs).
These devices provide very high front-end gain (in the order of 105 ) and very
low noise. However, due to heavy weight and large size, they are not suitable
for the space-based applications. Another type of detector are charge-coupled
devices (CCDs) which are good for low data rate as data rates depends upon the
read-out speed of the CCD. Due to this reason, they are best suited as spatial
acquisition detectors.
(ii) Beacon Detectors: Besides communication, ATP is also very essential for
establishing an FSO communication link. Various types of spatial acquisition
and tracking detectors are used at the receiver to detect the beacon signal
that is searching over a large region of uncertainty. As discussed earlier,
CCDs or charge injection devices (CIDs) configured as large array of detectors
are used to perform both spatial acquisition and tracking functions for FSO
communication system. The CCD-based array provides wide FOV and requires
only one steering cycle. The FOV of beacon detector has to be chosen very
carefully as large FOV will lead to large amount of background noise, and
small FOV will result in long search patterns/time. Another type of detector
used for detecting beacon signal is position-sensitive detector, e.g., quadrant
APD and PIN detectors. The quadrant photodetectors are 2  2 array of
individual photodetectors, separated by a small gap, fabricated on a single
chip. This small spacing between the quadrants is also called dead zone.
Ideally, the dead zone should be infinitely small, but in practice it is typically
in the order of 50 to 100 m. For large FOV, the spot size incident on the
detector’s dead zone could be lost. Therefore, instead of a finite dead zone
between the quadrants, a transition across the boundaries of the quadrant
is employed to avoid any abrupt change. These devices suffer from lot of
crosstalk which degrade the performance. Quadrant APDs are generally used
if receiver sensitivity requirement is high. Figure 3.6 shows the standard dead
zone and sharing transition in quadrant APD.
3.2 Optical Receiver 105

Fig. 3.6 Quadrant APD

showing standard dead zone
and shared transition

Table 3.4 Comparison between coherent and noncoherent receiver configurations

Parameters Coherent Noncoherent
Modulation parameters Amplitude and phase Intensity
Detection techniques Heterodyne or homodyne detection Direct detection (DD)
Adaptive control Necessary Not necessary

3.2.2 Receiver Configuration

The performance of any communication system for a given link distance and
data rate depends upon receiver configuration as well as modulation scheme
used. FSO communication uses either noncoherent (direct detection) or coherent
(homodyne or heterodyne) receiver detection techniques. For intensity modulation,
a simple receiver configuration, i.e., noncoherent detection, can be used. However,
for digital modulation schemes like phase-shift keying (PSK) or frequency-shift
keying (FSK), coherent receiver configuration is used which are more complex
than noncoherent receiver configuration. Coherent receivers have advantage over
noncoherent receivers in regard to their increased sensitivity level, data rates, and
transmission distance. But due to their increased complexity in receiver design
(as they have to restore the in-phase and quadrature components of the received
optical fields/state of polarization of the signal) and difficult implementation, this
configuration is not widely used in FSO communication system. Also, sometimes
due to poor mixing of incoming signal and LO signal, the level of improvement
in coherent receivers decreases significantly. Although coherent system is not a
preferred choice for FSO communication, however, when dealing with very high
data rates (100 Mbps or greater) or for power limited systems, these systems are
given preference over direct detection system. Comparison between coherent and
noncoherent receiver configurations is given in Table 3.4.
In this section, general expressions for output SNR in terms of required optical
power, PR in the presence of background noise for different receiver configurations,
i.e., noncoherent detection (direct detection) or coherent detection (homodyne or
heterodyne detection) are studied. Various receiver configurations considered are
106 3 FSO System Modules and Design Issues

Fig. 3.7 A general optical communication receiver valid for all configurations

(i) coherent PSK homodyne receiver, (ii) coherent FSK heterodyne receiver, (iii)
direct detection (PIN + Optical Amplifier) for OOK, (iv) direct detection APD for
OOK, and (v) direct detection (APD) for M-PPM. A general schematic diagram
valid for all the abovementioned receiver configurations is shown in Fig. 3.7.

3.2.2.1 Coherent PSK Homodyne Receiver

The homodyne PSK receiver has the highest receiver sensitivity. In case of coherent
receivers, photodetector in the receiver configuration act as a mixer for incoming
optical signals and background noise (i.e., within the range of BPF). Therefore, it
produces signal-background, background-background beat noise components along
with the desired signal component. In addition to above noise components, the
presence of LO will produce LO-background beat noise component too. All these
noise components will collectively degrade the performance of FSO system. For
homodyne receiver configuration, the frequency of LO is the same as that of
incoming signal frequency.
For SNR analysis of PSK homodyne receiver, optical amplifier and IF stage
blocks in Fig. 3.7 will not be used. The photocurrent is given as

ip / jER .t/ C EBG .t/ C EL .t/j2 ; (3.12)

where Es , EBG and EL are the electric field strengths due to signal, background,
and LO, respectively. Following [7], various signal and noise components under the
condition that optical bandwidth 4f is much greater than electrical bandwidth,
B.D 1=Tb , where Tb is the bit duration), are given as
p
Signal photocurrent W ip .t/ D 2R0 PR PL cos .t/ ; (3.13)


DC current W Idc D R0 PR C PL C mt Sn 4f ; (3.14)

 4KB TBFn
RMS noise current: i2bg D 2qIBG B C R20 Sn 2PR C2PL C mt Sn 4f 2B C ;
RL
(3.15)
3.2 Optical Receiver 107

where  .t/ D L  R . The parameters PR and R are the power and phase of
the received signal, respectively. Similarly, PL and L give the power and phase
of LO signal, respectively. Other parameters, viz., R0 is the responsivity, mt are
the number of background modes as seen by receiver FOV, Sn the background
noise power spectral density per spatial mode, B the electrical bandwidth, 4f
optical filter bandwidth, q electronic charge, IBG background noise current, KB
Boltzmann’s constant (D1:38  1023 J/K), T absolute temperature in Kelvin, Fn
noise figure of IF stages, and RL photodetector load resistance. In Eq. (3.14),
the first two terms arise due to received and LO signals, respectively. The third
term represents DC component of the current arising due to background noise. In
Eq. (3.15), the first term represents the shot noise. The second, third, and fourth
terms represent the contribution due to signal-background, LO-background, and
background-background beat noise, respectively. The last term in the equation
represents the thermal noise contribution of the photodetector load resistance and
the following stages.
The output SNR for binary PSK signaling scheme from Eqs. (3.13), (3.14),
and (3.15) is given by

S i2p .t/
D : (3.16)
N i2bg

After further simplification, the above equation becomes [5]:

S 2PR
D ; (3.17)
N hBFh

where
 
1 PR Sn f Sn PR Sn f KB TFn
Fh D 1C C C2 1C C C ; (3.18a)

PL PL h 2PL 2PL hLm

and
1 2
Lm D R PL RL : (3.18b)
2
In the above equations,
is the quantum efficiency, and  the frequency of received
signal. In balanced homodyne receiver, signal-background and background-
background noise and ASE-ASE components will cancel out. Therefore, Eq. (3.18a)
reduces to

1 PR Sn f Sn KB TFn
Fh D 1C C C2 C : (3.19)

PL PL h hLm
108 3 FSO System Modules and Design Issues

Fig. 3.8 Variation of Pe with 10

0

average PR for coherent PSK

λ = 1064 nm FSK
receivers 10
−2

Bit Error Rate

−4
10

−6
10

−8
10

−10
10
−60 −58 −56 −54 −52 −50 −48 −46 −44
Average Power, P dBm
R

The probability of error, Pe , is therefore given by:

 1=2
1 S
Pe D erfc : (3.20)
2 N

The variations of Pe with PR are computed from the Eq. (3.20) for
D 0:7; f D
5 nm, Sn D 1:9  1020 W/Hz,  D 1064 nm, Fn D 2 dB, T D 300 K, RL D
100 ; B = 1 GHz (or data rate, Rb D 1 Gbps), and PL D 10 dBm and are shown in
Fig. 3.8.

3.2.2.2 Coherent FSK Heterodyne Receiver

For heterodyne receiver, intermediate frequency (IF) stage is used in Fig. 3.7. Here,
the frequency of LO, !L , is offset from incoming signal frequency, !s , by a factor
equal to intermediate frequency, !IF . Therefore, the received signal photocurrent at
photodetector will be
p
Signal photocurrent W ip .t/ D 2R0 PR PL cos Œ.!L .t/  !s .t// C  .t/
p
D 2R0 PR PL cos Œ!IF .t/ C  .t/ :
(3.21)
The analysis for coherent FSK heterodyne receiver is similar to that of coherent PSK
homodyne receiver but the performance curve will shift toward the right by 6 dB.
Figure 3.8 shows the probability error curve for orthogonal FSK signaling scheme.

3.2.2.3 Direct Detection (PIN + OA) Receiver for OOK

Direct detection (PIN + OA) receiver follows the same block diagram as shown in
Fig. 3.7 except that IF stage and LO blocks will not be used here. Various signal and
noise components under the same condition as in homodyne receiver, i.e., f
B,
are given as
3.2 Optical Receiver 109

Signal photocurrent W Ip D GR0 PR ; (3.22)


DC current W Idc D R0 GPR C GSn f C Psp f ; (3.23)

where G is the gain of optical amplifier, Psp is the spontaneous noise power at the
output of the amplifier and is given by

Psp D nsp .G  1/ h; (3.24)

where nsp is population inversion parameter of optical amplifier. Other parameters

are same as mentioned above.

RMS noise current W


12 for bit 0 10 D 2qIdc B C R20 GSn 2GPR C GSn f 2B (3.25a)


CR20 Psp 2GPR C Psp f C GSn f 2B
C4KB TBFn =RL

and
 
02 for bit 0 00 D qR0 Psp f C GSn f 2B C R20 n2sp f 2B
CR20 G2 Sn2 f 2B C 4KB TBFn =RL : (3.25b)

For OOK signaling scheme, Pe for bit ‘1’ and ‘0’ are given by

1 h p i
Pe1 D erfc .R0 GPR  Th/ = 21 (3.26)
2
and
1 h p i
Pe0 D erfc Th= 20 ; (3.27)
2
where Th is the threshold level. If an optimum threshold level which equalizes Pe1
and Pe0 is used, then Pe from Eqs. (3.26) and (3.27) will be

1
Pe D Pe1 D Pe0 D erfc .S=2N/1=2 ; (3.28)
2
where

S R20 G2 P2R
D : (3.29)
N .1 C 0 /2
110 3 FSO System Modules and Design Issues

The SNR for this receiver configuration after simplification is given as [5]

S PR
D p p 2 ; (3.30)
N 2hB Fp1 C Fp0

where
  
1 Sn f n0sp hf Sn Sn f n0sp hf
Fp1 D 1C C C2 n0sp C  1C C

G PR PR h 2PR 2PR
KB TFn
C (3.31)
hG2 Lm
0

and
  
1 Sn f n0sp hf 0 Sn Sn f n0sp hb
Fp0 D C C 2 nsp C  C

G PR PR h 2PR 2PR
KB TFn
C : (3.32)
hG2 Lm
0

The other parameters are

 
1
n0sp D 1 nsp (3.33)
G

and

0 1 2
Lm D R PR RL : (3.34)
2 0
All other parameters in Eqs. (3.30), (3.31), (3.32), (3.33), and (3.34) are defined
earlier and are same as in PSK homodyne receiver. The variation of Pe with average
PR for (PIN + OA) direct detection receiver at  D 1064 nm is shown in Fig. 3.9.

3.2.2.4 Direct Detection (APD) Receiver for OOK

This receiver configuration follows the same block diagram as shown in Fig. 3.7
except that IF stage and LO blocks will not be there. The detector will be APD
instead of PIN. In this case, various signal and noise components under the condition
f
B are given by

Signal photocurrent W Ip D R0 MPR ; (3.35)


DC current W Idc D M2 FR0 PR C Sn f ; (3.36)
3.2 Optical Receiver 111

Fig. 3.9 Variation of Pe with

average PR for direct
detection receivers

RMS noise current W


12 for bit ‘1’ D 2qIdc B C M2 R20 Sn 2PR C Sn f 2B (3.37)
C4KB TBFn =RL

and

02 for bit ‘0’ D qM2 FR0 Sn f 2B C M2 R20 Sn2 f 2B (3.38)
C4KB TBFn =RL :

where M and F are multiplication gain and excess noise figure of APD as defined
earlier. Following the same approach as in direct detection (PIN + OA) receiver for
OOK, SNR can be analyzed by using Eq. 3.30 with Fp1 and Fp0 replaced by Fa1 and
Fa0 , respectively. Fa1 and Fa0 are given by [5]
 
F Sn f Sn Sn f KB TFn
Fa1 D 1C C2 1C C (3.39)

PR h 2PR hM2opt Lm
0

and
 
F Sn f Sn Sn f KB TFn
Fa0 D C2 C ; (3.40)

PR h 2PR hM2opt Lm
0

where
4KB TFn=RL
opt D
MxC2  : (3.41)
xqR0 PR C Sn f
112 3 FSO System Modules and Design Issues

Mopt is the optimum value of multiplication gain, M which maximizes SNR. All
other parameters are same as defined earlier. The variation of Pe with average PR
for (APD + OOK) direct detection receiver for x D 0:5 is shown in Fig. 3.9.

3.2.2.5 Direct Detection (APD) for M-PPM

As discussed earlier in Chap. 1, in M-PPM scheme, each word contains n bits of

information, i.e., M D 2n . Therefore, a complete data word is divided into M slots
of duration Ts seconds, and information is placed in any one of these slots. The
bandwidth occupied by M-PPM is given by

1
BPPM D ; (3.42)
Ts

where
Tb log2 M
Ts D : (3.43)
M
Hence, it is seen that bandwidth occupied by PPM is more than bandwidth occupied
by OOK (D T1b ). Using the same approach as in OOK, the probability of error Pe
for bits ‘1’ and ‘0’ in 2-PPM scheme are given by
" #
1 R20 M2 P2R
Pe D Pe1 D Pe0 D erfc   :1=2 (3.44)
2 2 12 C 02

This equation can be written in terms of SNR as

S R2 M2 P2R
D  02 : (3.45)
N 1 C 02

Substituting 12 and 02 from Eqs. (3.37) and (3.38), respectively, and after further
simplification and replacing B with BP , the above equation can be written as

S PR
D : (3.46)
N 2hBPPM .Fa1 C Fa0 /

In M-PPM, decision regarding the presence of a pulse in a time slot is made on the
basis of M  1 comparisons. The probability of wrongly decoding a word, Pew , is
given by

Z "Z #m1
1 .R0 MPR CI1 /
Pew D 1  p .I0 / dI0 p .I1 / dI1 ; (3.47)
1 1
3.3 Optical Post and Preamplifiers 113

where p .I1 / and p .I0 / are the probability density functions of I1 and I0 ; respectively.
These are considered to be Gaussian distributed with variance 12 and 02 , respec-
tively. After simplification, Eq. (3.47) reduces to
Z 1  M1 
1 1 R0 MPR C I1 I12
Pew D 1  q P 1  erfc p  exp  2 dI1 :
212 1 2 20 21
(3.48)
This Pew can be used to obtain the upper bound on the probability of error, Pe :
Relationship between Pew and Pe is given as

M=2
Pe  Pew : (3.49)
M1

If the second term inside the square bracket in Eq. (3.48) is very small ( 1/; then
Eq. (3.48) can be approximated as
" #1=2
M1 R20 M2 P2R
Pew  erfc   : (3.50)
2 2 12 C 02

3.3 Optical Post and Preamplifiers

Both post-amplifier and preamplifier, in addition of amplifying the input signal, will
add to it a signal due to spontaneous emission of light. A portion of this spontaneous
emitted light is in the same direction as the signal and gets amplified along with the
main signal. This added light, called the amplified spontaneous emission (ASE)
noise, is spread over a wide frequency range. The spontaneous noise power at the
output of the amplifier is given by

Psp D .G  1/ nsp hBo ; (3.51)

where nsp is spontaneous emission factor (or sometimes called population inversion
factor), G the amplifier gain, h the Planck’s constant,  the incoming frequency, and
Bo the optical filter bandwidth. The output power given by Eq. (3.51) is for each
polarization mode. Since we have two fundamental polarization modes, the total
noise power at the output of the amplifier will be 2Psp . The value of nsp depends
upon the population inversion and is given by

N2
nsp D ; (3.52)
N2  N1

where N1 and N2 are the atomic population in ground and excited states, respec-
tively.
114 3 FSO System Modules and Design Issues

The amplified output signal plus ASE noise when detected by a photodetector
consists of desired signal component along with the thermal, shot, and beat
noise components. Among shot noise are signal shot noise, ASE shot noise, and
background shot noise. The beat noise components consists of signal-background
beat noise, background-background beat noise, ASE-ASE beat noise, and signal-
ASE beat noise. The variances of all the noise components, namely, thermal noise,
shot noise, signal-ASE beat noise, ASE-ASE beat noise, and amplified background
noise, at the receiver output are given by [10, 12]:

4KB TFn Be
th2 D ; (3.53)
RL
2
sigshot D 2qR0 GPR Be ; (3.54)
2
ASEshot D 2qIsp Be D 2qR0 Psp Be D 2qR0 .G  1/ nsp hBe Bo (3.55)

and
2
sigBG D 2qIBG Be D 2qR0 GPBG Be : (3.56)

Combining Eqs. (3.54), (3.55), and (3.56), total variance due to shot noise is given
by
2 2 2 2
shot D sigshot C ASEshot C sigBG ; (3.57)
2
 
shot D 2qR0 GPR C .G  1/ nsp hBo C GPBG Be : (3.58)

Variance of various beat noise components are given as

2
sigASEbeat D 4R20 GPR nsp h .G  1/ Be ; (3.59)
2
 
sigBGbeat D 4R20 G2 PR nsp h .G  1/ Be ; (3.60)
2

2
ASEASEbeat D 2R20 nsp h .G  1/ .2Bo  Be / Be ; (3.61)
2
BGBGbeat D R20 NBG
2
G2 .2Bo  Be / Be ; (3.62)
2
 
ASEBGbeat  2R20 GNBG nsp h .G  1/ .2Bo  Be / Be : (3.63)

In the above equations, NBG is the background power spectral densities (W/Hz) at
the input of optical preamplifier. Therefore, total variance is given as
2 2 2 2 2 2
total D shot C sigASEbeat C sigBGbeat C ASEASEbeat C BGBGbeat
2
CASEBGbeat C th2 : (3.64)
3.4 Link Design Trade-Off 115

Since the amplifier gain is reasonably large, the contribution due to shot noise and
thermal noise are negligible as compared to signal-ASE and ASE-ASE beat noise.
The ASE-ASE noise can be made very small by reducing the optical bandwidth, Bo :
Therefore, the dominant noise component is usually signal-ASE beat noise. In that
case, the SNR at the amplifier output is given by the following equation:

.R0 GPR /2 .R0 GPR /2

SNRo D 2
D : (3.65)
sigASEbeat 4R2 GPnsp h .G  1/ B

In the above equation, B is the electrical bandwidth which is same as Be .

3.4 Link Design Trade-Off

In the design of FSO link for a given requirements, some trade-offs have to be made
among various design parameters. These are discussed below.

3.4.1 Operating Wavelength

The choice of operating wavelength depends upon many factors that includes:
(i) Availability of laser: While choosing the laser for any system, we need
to consider peak-to-average power ratio, available peak power, electrical-to-
optical conversion efficiency, and overall power consumption. Therefore, the
trade-off between available laser technologies which strongly depend upon
operating wavelength should be made to identify the appropriate choice.
(ii) Gain vs. beamwidth: In general, the gain of optical transmit or receive antenna
is given as G  .DR=/2 : Thus gain is inversely proportional to the operating
wavelength, and hence it is desirable to work at lower operating wavelengths
to get more gain. However, beamwidth of the system is proportional to .=D/.
This implies that at lower operating wavelength, beamwidth will be narrower
leading to increase in pointing errors. Therefore, a trade-off between higher
gain and reducing signal fades due to pointing error has to be considered.
(iii) Atmospheric absorption and scattering: The atmospheric absorption and scat-
tering depend upon the choice of operating wavelength. When a light beam
travels through the atmosphere, it may be absorbed or scattered by the
constituent particles of the atmosphere. Only the wavelengths outside the
main absorption band can be used for optical communication. The region of
maximum absorption is called forbidden band, and the region used for optical
communication is called atmospheric transmission band. The transmission
window for FSO communication system is in the visible and near-infrared
region that stretches roughly from 750 to 1600 nm. However, certain wave-
lengths in the near-infrared region suffer from strong atmospheric absorption
due to the presence of the water particles (moisture). The contribution of gas
116 3 FSO System Modules and Design Issues

absorption due to oxygen and carbon dioxide to overall absorption coefficient

can be neglected as the gas-specific absorption coefficient is very small as
compared to water absorption. However, at longer infrared wavelength range
.>2000 nm/, gas absorption can dominate the absorption properties of the
atmosphere. There are several transmission windows that are nearly transparent
between 750 and 1600 nm. These are 850, 1060, 1250, and 1550 nm. The
850 and 1064 nm are characterized by low attenuation windows with reliable
and inexpensive transmitter and receiver components. However, at 1064 nm,
Nd/YAG laser is preferred, and because of this FSO systems at this wavelength
are bulky. The 1250 nm window is very rarely used as it shows a drastic
increase in absorption at 1290 nm leaving a very small workable band. The
1550 nm is very well suited for high-quality and low-attenuation transmission.
The components in this range are easily available and reliable. The background
noise contribution at the receiver output decreases with the increase in the
operating wavelength. Consequently, 1550 nm wavelength is less vulnerable
to background noise.
(iv) Detector sensitivity: The sensitivity of the PIN detector is determined by its
detection efficiency. In case of APD, the detection sensitivity depends upon
gain M; quantum efficiency
, and excess noise factor F. A good APD detector
is characterized by high gain, large bandwidth, high efficiency, and low excess
noise factor. However, the availability of the detector is very much limited
by the operating wavelength. For example, Si detectors can provide a very
high gain bandwidth and low excess noise factor, but they have less detection
sensitivity at 1500 nm.

3.4.2 Aperture Diameter

The power efficiency in FSO link depends on the transmitter and receiver aperture
areas. In order to reduce the transmit power requirement, it is desirable to have
larger receiver aperture size. However, the receiver aperture area cannot be increased
indefinitely as it will enhance the background noise contribution and leads to
increase in mass of the terminal. Further, the size of aperture area affects the pointing
requirement. Since the beamwidth is inversely proportional to the transmitter
aperture diameter, larger aperture size will require tighter pointing accuracy and
higher sensitivity toward pointing loss. Therefore, an optimum choice of aperture
diameter has to be made in order to increase the power efficiency in FSO system.

3.4.3 Receiver Optical Bandwidth

The FSO link performance can be improved by reducing the background noise with
the help of BPFs. The filter bandwidth should be sufficient enough to pass the
information signal without any distortion. It should not be very wide, otherwise
3.5 Summary 117

Table 3.5 Parameters for FSO link design

S.No. Link budget Parameters
1. Transmitter parameters Operating wavelength
Transmit power
Transmitter aperture area
Transmitter optical efficiency
Transmitter antenna gain
2. Channel losses and noise Pointing loss
Atmospheric losses
Scintillation-induced loss
Atmospheric background noise
Free-space loss
3. Receiver parameters Receiver aperture area
Receiver sensitivity
Receiver optics efficiency
Receiver detector FOV
Receiver antenna gain
Bandwidth of narrow BPF
Optical post and preamplifier emission noise, i.e., ASE noise

background noise contribution will increase. Therefore, a BPF of appropriate

bandwidth is to be designed keeping in view the practical constraints.
Various parameters which are to be considered in the FSO link design are given
in Table 3.5.

3.5 Summary

The performance of FSO communication system depends directly on the efficiency

and sensitivity of optical components, i.e., optical transmitter, optical modulator,
optical amplifier, and optical receiver used in link design. The fundamental char-
acteristics of these optical components and associated noise sources are discussed
in this chapter. Receivers with both direct detection (average power limited) and
coherent detection (high power efficient) using various modulation format are
presented. The minimum required average optical power, i.e., receiver sensitivity
for achieving a given performance at a certain data rate, has been evaluated for PIN
and APD receivers for PSK, FSK, and OOK modulation formats. Link design trade-
offs has been discussed in order to provide a fair understanding of design for FSO
communication system.
 118 3 FSO System Modules and Design Issues

Bibliography

1. J.M. Budinger, M.J. Vanderaar, P.K. Wagner, S.B. Bibyk, Combinatorial pulse position
modulation for power-efficient free-space laser communications, in Proceedings of SPIE
(1993), pp. 214–225
2. A. Farrell, M. Furst, E. Hagley, T. Lucatorto et al., Surface and exo-atmospheric solar
measurements. Technical report, Physical Measurement Laboratory
3. Z.Z. Ghassemlooy, A.R. Hayes, B. Wilson, Reducing the effects of intersymbol interference
in diffuse DPIM optical wireless communications. Optoelectron. IEE Proc. 150(5), 445–452
(2003)
4. H. Hemmati, Near Earth Laser Communications (CRC Press/Taylor & Francis Group, Boca
Raton, 2009)
5. V.K. Jain, Effect of background noise in space optical communication systems. Int. J. Electron.
Commun. (AEÜ) 47(2), 98–107 (1993)
6. S.G. Lambert, W.L. Casey, Laser Communication in Space (Artech House, Boston, 1995)
7. W.R. Leeb, Degradation of signal-to-noise in optical free space data link due to background
illumination. Appl. Opt. 28, 3443–3449 (1989)
8. T. Ohtsuki, I. Sasase, S. Mori, Lower bounds on capacity and cutoff rate of differential
overlapping pulse position modulation in optical direct-detection channel. IEICE Trans.
Commun. E77-B, 1230–1237 (1994)
9. Practical uses and applications of electro-optic modulators. NEW FOCUS, Inc, Appli-
cation Note 2. [Weblink: https://www.newport.com/n/practical-uses-and-applications-of-
electro-optic-modulators]
10. R. Ramaswani, K.N. Sivarajan, Optical Networks: A Practical Perspective (Morgan
Kaufmann, San Francisco, 2002)
11. D. Shiu, J.M. Kahn, Differential pulse-position modulation for power efficient optical
communication. IEEE Trans. Commun. 47(8), 1201–1210 (1999)
12. P.J. Winzer, A. Kalmar, W.R. Leeb, Role of amplifed spontaneous emission in optical free-
space communication links with optical amplifcation-impact on isolation and data transmission
and utilization for pointing, acquisition, and tracking. Proc. SPIE Free-Space Laser Commun.
Technol. XI 3615, 134 (1999)

View publication stats