IOSR Journal of Electronics and Communication Engineering (IOSR-JECE)
e-ISSN: 2278-2834,p- ISSN: 2278-8735.Volume 9, Issue 3, Ver. III (May - Jun. 2014), PP 44-51
www.iosrjournals.org
www.iosrjournals.org 44 | Page
Implementation of Spread-Spectrum Techniques in Optical
Comunication
Bharat Vashistha
1
, Nishant Panwar
2
, Rebala Neel Reddy
3
, A. Jabeena
4
1,2,3,4,
School of Electronics Engineering, VIT University, Vellore , Tamil
Abstract: Method for applying spread-spectrum techniques to optical communication is presented. The
interference suppression capability of spread-spectrum systems is shown to be enhanced by optical transform
domain processing.Effects of jammer in DSSS communication system are demonstrated in thispaper.Several
possible implementations of this system are suggested, and applications to fiber optics, laser radar, free space
optical communications, and other systems are discussed.
Keywords: additive white gaussian noise, orthogonal frequency division multiplexing, bit error rate
I. Introduction
SPREAD-SPECTRUM communication consists of transmitting a given signal by modulating the
informationwith a large bandwidth, coded waveform such as a pn sequence. The transmitted signal occupies a
bandwidth much larger than the information bandwidth. Such systems possess a number of special properties
which distinguish them from narrowband communication techniques. A primary advantage of such systems is
resistance to jamming and interference. A broad spectral bandwidth signal is more difficult to distinguish from
ambient noise, which adds to the security of the channel.Spread-spectrum techniques have not been utilized in
optical communication systems, despite their increasingpopularity and inherently wide bandwidths, due to a
lackof effective modulation and coding methods available atoptical frequencies. We propose several methods
for incorporating spread-spectrum techniques into optical communications. An architecture for optical spread-
spectrum encoding is developed which performs pn code modulation on the Fourier transform of the data
sequence. This eliminates the need for high-frequency (GHz) optical modulators, as required by time domain pn
code systems. We also present a correlation receiver design for the optical spread-spectrum system, which is
simpler than the corresponding time domain receiver
Several possible implementations of these designs are suggested, as well as various applications
including fiber optic communications, laser radar, optical code division multiple accessnetworks, and laser
range finding.
A typical direct sequence spread-spectrum communications system is shown in Fig. 1. An information sequence,
S ( t ) , is modulated by a pn code sequence, c( f ) . The
modulated signal is corrupted in the communicationschannel by interference, I ( t ) , and additive, almost
whiteGaussian noise, n ( t ) . The corrupted signal is recovered by a matched filter containing the code
sequence.The ability of a spread-spectrum system to resist jammingis determined by the processing gain, which
in turn is given by the ratio of transmission bandwidth to data bandwidth.
Large processing gains provide a high degree of jamming immunity. Since processing gain cannot be
increased indefinitely, it is desirable to supplement the jamming resistance. This has led to the use of transform
domain processing techniques. A transform domain receiver is shown in Fig. 2. The received signal is x ( r ) = S
( r ) c ( r ) , plus channel interference and noise. Filtering by the transfer function H(w)is performed by
multiplication followed by inverse transformation. This real-time frequency domain multiplication has been
demonstrated both theoretically and experimentally. An alternate receiver implementation replaces the matched
filter by multiplication with the complex conjugate of the signal spectra in the transform domain.
Transform domain-processing techniques effectivelysuppress narrow band jammers in a spread-
spectrum system. The jammer may be removed by the system illustrated in Fig. 3. Input consists of the code and
jammer on an RF carrier; a high power* narrowband jammer appears as an impulse in the transform domain. A
gating function removes the portion of the spectrum containing the jammer. The gate output is the pn code;
Implementation of Spread-Spectrum Techniques in Optical Comunication
www.iosrjournals.org 45 | Page
since the code has a large bandwidth, the notch filter has not seriously degraded the signal spectrum. Correlation
is performed by multiplication in the transform domain, followed by an inverse Fourier transform.
II. Transform Domain Processing In The Optical Domain
Optical signal processing techniques are suited to applications in transform domain processing. We
shall consider the Chirp Transform, which can be used to implement a real-time Fourier transform. The Chirp
Transform is illustrated by Fig. 4(a). An arbitrary signal f(t) is multiplied by a down-chirp, then passes through a
linear system whose impulse response is an up-chirp. The result is multiplied by a down-chirp; the output is the
Fourier transform of the input. This is also known as the multiply- convolve- multiply (MCM) algorithm. The
system of Fig. 4(b) performs the same operation by a dual process called the convolve-multiply-convolve
(CMC) algorithm. In both cases, the same result holds if the impulse responses are down-chirps and we multiply
by the up-chirps. If the input is F(w ) , the algorithms invert the Fourier transformto yield f( - t) .Either
algorithm can be used in optics to perform spatial Fourier transforms of optical signals. The MCM algorithm
may also be applied to optical pulse compression systems. This algorithm can be realized in the time domain,
which forms the basis for the use of spread-spectrumsystems at optical frequencies.
Implementation of Spread-Spectrum Techniques in Optical Comunication
www.iosrjournals.org 46 | Page
Optical Pulse Compression
Pulse compression via fiber-grating orfiber-prism compression is first proposed by Tomlinson a fiber
with positive dispersions used to broaden the pulse spectrum and generate a square intensity profile with very
linear chirp across the pulse. SPM generates new frequencies and broadens the pulse spectrum, and the GVD
linearizes the chirp and squares the pulse. The linear chirp can then be compensated by a dispersive element
with negative dispersion (such as a grating pair, prism pair, or a chirped fiber grating operating in reflection,
producing a nearly transform- limited compressed pulse. This technique was applied successfully to produce
some of the shortest optical pulses. It should be stressed that this method relies on nonlinear effects and spectral
broadening, which is in contrast to simple chirp compensation by devices such as chirped gratings. These chirp
compensators are linear systems (similar to the second section in our compressor) that do not generate new
bandwidth.
The basic configuration for optical pulse compression and shaping is shown in Fig. 5 after the
treatment of [1] and [2]. Note that Fig. 5 is an optical implementation of Fig. 4(a), where f(t) is the pulse to be
compressed. This pulse compression is achieved by inducing a chirp, or linear frequency sweep, on an optical
pulse and subsequently re-phasing the chirped frequency components. In other words, the phase of the chirped
signal spectrum is adjusted so that it corresponds to a narrow pulse (impulse) in the time domain. Although there
are several means of obtaining chirped optical pulses, a single mode optical fiber induces uniform frequency
modulation across the entire pulse profile. The frequency chirp is generated by self-phase modulation, which
arises from the interaction of the propagating light and the intensity dependent portion of the fiber's refractive
index .It is then necessary to re-phase the spectral components to compress the pulse in time. The system in Fig.
5 uses a diffraction grating pair as a dispersive delay time. This system can be modified to encode an optical
pulse in the frequency domain. A chirped optical pulse can be produced by the nonlinear process of self-phase
modulation (SPM). The fiber's refractive index is given by
n=n
0
+n
2
I(t) (1)
Where I(t) is the intensity profile of the light, and n2 is a positive material constant [3]. The propagation
constant is given by
k = ωn
0
/c + AI(t) (2)
where c is the speed of light and the constant A represents a collection of terms. The phase of the optical pulse
becomes
θ = ω
0
t -ωzn
0
/c -AI(t) (3)
where z is the propagation distance. The instantaneous frequency is thus proportional to the negative time deriv-
ative of the intensity profile,
ω
i
= dθ/dt = ω
0
– A.d[I(t)]/dt (4)
and the properties of the resulting chirp depend on the time-varying intensity. However, for many pulse shapes
of interest such as Gaussian, the time derivative of (4) leads to a non-uniform frequency chirp. Thus, the
resulting frequency chirp produced by SPM alone is not linear with time over the full intensity profile. The
degree of nonlinearity depends on the temporal profile of the incident optical pulse; various cases have been
treated in the literature. The linearity of the chirp can be improved by the effect of positive group velocity
dispersion (GVD) in the fiber. This effect is calculated by expanding
Implementation of Spread-Spectrum Techniques in Optical Comunication
www.iosrjournals.org 47 | Page
0
the propagation constant, k ( w) , about the center frequencywthe relevant term is
β =(
2
k/ω
2
)z = k
2
z (5)
The combined effect of SPM and G V D is called disper-sive self-phase modulation (DSPM);it produces an ap-
proximate square, linearly up-chirped pulse from a single frequency input of non-uniform intensity. Once again,
the required non-uniform intensity profile may be Gaussian, hyperbolic secant squared, or some other profile
whose time derivative [as given in(4)] is sufficiently pro-nounced.
Asmentioned before, the system of Fig. 5 implements the following relation:
(f(t)
) *
=F(ω)
(6)
whereF(ԝ ) is the Fourier transform of f(t),ԝ= 2πf is the angular frequency, and β is a chirping factor to be
determined.The chirp parameter for a grating pair is given by[14]
β = π(f)
2
.d/bT (7)
where d is the grating constant = 1 /number of lines per mm, b is the grating separation, and T is a constant
determined by the angle of incidence. Thus, the system of Fig. 5 realizes (6); the analogous functions for each
component are shown in Fig. 6. The time-bandwidth product of a grating pair delay line may also be expressed
in terms of the grating's physical properties as
f)() =mbT/d
*
(f/f)
2
(8)
wherem is the diffracted order. Note that the time-bandwidth product may be increased by using higher
diffracted orders; this could be realized with blazed diffraction gratings, designed to diffract most of the optical
power into higher orders. After one pass through the grating pair, the optical signal, g
1
( t ) , is given by
g
1
(t) = F(βt/π)
(9)
The first pass through the gratings has taken the Fourier transform of the original input signal, because the
grating
pair input had the opposite chirp as the impulse response,
Implementation of Spread-Spectrum Techniques in Optical Comunication
www.iosrjournals.org 48 | Page
h( t ) , of the grating pair. The signal of (6) now possesses a chirp of the same sign as h (t),if this signal reflects
from the mirror and transverses the grating pair a second time,no further Fourier transformation can occur. This
second pass through the gratings doubles the chirp factor; the final output of the system, g
2
(t), is
g
2
(t) = F(βt/π)
(10)
The system has performed pulse compression as an MCM system. After one pass through the gratings, the
Fourier spectra of the signal are spatially separated in a plane. By placing some form of transmission mask in
this
plane, it is possible to modulate the frequency components of the optical signal .If a transmission mask function
M(f)is placed in this plane as indicated by Fig.7,then the final output will be
g
2
(t) = F(βt/π)M(βt/π)
(11)
III. Spread Spectrum Techniques In Optical Communication
There are many possible designs for an optical spread spectrum communication system based on the
pulse compression architecture described previously. Fig. 7 illustrates how the chirped light may be modulated
with binary data; an optical signal with positive GVD is assumed, although negative GVD may also be realized
as discussed earlier. The appropriate pn code must now be impressed upon these data. One implementation
involves placing a transmission-type mask between the grating pair and the feedback mirror, as shown in Fig. 7.
Since the frequency components are spatially distributed in this plane, it is possible to perform both amplitude
and phase modulation on the optical signal. The optical signal passes through the transmission mask twice; the
second pass can only be neglected if a binary amplitude mask is used (consisting of either opaque or transparent
pixels). If more complicated masks are required, then the square root of the desired amplitude function must be
implemented on the mask. Any type of phase coding must account for the double phase delay incurred by a
second pass through the mask. The optical signal at the encoding plane is the Fourier transform of the input
pulse; thus, the encoding mask must be the Fourier transform of the desired code. The encoded optical pulse is
transmitted along a fiber optic link. An optical receiver and decoding scheme for this signal is shown in Fig. 10.
The optical signal is passedthrough a grating pair, which spatially separates the frequencycomponents without
affecting the encodedsignal.The grating pair is now separated by twice the distance provided at the encoder, to
account for the doubled chirp rate. This spatially dispersed signal is then passed through the complex conjugate
of the encoding mask, M*(f). Multiplication in the transform domain is equivalent to correlation in the time
domain. If the decoding mask matches the signal modulation, a correlation peak will be observed. Otherwise,
the output will resemble random noise, since the cross correlation of two different codes is near zero. The
optical signal passes through another grating pair to compress the spectra before detection. The correlation
receiver design is greatly simplified in the transform domain. Spread-spectrum coding techniques at optical
frequencies can be implemented in this way.Note that transmitting the encoded optical pulse along a fiber optic
cable will not necessarily induce any further frequency chirp. The encoded optical pulse will be of lower
intensity and will not exhibit a pronounced intensity variation with time, so the SPM described by (4) would not
be significant. If the transmission characteristics are such that further chirping does occur, then the chirp
parameter of the received signal will not match the chirp parameter of the receiver grating pair. Thus, when the
received signal is spread out by the grating pair, it will not be properly aligned with the transmission mask and
may not be correctly decoded. However, from (7), the chirp parameter of the receiver grating pair can be
adjusted by changing the grating spacing. In this manner, the receiver can be tuned to match the chirp parameter
of the received signal. This assumes that the additional chirp induced during transmission is known, or can be