International Journal of Engineering & Technology, 7 (3.

13) (2018) 176-182

International Journal of Engineering & Technology

Website: www.sciencepubco.com/index.php/IJET

Research paper

Nonlinear effects in WDM Networks

Kenza GAIZI1 2*, Pr. Dr. Fouad Mohammed ABBOU 1, Pr. Dr. Farid ABDI2
1
School of Science and Engineering, Al Akhawayn University Ifrane, Morocco
2
Laboratoire de Signaux, Systèmes et Composants, Université Sidi Mohammed Ben Abdellah, Faculté des Sciences et Techniques,
Fès, Morocco
*Corresponding author E-mail: [email protected]

Abstract

This paper studies nonlinear effects in WDM Network. The focus is on both the propagation of a unique signal as well as a set of them in
optical fibers in a WDM network. The paper presents an analytical model to study the effects of having nonlinearities in a WDM system.
Three main nonlinear effects are studied here are Cross Phase Modulation, Self-Phase Modulation, and Four Wave Mixing. Simulations
are performed on up to 265 channels WDM System.

Keywords: WDM network; nonlinear effects; Four Wave Mixing; Cross Phase Modulation.

let‟s assume that the refractive index increases with the intensity
of light. This means that the whole pulse is actually not going to
1. Introduction see the same propagation parameters. That is to say, at the centre
of the pulse when the light is intense, the refractive index seen by
Before diving into the subject, it is important to distinguish be- the pulse is different than on the edge of the pulse when the light
tween two types of Fiber Optics: (1) Linear and (2) Nonlinear. The is less intense. Because of that, the refractive index on the edge is
Linearity of the first category comes from the fact that the medium less than the refractive index seen by the pulse in its centre. If we
properties of the fiber do not depend on the signal. On the other try to send a continuous signal in the optical fibre, any small per-
hand, when the material properties of the fiber get affected by the turbation will break the signal into pulses. This is phenomenon is
signal itself we would then be in the second category of Nonlinear referred to, in the fields of nonlinear optics and fluid dynamics, as
Fiber Optics. In this paper, we will be discussing the propagation modulation instability. What happens here is that the deviations
of signal in the fiber taking into account nonlinear effects. We from a periodic waveform are reinforced by nonlinearities, gener-
would need to answer the following question: If the medium prop- ating by that spectral sidebands while also causing a breakup of
erties change in a fiber due to nonlinear effects, what are the im- the waveform into a set of pulses (wave trains) [6], [7], [8]. Given
plications on the signal? this, it‟s important to remember that without taking into considera-
tion the pulse nature of the signal, the nonlinear propagation will
2. Signal Propagation within Optical Fibre not be very effective.
In this paper, I will present a simple formulation and we‟ll answer
2.1. Propagation of One Signal within the Optical Fibre the question why is nonlinear fibre optics important. We‟re also
going to investigate if it is important to include the effects of non-
The 80s was marked as an era during which many researchers linearity into fibre optics communications or not. If the answer is
were invested in experimental work to establish the equivalence of yes, then, we‟re going to study if we can we make use of these
light and electromagnetic propagation [1]. Marking the beginning nonlinear effects to improve the signal propagation from the opti-
of radio telecommunications, in 1887, Heinrich Hertz was able to cal communication point of view?
generate and detect radio waves in the laboratory [2]. This came
2.1.1. Analytical Model
validating Maxwell‟s electromagnetic theory [3], according to
which, electromagnetic waves could be emitted from an oscillat- We‟re going to begin with taking a very basic model. When the
ing electric dipole, propagating via space at the speed of light; electric field is imposed on a dielectric material, the result is a tiny,
while exhibiting wave-like characteristics of light propagation [1], compared to the scale of atom dimensions, displacement of both
[2]. Let‟s take an example of a pulse that we need to transmit on positive and negative charge. Every molecule of the material is
the optical fibre in the linear domain. The properties of the propa- characterized by an induced dipole moment and the material is
gation of such pulse are only decided by the medium properties [4] „polarized‟ [9]. The induced polarization in the material given by
[5]. We could simply then solve the wave equation without worry- susceptibility of the medium. Normally, we consider only the first
ing about the amplitude of the pulse, and we can get the normal order susceptibility due to the dielectric material. However, that‟s
distribution, propagation constant and find out about the propaga- just an approximation. In real life, when light intensity is high, the
tion characteristics of a path. Now imagine that the medium prop- first order term is simply not adequate and one needs to consider
erties are modified by the pulse itself. Let‟s say that the refractive higher order terms into the polarization of the material to account
index of the optical fibre is related to the intensity of light. Now, for nonlinear effects. We consider a medium nonlinear, if there‟s a

Copyright © 2018 Authors. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted
use, distribution, and reproduction in any medium, provided the original work is properly cited.
International Journal of Engineering & Technology 177

nonlinear relation between the induced polarization P and the over which the beam focus is related to the size of the spot. So, in
applied electric field E [10]. In general, the induced polarization in a bulk medium, when we try to tightly focus the optical beam, the
the material is P, given as: interaction length gets affected as well [12]. The Interaction length
in bulk optics is:

χ (1) is the first order susceptibility, χ (2) is the second order suscep-
tibility, χ (3) is the third order susceptibility, and so on. E is the
imposed electric field [11]. The dominant term from the Induced Let‟s now compute the figure of merit:
Polarization equation is χ (1) and it contributes to dielectric con-
stant, the second term χ (2) is small for glass (SiO2), and the third
term χ (3), also known as the third order susceptibility says that the
refractive index is proportional to the square of the electrical field,
This shows the figure of merit, or the interaction efficiency is
and that‟s the one that‟ll contribute in glass to the nonlinear effects.
independent of the size of the focusing beam. So, in bulk material,
Notice here that when the intensity is very low, the second and
if we want to enhance the effect of nonlinearity, and want to do a
third terms are negligible [12].
certain wavelength, the only option we have is to increase the
optical power. So for research focusing on bulk material nonline-
arities, high power is required so that nonlinearities can start get-
ting induced effectively and allow researchers to study their ef-
fects.
n2 is the nonlinearity coefficient (a.k.a Kerr nonlinearity) [13]. It is
material dependent and also dependent on the third order suscepti- 2.1.3. Nonlinearities in Bulk Material
bility of the medium as shown below:
Now, let‟s compare the situation with Optical Fibre. In Optical
fibre, when the light gets in, it remains focused, or confined to a
region which is the size of the core. Since the loss on the optical
fibre is very small, the intensity reduces but very slowly, allowing
it to remain almost constant over some tens of kilometres. In the
Comparing the nonlinear coefficient of glass to other mate- fibre case, let‟s define the radius of the optical fibre core as a [12].
rials we can see that the glass‟s coefficient is 2 orders of Hence, the light intensity I would be defined as:
magnitude smaller. Now if this is the case for glass com-
pared to other bulk materials, should we really care about
the effects of nonlinearities in optical networks at all? To
properly answer this question, we need to understand how
do nonlinearities manifest in optical fibres compared to the When the light propagates through the optical fibre, it exponential-
ly decreases. Let‟s define the power attenuation constant as α and
bulk materials and if there‟s situational difference. the power function as per [12] [14] below:
2.1.2. Nonlinearities in Bulk Material

Let‟s start with the bulk material. When the light beam enters the
bulk material, the nonlinear interaction will be in the light and the
matter of that bulk material. On the other hand, when the light
enters the fibre, it travels a long distance during which it keeps Now, if we take a distance we will have:
interacting with the fibre, resulting in a cumulative nonlinear ef-
fect. To compare between the bulk and the fibre material, one can
check how much enhancement of nonlinear interaction takes place
in fibre material compared to a bulk material. Let‟s now define a
Figure of merit or efficiency parameter for nonlinear interaction, We can then say, in the case of optical fibre, that Leff ≈ 1/ α. Now,
say where: let‟s now compute the figure of merit for the optical fibre:

I is the intensity of light, which is the power intensity per meter

area (w/m2), Leff is the interaction length in (m) [12]. Equation (4)
shows that increasing efficiency of nonlinear interaction can be 2.1.4. Nonlinearities in Bulk Material and Optical Fiber
achieved by either increasing the intensity of light, the interaction
length, or both if that‟s possible. We can find how much interaction efficiencies increase in the
Let‟s say, we would like to increase the light intensity in some optical fibre compared to bulk material, we can find out by taking
bulk material. To do that one would need to focus the light beam the ratio of ηbulk and ηfiber as shown below:
using a lens. So, by focusing of the light, the spot size becomes
smaller and smaller and then the light intensity can be enhanced. If
we have a Gaussian Light Beam, with a spot radius of ω0 radius,
or a diameter of 2ω0. We can represent the Intensity of Light I as
the optical beam power divided by the area of the spot as per the
equation: Let‟s put some parameters for a typical optical fibre:

Where P is the light beam power and ω0 is the light spot radius in
the bulk material from this equation, we can see that the length
178 International Journal of Engineering & Technology

The quantity: {1 + χ (1)} is the dielectric constant of the medium.

In general such quantity is complex especially if there are losses in
the medium.

So for the same parameters, the nonlinear effects inside optical

fibres is enhanced by 1 billion times than those inside of the bulk
material. This also means that whatever affects you can see in bulk
material with 1 Watt of power, the same effect inside optical fibre
can be seen with only 1 nano Watt of power. This means that with
the power which we deal with in Optical Fibre in the order of
microWatt or milliWatt, the nonlinear effects will certainly be
present because once the power will be confined in optical fibre, it
will keep interacting with the fibre over long kilometres.
Comparing the Light intensity required to get nonlinear effects in
Fibre v.s. in bulk material. We see that in Fibre, such Intensity is
at least two orders of magnitude smaller. Now, in fibre optic
communications, we aren‟t working with single fibres, but rather
with some complex systems of multi fibre optics such as WDM
and DWDM. Assuming we have an intensity of 1 microWatt per
meter square induced per fibre. In a 100 channels WDM System
that‟ll already be 100 microWatt/ meter square. WDM systems are
using currently way more channels that this example, and the
power intensity inside optical fibre is certainly going to be large
enough to really worry about the nonlinear effects in these system
while giving us great motivation to study them and analyse ways
to make the best use out of them to rather make improvements to
the system.

2.1.5. Light Propagation in Optical Fiber with Nonlinearities

To analyse the propagation of light in the optical fibre in the pres-

ence of nonlinearity, we‟ll start from the very basic Maxwell‟s
Equations [3] below:

That is Fourier Transform, spectrum of the electric field as a func-

tion of distance:

This is assuming that permeability is not a function of time. Also

since we‟re dealing with dielectric material, we can assume that
the permeability is the same as that of free space, hence µ0.

Assuming the conduction J of the medium is 0

2.1.6. Summary

To summarize, when the light is intense, the induced polarization

has high order susceptibility terms also and they cannot be ne-
glected [15]. For a material like glass, the second order suscepti-
bility contribution is negligible because SiO2 molecule is symmet-
ric, so third order susceptibility term is what contributes to the
nonlinear effects. Even if the glass is not a very good nonlinear
material because its nonlinear coefficient is two orders of magni-
tude smaller compared to many well-known nonlinear materials,
since the interaction length inside the optical fibre is very large,
International Journal of Engineering & Technology 179

the nonlinear effects are very pronounced inside the optical fibre. power in these channels, there will be change in the phase of each
Compared to the bulk material, the interaction efficiency in optical signal due to the existence of the other signals. Also note here, if n
fibre is a billion time more so this is a strong case to investigate channels are transmitting, then, each channel is going to be affect-
the nonlinear effects inside the optical fibre. We then started with ed by the power present and remaining in 1-n channels. The col-
the Maxwell‟s equations [3] and wrote down the wave equation in lective effect of Cross Phase Modulation is way stronger than
terms of the polarization which includes the nonlinear terms. what one will see from self-phase modulation.

2.2. Propagation of Multiple Signals within the Optical Now in order to formulate the problem, we can write the field
Fibre distribution for the j channel as:

In the previous section, we‟ve studied what happens inside of the

optical fibre if there‟s one signal propagating. In the section below,
we‟ll see what happens inside the optical fibre with multiple sig- Now, if we follow the same steps as were done for the Self Phase
nals propagating through it. So, this section will then treat two Modulation, we‟ll end up with two equations: One for the field
phenomenon: Cross Phase Modulation and Four Wave Mixing. distribution: Fj (x, y) and one for the evolution of the envelope: A j
Both phenomena are related to the term which is due to third order (z)e −jβ0jz is a function of distance on the optical fibre.
susceptibility. Taking the nonlinear variation of the Schrodinger equation, we‟re
talking about the Nonlinear Schrodinger (NLS) equation, an equa-
2.2.1. Cross Phase Modulation tion belonging to classical field theory [16]. This equation has,
among others, applications related to the propagation of light in
Let‟s start here with some equations. Let‟s see we have the electric
nonlinear optical fibres [17]. The NLS equation is given as:
field which has two frequency components. The first components
having an amplitude E1 and a frequency 1 and the second has the
amplitude E2 and the frequency 2 and we write the equation:

If we substitute this into the polarization expression, we will get

the nonlinear polarization PNL created inside the material. This
polarization multiple nonlinear components: One at ω1, another
one at ω2 and other components in the same frequency range as
the previous two.

With j being the frequency and n2 the nonlinear coefficient. Note

from the delta of the refractive index above, that the nonlinear
effects of another signal are twice as higher than those of the sig-
nal itself. When the term |E3−j |2 = 0, the refractive index change
will be only dependent on the signal itself as the signal phase in
this case gets modified by itself. This change in the refractive
index gives a phenomenon that‟s called the Self Phase Modulation:

When that term |E3−j |2 ≠ 0, this means that the change in the re- In fibre where dispersion is large, there will be sign difference
fractive index is not only dependent on the signal itself but due between the velocity of the two signals, and then the walk off time
also to other signal that‟s propagating. The phase then is going to will be very short, so the nonlinear interaction between these two
be changed by both the signal as well as the other coexistent signal. pulses will be very small. Now if one uses the Dispersion Flat-
This phenomena is called the Cross Phase Modulation (XPM): tened Fibre (DFF), which is a fibre characterized by a smaller
mode-field diameter which concentrates optical power in a smaller
volume where the increased power density in fibre can cause non-
linear effects [18]. The velocity of the signals going through the
DFF will be almost similar, the walk off time will be very long,
In WDM Systems, there will be multiple channel transmissions and the nonlinear interaction between these pulses will be quite
within the optical fibre and XFM. We know that in WDM Sys- significant in a way that will increase the effect of Cross Phase
tems, if large number of channels propagate, if there‟s sufficient Modulation.
180 International Journal of Engineering & Technology

To avoid Cross Phase Modulation effects in the WDM signal, it is ω1 + ω3 − ω2,

desirable to keep some dispersion into the system. In the WDM ω3 + ω2 − ω1,
system where number of channel is very high, say 128 channels 2ω1 − ω2,
WDM System every channel will have the effect of the remaining 2ω1 − ω3,
127 channels. Moreover, every channel, even if it just carries the 2ω2 − ω3,
order of 2-3 milliWatts, the total power of 128 channels will be 2ω2 − ω3,
easily in the order of 500 milliWatts, and then the nonlinear phase 2ω3 − ω1,
which will be created because of this will be significant. So, in a 2ω3 − ω2 (B)
WDM System one will see that the Cross Phase Modulation will Fig. 3. Equispaced 3 channels in WDM System
be much stronger than the Self Phase Modulation because each
channel will still carry a power that is very small say 2-3 milliWatt. From the previous sections, we have defined a formula P NL where:
At such power, the Self Phase Modulation will be very weak, but
collectively, the Cross Phase Modulation will be very strong. From the above, we see that section A‟s frequency quantities have
In a WDM System if we want channels not to get affected by the same range or lie within the same frequency band while sec-
neighbouring channels due the Cross Phase Modulation, we tion B‟s frequency quantities fall in a different band altogether.
should get the fibre that has dispersion. You might have read in For instance, ω4 is almost three times ω1. We will not be looking
the literature, that it‟s desirable to make the dispersion small [19] into those. The phenomenon that frequencies of section B will lie
[20], so that the pulse broadening do not take place. However, within the same frequency band is what we call Four Wave Mix-
now when we are talking about WDM channels, we find that mak- ing, which is also called the inter modulation. This phenomenon is
ing dispersion 0 is not a good option, as channels then will start exactly what you see within an amplifier. Let‟s consider a WDM
interacting and there will be Cross Talk between these channels. system where there are 3 channels and as we know in WDM, m
So, XPM affects the signals in different frequencies, it also gives wavelengths are equispaced.
the same phenomenon like the spectral broadening which is not This means by this nonlinear process, the two channels ω1 and ω2
symmetrical around the original frequency, you might get asym- are going to put power into ω3. That could be interpreted in terms
metric spectrum developed because of this phenomenon of Cross
Phase Modulation. The same phenomenon gives another phenom-
enon called Four Wave Mixing, which we‟ll discover in the fol-
lowing section.

2.2.2. Four Wave Mixing

Four Wave Mixing in the optical system is similar to inter channel

mixing or intermodulation products in the electrical system. When
an amplifier goes into saturation, if you put two frequencies inside
the amplifier, the third frequency is generated because of the non-
linearities. These two frequencies give the product which is sum
and difference of frequencies. The same phenomenon is seen in
optical fibres because of the third order susceptibility. If we con-
sider a nonlinear polarization with three signals, or three frequen- of Cross Talk between ω1 and ω3 and ω2 and ω3. There will be a
cies simultaneously put in the optical fibre: Cross Talk phenomenon between all WDM Channels. To avoid
this phenomenon, it looks like we shouldn‟t propagate channels
PNL = ε0 χ(3) ⋮ E1 E2 E3 (43) that are equispaced. Making dispersion 0 on optical fibre is not a
good option. We should rather be keeping some residual disper-
sion to reduce the nonlinear effects in a multichannel transmission.

As a side note, it‟s worth mentioning that the phenomenon of Four

Wave Mixing is not always considered undesirable. It is actually
quite beneficial for wavelength conversion. In paper [21], the
authors have shown that using a length of 1 mm in Semiconductor
Optical Amplifiers (SOAs) improves the efficiency by 20 dB with
respect to standard devices, which is 0.5 mm long.

2.2.3. Simulations
These three frequencies will produce a fourth frequency ω4 = ω1
±ω2 ±ω3 and they will be having the wave numbers:

k = k1 ± k2 ± k3 (A)

So from three frequencies: ω1, ω2, ω3 the following frequencies

Fig. 1. FWM in a 3 Channels WDM System

The graph in figure 3 represents the number of mixing product

terms (M) as a function of number of wavelength where:
will be generated:
ω1 + ω2 − ω3,
International Journal of Engineering & Technology 181

As you have seen from the previous section, for a total of 3 input Figures 6, and 7 represent the same principle but for cases of more
wavelengths, we‟ve obtained a total of 9 mixing product terms. subcarriers: 128, and 256 consecutively.

Fig. 4. FWM Product vs. Channel Count

Figure 4 represents how this product is changing as the number of Fig. 8. Partially degenerate FWM Products and self-contributed FWM
products in a 64 Channels WDM System
input wavelengths increases. As you can see on the graph, for an
input of 10 wavelengths, we obtain 450 mixing product terms. If
we triple the number of input wavelengths to 30, this results in If we look at the next graph in Figure 8, we can see a representa-
13050 total mixing products. That is 2900 times higher. This is tion of the partially degenerate FWM products in a 64 Channels
almost a turning point in the graph. Post that, we see that the WDM System. Before explaining what‟s on the graph, a quick
FWM product continue to increase very rapidly reaching recall of the theory is necessary. Four Wave Mixing, is known as a
1,090,050 terms for only 130 input wavelengths. 3rd order nonlinear process where 3 frequencies (ωpump−1, ωpump−2,
and ωsignal) mix in a nonlinear medium, resulting in 2 extra fre-
quencies (a lower frequency ωStokes, and a higher frequency
ωantiStokes) [22] [23]. When ωpump−1 = ωpump−2 = ωpump, the
process is referred to as Partially Degenerate Four Wave Mixing
(PDFWM). We can see from the graph that the degenerate prod-
ucts number is constant within the subcarrier band, meaning that
these subcarriers are affected in an equal fashion. The same prin-
ciple applies when we studied WDM Systems of 128 and 256
Channels, as illustrated in Figure 9, and 10.

Fig. 5. FWM product as a function of Target Index Channel in 64 Chan-

nels WDM System

The next graph in Figure 5 depicts Four Wave Mixing as a func-

tion of the Target Index Channel in 64 Channels WDM System. In
the case considered with 64 subcarriers, the 33th target index
(subcarrier) plays a role in around 1485 FWM combinations while
the 1st and the 64th subcarriers have contribution to only approx-
imately 1000 FWM combinations.
Fig. 9. Partially degenerate FWM Products and self-contributed FWM
products in a 128 Channels WDM System

Fig. 6. FWM product as a function of Target Index Channel in 128 Chan-

nels WDM System
Fig. 10. Partially degenerate FWM Products and self-contributed FWM
products in a 256 Channels WDM System

3. Conclusion
In summary, we‟ve begun analysing nonlinear effects in bulk ma-
terial vs. fibre and found that we‟ve seen that high power is re-
quired so that nonlinearities can start getting induced effectively
and allow researchers to study their effects. Specifically, you can
see in bulk material with 1 Watt of power, the same effect inside
optical fibre can be seen with only 1 nano Watt of power. We then
Fig. 7. FWM product as a function of Target Index Channel in 256 Chan-
moved to study the propagation of light within the optical fibre
nels WDM System
182 International Journal of Engineering & Technology

and saw how very pronounced inside the optical fibre they are, as Boyd/dp/0123694701?SubscriptionId=0JYN1NVW651KCA56C10
a result of the very large interaction length inside the optical fibre. 2& tag=techkie-
We‟ve also seen how fibre dispersion has little effect on XPM, 20&linkCode=xm2&camp=2025&creative=165953& crea-
tiveASIN=0123694701
while a big negative impact on FWM, while more of it results in
[16] S. A. Rashkovskiy. (2018, feb) Nonlinear schrodinger equation and
rather less SPM. The next sections highlighted the analytical mod- ¨ classical-field description of the lamb retherford experiment.
el of these nonlinear effects as well as their simulations following [Online]. Available: https://arxiv.org/pdf/1802.01979.pdf
the same pattern as we‟ve kept testing with increasing number of [17] D. Felice, “A study of a nonlinear schrodinger equation for optical ¨
channels within the WDM System. fibers,” Ph.D. dissertation, Facolta di Scienze Matematiche, Fisiche
e ` Naturali, 2016.
[18] F. C. F. O. F. S. Co.). (2016, nov) Optical fiber dispersion. [Online].
References Available:
https://www.fiberoptics4sale.com/blogs/archiveposts/95047942-
[1] M. Fox, Optical Properties of Solids, ser. Oxford Master Series in optical-fiber-dispersion
Physics. OUP Oxford, 2010. [Online]. Available: [19] J. D. Downie, J. Hurley, and X. Zhu, “Xpm and sbs nonlinear ef-
https://books.google.ie/books?id=5WkVDAAAQBAJ fects on mlse with varying uncompensated dispersion,” Opt. Ex-
[2] “Heinrich hertz‟s wireless experiment (1887),” 2017. [Online]. press, vol. 17, no. 24, pp. 22 240–22 245, Nov 2009. [Online].
Available: http://people.seas.harvard.edu/∼jones/cscie129/nu lec- Available: http://www.opticsexpress.org/abstract.cfm?URI=oe-17-
tures/ lecture6/hertz/Hertz exp.html 24-22240
[3] J. C. Maxwell, “A dynamical theory of the electromagnetic field,” [20] M. Takahashi, R. Sugizaki, J. Hiroishi, M. Tadakuma, Y. Taniguchi,
Philosophical Transactions of the Royal Society of London, vol. and T. Yagi, “Low-loss and low-dispersion-slope highly nonlinear
155, no. 0, pp. 459–512, jan 1865. [Online]. Available: fibers,” J. Lightwave Technol., vol. 23, no. 11, p. 3615, Nov 2005.
https://doi.org/10.1098/rstl.1865.0008 [Online]. Available: http://jlt.osa.org/abstract.cfm?URI=jlt-23-11-
[4] R. Ramaswami, K. Sivarajan, and G. Sasaki, Optical Networks: A 3615
Practical Perspective, 3rd Edition. Morgan Kaufmann, 2009. [21] A. D‟Ottavi, F. Girardin, L. Graziani, F. Martelli, P. Spano, A.
[Online]. Available: https://www.amazon.com/OpticalNetworks- Mecozzi, S. Scotti, R. Dall‟Ara, J. Eckner, and G. Guekos, “Four-
Practical-Perspective-3rd/dp/0123740924?SubscriptionId= wave mixing in semiconductor optical amplifiers: a practical tool
0JYN1NVW651KCA56C102&tag=techkie-20&linkCode=xm2& for wavelength conversion,” IEEE Journal of Selected Topics in
camp=2025&creative=165953&creativeASIN=0123740924 Quantum Electronics, vol. 3, no. 2, pp. 522–528, Apr 1997.
[5] R. Ramaswami and K. Sivarajan, Optical Networks: A Practical [22] T. N. H. D. V. Aravind P. Anthur, Regan T. Watts and L. P. Barry.
Perspective (Morgan Kaufmann Series in Networking). Morgan (2013, aug) A general phase noise relationship for four-wave mix-
Kaufmann, 2001. [Online]. Available: ing. [Online]. Available: https://arxiv.org/pdf/1308.0914.pdf
https://www.amazon.com/Optical-NetworksPractical-Perspective- [23] S. K. O. R. I. L. Y. N. I. Motoki Asano, Yuki Takeuchi and T.
Networking/dp/1558606556?SubscriptionId= Yamamoto. (2016, may) Stimulated brillouin scattering and bril-
0JYN1NVW651KCA56C102&tag=techkie-20&linkCode=xm2& louin-coupled four-wave-mixing in a silica microbottle resonator.
camp=2025&creative=165953&creativeASIN=1558606556 [Online]. Available: https://arxiv.org/pdf/1605.07287.pdf
[6] G. Agrawal, Nonlinear Fiber Optics, Second Edition (Optics and
Photonics). Academic Press, 1995. [Online]. Available:
https://www.amazon.com/NonlinearFiber-Optics-Second-
Photonics/dp/0120451425?SubscriptionId=
0JYN1NVW651KCA56C102&tag=techkie-20&linkCode=xm2&
camp=2025&creative=165953&creativeASIN=0120451425
[7] T. B. Benjamin and J. E. Feir, “The disintegration of wave trains on
deep water part 1. theory,” Journal of Fluid Mechanics, vol. 27, no.
03, p. 417, feb 1967. [Online]. Available: https:
//doi.org/10.1017/s002211206700045x
[8] T. B. Benjamin and K. Hasselmann, “Instability of periodic wa-
vetrains in nonlinear dispersive systems [and discussion],” Proceed-
ings of the Royal Society A: Mathematical, Physical and Engineer-
ing Sciences, vol. 299, no. 1456, pp. 59–76, jun 1967. [Online].
Available: https://doi.org/10.1098/rspa.1967.0123
[9] S. Obayya, Computational Photonics. Imprint unknown, 2011.
[Online]. Available:
https://www.amazon.com/ComputationalPhotonics-Salah-
Obayya/dp/1119957508?SubscriptionId=
0JYN1NVW651KCA56C102&tag=techkie-20&linkCode=xm2&
camp=2025&creative=165953&creativeASIN=1119957508
[10] L. R. D. Sam-Shajing Sun, Introduction to Organic Electronic and
Optoelectronic Materials and Devices (Optical Science and Engi-
neering). CRC Press, 2008. [Online]. Available:
https://www.amazon.com/Introduction-ElectronicOptoelectronic-
Materials-Engineering/dp/0849392845?SubscriptionId=
0JYN1NVW651KCA56C102&tag=techkie-20&linkCode=xm2&
camp=2025&creative=165953&creativeASIN=0849392845
[11] C. Li, Nonlinear Optics: Principles and Applications. Springer, aug
2016. [Online]. Available: https://www.xarg.org/ref/a/9811014876/
[12] P. R. Shevgaonkar. (2012, jan) Introduction to non-linear fiber op-
tics. [Online]. Available: https://goo.gl/2MMaqD
[13] S. Dhara, F. Araoka, M. Lee, K. V. Le, L. Guo, B. K. Sadashiva, K.
Song, K. Ishikawa, and H. Takezoe, “Kerr constant and third-order
nonlinear optic susceptibility measurements in a liquid crystal com-
posed of bent-shaped molecules,” Physical Review E, vol. 78, no. 5,
nov 2008. [Online]. Available:
https://doi.org/10.1103/physreve.78.050701
[14] D. B. Potter. (2010, aug) Module 3 - attenuation in optical fibers.
[Online]. Available: http://opti500.cian-erc.org/opti500/pdf/sm/
Module3%20Optical%20Attenuation.pdf
[15] R. W. Boyd. (2008) Nonlinear optics, third edition. [Online].
Available: https://www.amazon.com/Nonlinear-Optics-Third-
Robert-