Component Library
Optical Communication System Design Software
Version 11
OptiSystem
Component Library
Optical Communication System Design Software
All OptiSystem documents, including this one, and the information contained therein, is copyright material.
No part of this document may be reproduced, stored in a retrieval system, or transmitted in any form or by any means whatsoever,
including recording, photocopying, or faxing, without prior written approval of Optiwave.
Disclaimer
Optiwave makes no representation or warranty with respect to the adequacy of this documentation or the programs which it
describes for any particular purpose or with respect to its adequacy to produce any particular result. In no event shall Optiwave, its
employees, its contractors or the authors of this documentation, be liable for special, direct, indirect, or consequential damages,
losses, costs, charges, claims, demands, or claim for lost profits, fees, or expenses of any nature or kind.
Technical Support
If you purchased Optiwave software from a distributor that is not listed here, please send technical
questions to your distributor.
Optiwave Canada/US
Tel (613) 2244700 Email support@optiwave.com
Fax (613) 2244706 URL www.optiwave.com
OptiAmplifier.........................................................................................................1545
OptiGrating...........................................................................................................1553
WDM_Phasar Demux 1xN ...................................................................................1557
WDM_Phasar Mux Nx1........................................................................................1559
OptiBPM Component NxM...................................................................................1563
Save Transverse Mode ........................................................................................1567
OptiSPICE Output ................................................................................................1571
OptiSPICE NetList................................................................................................1573
Scilab Component................................................................................................1601
EDA Cosimulation Library ........................................................................ 1605
1
Notes:
2
DUOBINARY PULSE GENERATOR
Ports
Parameters
Simulation
3
DUOBINARY PULSE GENERATOR
Technical background
The equivalent subsystem is:
4
ELECTRICAL JITTER
Electrical Jitter
Ports
Parameters
Main
Simulation
5
ELECTRICAL JITTER
Technical background
The jitter is a shortterm, noncumulative variation of the significant instants of a digital
signal from their positions in time. Jitter amplitude is measured in unit intervals (UI),
where 1 UI is the phase deviation of one clock period. The peaktopeak UI deviation
of the phase function with respect to time is referred as jitter amplitude. The output
signal is:
A
out ( t ) = E in ⎛⎝ t + tr +  sin ( 2πft )⎞⎠
2B
where A is the deterministic jitter amplitude, B is the signal bit rate, and f is the jitter
frequency. And tr is the random jitter that has a Gaussian probability distribution with
zero mean and standard deviation defined by the parameter Random jitter amplitude
(rms value).
6
NOISE SOURCE
Noise Source
Ports
Parameters
Main
Simulation
7
NOISE SOURCE
Noise
Random numbers
Technical background
The average output Power or Power spectral density are parameters that you specify.
This model generates electrical sampled signals or electrical sampled noise
according to:
E out = [ x ( t ) + jy ( t ) ] P ⁄ 2
A Gaussian distribution describes the probability density function for the real and
imaginary part of E. P is the average power when PSD parameter is false, if PSD is
true then P is calculated from the power spectral density multiplied by the Sample
rate.
8
RZ PULSE GENERATOR
RZ Pulse Generator
Ports
Parameters
Main
Position 0 bit
Defined as the time from when the rising edge reaches 10% of the
amplitude to the time it reaches 90% of the amplitude
Defined as the time from when the falling edge reaches 90% of the
amplitude to the time it reaches 10% of the amplitude
9
RZ PULSE GENERATOR
Simulation
Technical background
According to the parameter Rectangle shape, this model can produce pulses with
different edge shapes:
Exponential
⎧ – ( t ⁄ cr )
⎪1 – e ,0 ≤ t < t 1
⎪
⎪ 1, t 1 ≤ t < t 2
E(t ) = ⎨
⎪ e – ( t ⁄ cf ) ,t ≤ t < t
⎪ 2 c
⎪ 0, t c ≤ t < T
⎩
Gaussian
⎧
2
–( t ⁄ cr )
⎪1 – e ,0 ≤ t < t 1
⎪
⎪ 1, t 1 ≤ t < t 2
E(t) = ⎨ 2
⎪ e – ( t ⁄ c f ) ,t ≤ t < t
⎪ 2 c
⎪ 0, t c ≤ t < T
⎩
10
RZ PULSE GENERATOR
Linear
⎧ t ⁄ c r ,0 ≤ t < t 1
⎪
⎪ 1, t 1 ≤ t < t 2
E(t) = ⎨
⎪ t ⁄ c f ,t 2 ≤ t < t c
⎪
⎩ 0, t c ≤ t < T
Sine
⎧ sin ( π.t ⁄ c r ) ,0 ≤ t < t 1
⎪
⎪ 1 ,t 1 ≤ t < t 2
E(t ) = ⎨
⎪ sin ( π.t ⁄ c f ) ,t 2 ≤ t < t c
⎪
⎩ 0 ,t c ≤ t < T
where cr is the rise time coefficient and cf is the fall time coefficient. t1 and t2, together
with cr and cf, are numerically determinate to generate pulses with the exact values
of the parameters Rise time and Fall time. tc is the duty cycle duration, and T is the
bit period.
11
RZ PULSE GENERATOR
Notes:
12
NRZ PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
Defined as the time from when the rising edge reaches 10% of the
amplitude to the time it reaches 90% of the amplitude
13
NRZ PULSE GENERATOR
Simulation
14
NRZ PULSE GENERATOR
Technical background
According to the parameter Rectangle shape, this model can produce pulses with
different edge shapes:
Exponential
⎧ – ( t ⁄ cr )
⎪1 – e ,0 ≤ t < t 1
⎪
E(t ) = ⎨ 1 ,t 1 ≤ t < t 2
⎪ –( t ⁄ c )
⎪ e f
,t 2 ≤ t < T
⎩
Gaussian
⎧ –( t ⁄ cr )2
⎪ e ,0 ≤ t < t 1
⎪
E(t) = ⎨ 1 ,t 1 ≤ t < t 2
⎪ 2
⎪ e –( t ⁄ c f ) ,( t ≤ t < T )
⎩ 2
15
NRZ PULSE GENERATOR
Linear
⎧ t ⁄ c r ,0 ≤ t < t 1
⎪
E ( t ) = ⎨ 1, t 1 ≤ t < t 2
⎪
⎩ t ⁄ c f ,t 2 ≤ t < T
Sine
where cr is the rise time coefficient and cf is the fall time coefficient. t1 and t2, together
with cr and cf, are numerically determined to generate pulses with the exact values of
the parameters Rise time and Fall time, and T is the bit period.
16
GAUSSIAN PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
Order 1 — [1,100]
Order of the function
17
GAUSSIAN PULSE GENERATOR
Simulation
Technical background
This model generates Gaussian or superGaussian electrical pulses according to the
bit sequence at the input. For each bit
2N
1 t.k
⎛ –  ⎛⎝ ⎞⎠ ⎞
2 T FWHM
⎜
E ( t ) = B. ⎜ A p .e + A bias⎟⎟
⎝ ⎠
where Ap is the parameter peaktopeak Amplitude, and Abias is the parameter Bias.
B is the bit value (1 or 0) and depends on the input bit sequence. k is the fitting
coefficient determined numerically to generate pulses with the exact values of the
parameter Width TFWHM, and N is the Order of the Gaussian (N=1) or superGaussian
pulses (N>1).
18
HYPERBOLICSECANT PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
19
HYPERBOLICSECANT PULSE GENERATOR
Simulation
Technical background
This model generates electrical pulses according to the bit sequence at the input. For
each bit:
t.k 2
E ( t ) = B. ⎛ A p ⁄ cosh ⎛ ⎞ + A bias⎞
⎝ ⎝ T FWHM⎠ ⎠
where Ap is the parameter peaktopeak Amplitude, and Abias is the parameter Bias.
B is the bit value (1 or 0) and depends on the input bit sequence. k is the fitting
coefficient determined numerically to generate pulses with the exact values of the
parameter Width, TFWHM.
20
SINE GENERATOR
Sine Generator
Ports
Parameters
Main
21
SINE GENERATOR
Simulation
22
TRIANGLE PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
23
TRIANGLE PULSE GENERATOR
Simulation
24
SAWUP PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
25
SAWUP PULSE GENERATOR
Simulation
26
SAWDOWN PULSE GENERATOR
Ports
Parameters
Main
27
SAWDOWN PULSE GENERATOR
Simulation
28
IMPULSE GENERATOR
Impulse Generator
Ports
Parameters
Main
Simulation
29
IMPULSE GENERATOR
Notes:
30
RAISED COSINE PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
31
RAISED COSINE PULSE GENERATOR
Simulation
Technical background
This model generates electrical pulses according to the bit sequence at the input. For
each bit:
t.k 2
E ( t ) = B. ⎛ A p . cos ⎛ ⎞ + A bias⎞
⎝ ⎝ T FWHM⎠ ⎠
where Ap is the parameter peaktopeak Amplitude, and Abias is the parameter Bias.
B is the bit value (1 or 0) and depends on the input bit sequence. k is the fitting
coefficient determined numerically to generate pulses with the exact values of the
parameter Width, TFWHM.
32
SINE PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
33
SINE PULSE GENERATOR
Simulation
Technical background
This model generates electrical pulses according to the bit sequence at the input. For
each bit:
t.k
E ( t ) = B. ⎛ A p . cos ⎛ ⎞ + A bias⎞
⎝ ⎝ T FWHM⎠ ⎠
where Ap is the parameter peaktopeak Amplitude, and Abias is the parameter Bias.
B is the bit value (1 or 0) and depends on the input bit sequence. k is the fitting
coefficient numerically determinate to generate pulses with the exact values of the
parameter Width TFWHM.
34
MEASURED PULSE
Measured Pulse
Ports
Parameters
Main
Position 0 bit
Filename Pulse.dat — —
Numerical
35
MEASURED PULSE
Simulation
Graphs
Technical background
This model generates electrical signal loading measurements from a file. The input
file is formatted containing two values per line, the time in seconds and signal
amplitude in arbitrary units. The time scale is normalized to fit in one bit period  the
duration of the pulse. For example, the file representing one measurement has the
following form:
0 0
1e6 0.5
2e6 0.5
3e6 0
...
36
MEASURED PULSE SEQUENCE
Ports
Parameters
Main
Filename Sequence.dat — —
Numerical
37
MEASURED PULSE SEQUENCE
Simulation
Graphs
Technical background
This model generates electrical signal loading measurements from a file. The input
file is formatted containing two values per line, the time in seconds and signal
amplitude in arbitrary units. For example, the file representing one measurement has
the following form:
0 0
1e6 0.5
2e6 0.5
3e6 0
...
38
BIAS GENERATOR
Bias Generator
A d.c. source.
Ports
Parameters
Main
Simulation
39
BIAS GENERATOR
Notes:
40
MARY PULSE GENERATOR
Ports
Parameters
Main
Position 0 bit
Simulation
41
MARY PULSE GENERATOR
Technical background
This model generates pulses according to:
⎧
⎪ b, 0 ≤ t < t 1
⎪
v out ( t ) = ⎨ av in ( t ) + b, t 1 ≤ t < t 1 + t c
⎪
⎪ b, t 1 + t c ≤ t < T
⎩
where v in is the input Mary signal, a is the linear gain, and b is the parameter Bias.
T is the bit period, t c is the duty cycle, and t 1 is the pulse position.
42
MARY RAISED COSINE PULSE GENERATOR
Generates multilevel raised cosine pulses according to the Mary signal input.
Ports
Parameters
Main
Position 0 bit
43
MARY RAISED COSINE PULSE GENERATOR
Simulation
Technical background
This model generates pulses according to:
⎧
⎪ b, 0 ≤ t < t 1
⎪
v out ( t ) = ⎨ ah ( t ) + b, t 1 ≤ t < t 1 + w
⎪
⎪ b, t 1 + w ≤ t < T
⎩
where v in is the input Mary signal, a is the linear gain, and b is the parameter Bias.
T is the bit period, w is the pulse width, and t 1 is the pulse position. h is given by:
πt
sin ⎛  ( 1 + α )⎞
πt ⎝ T ⎠
cos ⎛  ( 1 + α )⎞ + 
⎝T ⎠ 4αt

T
h ( t ) = 4α 
2

4αt
π T ⎛ 1 – ⎛⎝ ⎞⎠ ⎞
⎝ T ⎠
44
PREDISTORTION
Predistortion
Ports
Parameters
Main
Gain 1 ]INF,+INF[
Simulation
45
PREDISTORTION
Technical background
If parameter Predistortion is Arcsin, the function applied to the input signal is:
1
v out ( t ) =  arc sin ( v in ( t ) ) ⋅ a + b
π
where v in is the input signal, a is the linear gain, and b is the bias.
If parameter Predistortion is Polynomial, the function applied to the input signal is:
2 N
v out ( t ) = ( c 0 + c 1 v in ( t ) + c 2 v ( t ) in + … + c N v ( t ) in ) ⋅ a + b
46
PREDISTORTION
Transmitters Library
Pulse Generators
Optical
47
PREDISTORTION
Notes:
48
OPTICAL GAUSSIAN PULSE GENERATOR
Ports
Parameters
Main
Emission frequency
Position 0 bit —
Order 1 — — [1,100]
49
OPTICAL GAUSSIAN PULSE GENERATOR
Chirp
Polarization
Simulation
Parameterized Parameterized — —
50
OPTICAL GAUSSIAN PULSE GENERATOR
Technical background
This model generates Gaussian or superGaussian optical pulses according to the bit
sequence at the input. For each bit, the output optical power is:
2N
1 t.k
⎛ –  ⎛ ⎞ ⎞
2 ⎝ T FWHM⎠
⎜
P ( t ) = B. A p .e + A bias⎟
⎜ ⎟
⎝ ⎠
where Ap is the parameter peaktopeak Power, and Abias is the parameter Bias. B is
the bit value (1 or 0) and depends on the input bit sequence. k is the fitting coefficient
determined numerically to generate pulses with the exact values of the parameter
Width, TFWHM, and N is Order of the Gaussian (N=1) or superGaussian pulses (N>1).
dϕ α d
 = e 
 ln P ( t ) + κP ( t )
dt 2 dt
where ϕ is the signal phase, αe is the parameter Linewidth enhancement factor, and
κ is the parameter Adiabatic chirp.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
where the power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
51
OPTICAL GAUSSIAN PULSE GENERATOR
Notes:
52
OPTICAL SECH PULSE GENERATOR
Ports
Parameters
Main
Emission frequency
Position 0 bit —
53
OPTICAL SECH PULSE GENERATOR
Chirp
Polarization
Simulation
Parameterized Parameterized — —
54
OPTICAL SECH PULSE GENERATOR
Technical background
This model generates optical pulses according to the bit sequence at the input. For
each bit, the output optical power is:
t.k
P ( t ) = B. ⎛ A p ⁄ cosh ⎛ ⎞ + A bias⎞
⎝ ⎝ T FWHM⎠ ⎠
where Ap is the parameter peaktopeak Power, and Abias is the parameter Bias. B is
the bit value (1 or 0) and depends on the input bit sequence. k is the fitting coefficient
determined numerically to generate pulses with the exact values of the parameter
Width, TFWHM.
dϕ α d
 = e 
 ln P ( t ) + κP ( t )
dt 2 dt
where ϕ is the signal phase, αe is the parameter Linewidth enhancement factor, and
κ is the parameter Adiabatic chirp.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
55
OPTICAL SECH PULSE GENERATOR
Notes:
56
OPTICAL IMPULSE GENERATOR
Ports
Parameters
Main
Emission frequency
Chirp
57
OPTICAL IMPULSE GENERATOR
Polarization
Simulation
Parameterized Parameterized — —
58
OPTICAL IMPULSE GENERATOR
Technical background
This model generates optical pulses according to the bit sequence at the input. For
each bit, the output optical power is:
P ( t ) = B. ( A p δ ( t – t p ) + A bias )
where Ap is the parameter peaktopeak Power, and Abias is the parameter Bias. B is
the bit value (1 or 0) and depends on the input bit sequence. δ is the impulse function
and tP is the parameter Pulse position.
dϕ α d
 = e 
 ln P ( t ) + κP ( t )
dt 2 dt
where ϕ is the signal phase, αe is the parameter Linewidth enhancement factor, and
κ is the parameter Adiabatic chirp.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
59
OPTICAL IMPULSE GENERATOR
Notes:
60
MEASURED OPTICAL PULSE
Ports
Parameters
Main
Emission frequency
Position 0 bit —
Filename Optical — — —
pulse.dat
Filename with the measured data
61
MEASURED OPTICAL PULSE
Polarization
Numerical
Simulation
Parameterized Parameterized — —
Graphs
62
MEASURED OPTICAL PULSE
Technical background
The input file is formatted containing two items per line — the time in seconds and the
signal measurement (Power in watts, Phase in radians, Real and Imag in Volts). The
time scale is normalized to fit in one bit period  the duration of the pulse. According
to the parameter File format, the second item can be one value (Power or Phase), or
two values (Power and Phase or Real and Imag).
0 0
1e6 0.5
2e6 0.5
3e6 0
...
Power Phase
0 0 0
3e6 0 0
...
Real Imag
0 0 0
3e6 0 0
...
63
MEASURED OPTICAL PULSE
0 0
1e6 3.14
2e6 3.14
3e6 0
...
This model generates optical pulses according to the bit sequence at the input. For
each bit, the output optical power is:
P ( t ) = B. ( A p M ( t ) + A bias )
where Ap is the parameter peaktopeak Power, and Abias is the parameter Bias. B is
the bit value (1 or 0) and depends on the input bit sequence. M is the measured data.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
64
MEASURED OPTICAL PULSE SEQUENCE
Ports
Parameters
Main
Emission frequency
Filename Sequence.dat — — —
Polarization
65
MEASURED OPTICAL PULSE SEQUENCE
Numerical
Simulation
Graphs
66
MEASURED OPTICAL PULSE SEQUENCE
Technical background
This model generates optical signal loading measurements from a file.
The input file is formatted containing two items per line — the time in seconds and
signal measurement (Power in watts, Phase in radians, Real and Imag in Volts).
According to the parameter File format, the second item can be one value (Power or
Phase) or two values (Power and Phase or Real and Imag).
0 0
1e6 0.5
2e6 0.5
3e6 0
...
Power Phase
0 0 0
3e6 0 0
...
Real Imag
0 0 0
3e6 0 0
...
67
MEASURED OPTICAL PULSE SEQUENCE
0 0
1e6 3.14
2e6 3.14
3e6 0
...
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
68
TIME RESOLVE CHIRP (TRC) MEASUREMENT DATA
This component is an interface between OptiSystem and time resolve chirp (TRC) [1]
measurement instruments, such as the OSA Agilent 86146B with TRC option.
Ports
Parameters
Main
Emission frequency
Filename Sequence.dat — — —
Polarization
69
TIME RESOLVE CHIRP (TRC) MEASUREMENT DATA
Numerical
Simulation
Graphs
Technical background
This component generates optical signal loading measurements from a file. These
measurements are TRC data that describe the power and chirp evolution of the
optical signal in time [1].
TRC provides frequency vs. time information about a modulated lightwave signal.
Also called dynamic chirp, the TRC graph provides useful information on the ability of
a modulated signal to propagate over long distances in optical fiber.
Using measurement equipment such as the Agilent 86146B, with the filter mode
capability, Agilent 86100 Infinium Digital Communications Analyzer (DCA) dedicated
software (86146B Option TRL), and a personal computer, the time resolved chirp
(TRC) of a modulated laser can be calculated.
From the measurement, a file with the TRC data is generated. OptiSystem can load
this file and the effect of laser chirp on a wide variety of system performance metrics
70
TIME RESOLVE CHIRP (TRC) MEASUREMENT DATA
The input file is formatted containing three items per line  the time in seconds, the
signal power is Watt (Linear scale) or dBm, and the signal chirp (Hz).
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as:
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
71
TIME RESOLVE CHIRP (TRC) MEASUREMENT DATA
References
[1] Agilent Technologies, “Making TimeResolved Chirp Measurements Using the Optical
Spectrum Analyzer and Digital Communications Analyzer”, Agilent Application Note 15507,
2002.
72
SPATIAL OPTICAL GAUSSIAN PULSE GENERATOR
This component is Gaussian pulse generator that includes transverse mode profiles
in the optical output. It is a subsystem built using the Optical Gaussian Pulse and the
Multimode Generators.
Ports
Parameters
Main
Emission frequency
73
SPATIAL OPTICAL GAUSSIAN PULSE GENERATOR
Chirp
Name and description Default value Default unit Units Value range
Chirp definition Linear [Linear,
Measured]
Polarization
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1
74
SPATIAL OPTICAL GAUSSIAN PULSE GENERATOR
Name and description Default value Default unit Units Value range
Pol. X inv. radius of curvature 0 1/um [0, 1e+100]
Simulation
Technical Background
The layout of the Spatial Optical Gaussian Pulse Generator is presented in Figure 1.
Refer to Optical Gaussian Pulse Generator and Multimode Generator component
documentation for the technical background of the models.
75
SPATIAL OPTICAL GAUSSIAN PULSE GENERATOR
76
SPATIAL OPTICAL SECH PULSE GENERATOR
This component is sech pulse generator that includes transverse mode profiles in the
optical output. It is a subsystem built using a the Optical Sech Pulse and the
Multimode Generators.
Ports
Parameters
Main
Emission frequency
77
SPATIAL OPTICAL SECH PULSE GENERATOR
Chirp
Name and description Default value Default unit Units Value range
Chirp definition Linear [Linear,
Measured]
Polarization
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1
78
SPATIAL OPTICAL SECH PULSE GENERATOR
Name and description Default value Default unit Units Value range
Pol. X inv. radius of curvature 0 1/um [0, 1e+100]
Simulation
Technical Background
The layout of the Spatial Optical Sech Pulse Generator is presented in Figure 1. Refer
to Optical Sech Pulse Generator and Multimode Generator component
documentation for the technical background of the models.
79
SPATIAL OPTICAL SECH PULSE GENERATOR
80
SPATIAL OPTICAL IMPULSE GENERATOR
This component is impulse generator that includes transverse mode profiles in the
optical output. It is a subsystem built using a the Impulse and the Multimode
Generators.
Ports
Parameters
Main
Emission frequency
Chirp
Name and description Default value Default unit Units Value range
Alpha parameter 0 rad/W [1000, 1000]
81
SPATIAL OPTICAL IMPULSE GENERATOR
Name and description Default value Default unit Units Value range
Adiabatic chirp 0 1/s [1000, 1000]
Polarization
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1
82
SPATIAL OPTICAL IMPULSE GENERATOR
Name and description Default value Default unit Units Value range
Pol. Y spot size 5 um [1e100, 1e+100]
Simulation
Technical Background
The layout of the Spatial Optical Impulse Generator is presented in Figure 1. Refer to
Optical Impulse Generator and Multimode Generator component documentation for
the technical background of the models.
83
SPATIAL OPTICAL IMPULSE GENERATOR
Notes:
84
SPATIAL OPTICAL IMPULSE GENERATOR
Transmitters Library
Optical Sources
• CW Laser
• Laser Rate Equations
• Laser Measured
• Fabry Perot Laser
• LED
• White Light Source
• Pump Laser
• Pump Laser Array
• Controlled Pump Laser
• CW Laser Array
• CW Laser Array ES
• CW Laser Measured
• Directly Modulated Laser Measured
• VCSEL Laser
• Spatiotemporal VCSEL
• Spatial CW Laser
• Spatial VCSEL
• Spatial Laser Rate Equations
• Spatial LED
85
SPATIAL OPTICAL IMPULSE GENERATOR
Notes:
86
CW LASER
CW Laser
Ports
Parameters
Main
Emission frequency
Polarization
87
CW LASER
Simulation
Noise
Random numbers
88
CW LASER
Technical background
In the CW case, the average output Power is a parameter that you specify. Laser
phase noise is modeled using the probability density function:
2
Δϕ
1 – 
4πΔfdt

f ( Δϕ ) =  ⋅ e
2π Δfdt
where Δϕ is the phase difference between two successive time instants and dt is the
time discretization. A Gaussian random variable for the phase difference between two
successive time instants with zero mean and a variance equal to 2 π Δ f has been
assumed, with Δf as the laser Linewidth.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
where the power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as follows:
2 k ( 1 – k ) cos ( θ )
tan ( 2α ) = 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
89
CW LASER
Notes:
90
LASER RATE EQUATIONS
Ports
Parameters
Main
91
LASER RATE EQUATIONS
Physical
92
LASER RATE EQUATIONS
Numerical
Graphs
To 40 mA [0, +INF]
Simulation
Parameterized Parameterized — —
93
LASER RATE EQUATIONS
Noise
Random numbers
Technical background
The modulation dynamics of the laser are modeled by coupled rate equations which
describe the relation between the carrier density N ( t ) , photon density S ( t ) , and
optical phase φ ( t ) :
dN ( t ) I(t) N(t ) 1
 =  –  – g o ⋅ ( N ( t ) – N t ) ⋅  ⋅ S ( t ) (2)
dt q⋅V τn (1 + ε ⋅ S(t))
dS ( t ) 1 S(t) Γ ⋅ β ⋅ N(t)
 = Γ ⋅ g o ⋅ ( N ( t ) – N t ) ⋅  ⋅ S ( t ) –  +  (3)
dt (1 + ε ⋅ S(t)) τp τn
dφ ( t ) 1 1
 =  ⋅ α ⋅ Γ ⋅ g o ⋅ ( N ( t ) – N t ) –  (4)
dt 2 τp
94
LASER RATE EQUATIONS
The optical power and chirp response of the semiconductor laser to a current
waveform I ( t ) is determined by the above equations. Parameters Bias current and
Modulation peak current are scale factors applied to the input electrical signal.
I ( t ) = I DC + I in ( t ) × I Pk (5)
Where I in ( t ) is the input signal current, I DC is the parameter Bias Current and
I Pk is the parameter Modulation peak current. If parameter Bias Current and
Modulation peak current have zero values, the internal current is given by I in ( t ) only.
The time variations for the optical and laser chirp are:
S ⋅ V ⋅ ηo ⋅ h ⋅ v (6)
P = 
2 ⋅ Γτ p
1 dφ
Δv =  ⋅  (7)
2 ⋅ π dt
95
LASER RATE EQUATIONS
The Laser Rate Equations supports individual samples for timedriven simulation.
References
[1] J. C. Cartledge and G. S. Burley, “The Effect of the Laser Chirping on Lightwave System
Performance”, J. Lightwave Technology, vol. 7, pp. 568573, March 1989.
[2] Agrawal GP, Dutta NK. Semiconductor lasers, 2nd ed. New York: Van Nostrand Reinhold, 1993.
96
LASER MEASURED
Laser Measured
Extracts values of the rate equation parameters using measurements and simulates
the modulation dynamics of a laser.
Ports
Parameters
Main
97
LASER MEASURED
Measurements
98
LASER MEASURED
Initial estimate
99
LASER MEASURED
Numerical
Graphs
To 40 mA [0, +INF]
Simulation
Parameterized Parameterized — —
100
LASER MEASURED
Noise
Random numbers
Technical background
The laser measured model extracts values of the rate equation parameters using
measurements of the threshold current, optical power, resonance frequency, and
damping factor to simulate a DFB laser.
Based on the results featured in [1], the values of the rate equation parameters are
calculated in a way that parameters simultaneously yield the measured values of Y
(damping factor), Z (resonance frequency factor), Ith (threshold current), and P
(Power bias). The parameter extraction procedure is based on minimization of the
sum of squared errors between the measured values of (Y, Z, Ith, P) and values
calculated from rate equation parameters. The minimization is over the values of the
rate equation parameters which are:
Damping factor
S 1 1 1
Y = g 0  +  – Γ ⋅ g 0 ( N – N t ) 2 + 
( 1 + ε ⋅ S ) τn (1 + ε ⋅ S) τp
S 1 g0 1 1
Z = g 0  ⋅  + ( β – 1 ) ⋅ Γ ⋅  ( N – N t ) 2 + 
( 1 + ε ⋅ S ) τp τn (1 + ε ⋅ S) τp ⋅ τn
Threshold current
q ⋅ V 1 + Nt ⋅ Γ ⋅ go ⋅ τp
I th =  ⋅ 
τn Γ ⋅ go ⋅ τp
101
LASER MEASURED
Power bias
S ⋅ V ⋅ η0 ⋅ h ⋅ v
P = 
2 ⋅ Γτ p
NandS are the steadystate values of the carrier and photon densities
corresponding to the bias current of the laser
where ( Y mea, Z mea ,P mea ,I mea ) are the measured values and ( Y cal, Z cal ,P cal ,I cal ) are the
calculated values using the initial estimates of the rate equation parameters.
The parameters available in the main tab allow the user to enter the values for current,
or for power in steady state. Using these numbers, the model will estimate the values
of the current.
Note: It is recommended to enter the values for current, rather than power, when
using the measured laser (as this is the realistic case).
The parameters in the measured tab are used to extract the physical/geometrical
properties of the laser. This extraction is completely independent of the parameters in
the main tab (current/power).
After finding the rate equation parameters, the laser measured works similarly to the
laser rate equations model. RIN is calculated according to [2][3].
102
LASER MEASURED
I ( t ) = I DC + I in ( t ) × I Pk (1)
Where I in ( t ) is the input signal current, I DC is the parameter Bias Current and
I Pk is the parameter Modulation peak current. If parameter Bias Current and
Modulation peak current have zero values, the internal current is given by I in ( t ) only.
The user can also calculate the subtracted IM response from the measured IM
response curves (Figure1) and load a file with this information into the component.
This will allow a preoptimization step, where the component fits the parameters Z and
Y to the measured results.
The file format for the subtracted IM response data is the following:
Frequency0 SubtractedIM0
Frequency1 SubtractedIM1
Frequency2 SubtractedIM2
103
LASER MEASURED
...
FrequencyN SubtractedIMN
The laser measured can also include the turnon delay parameter in the optimization
process. In this case, the turnon delay value specified defines the time needed for
the carrier density to reach the threshold carrier density when the laser current rises
to the reference current. The calculation of the turnon delay is based on the definition
find in [1].The laser linewidth parameter can be included in the optimization process
by defining the linewidth value for the laser when the bias current is the reference
current parameter [4]. The RIN is calculated according to [2][3] and the user has to
define the average RIN value in the defined frequency range.
If parameter Include noise is enabled, the Langevin noise terms for photon and
electron densities are included in the model[4]. If Include phase noise is enabled, the
Langevin noise term for the phase is included in the model. The Laser Measured
supports individual samples for timedriven simulation.
References
[1] Cartledge, J. C. and Srinivasan, R. C. “Extraction of DFB laser rate equation parameters for
system simulation purposes”, J. Light. Techn., 15, 852860, (1997).
[2] Yamada, M. "Variation of intensity noise and frequency noise with the spontaneous emission
factor in semiconductor lasers". IEEE Journal of Quantum Electronics. Volume 30, Issue 7, July
1994 Page(s):1511  1519.
[3] Agrawal, G.P., FiberOptic Communication Systems, Second edition. John Wiley & Sons, Inc.,
N.Y., (1997).
[4] Agrawal GP, Dutta NK. Semiconductor lasers, 2nd ed. New York: Van Nostrand Reinhold,
1993.
[5] K.Petermann, Laser Diode Modulation and Noise, Kluwer Academic Publishers,1988
104
FABRY PEROT LASER
Ports
Parameters
Main
105
FABRY PEROT LASER
Physical
Side Mode
106
FABRY PEROT LASER
Numerical
Simulation
Parameterized Parameterized — —
Noise
Random numbers
Technical background
The modulation dynamics of the FP laser are modeled by the coupled rate equations
which describe the relation between the carrier density N(t), photon densities Si(t),
and optical phases φi(t):
107
FABRY PEROT LASER
dS i ( t ) 1 ΓβN ( t ) (2)
 = G i ⋅ ( N ( t ) – N t ) ⋅  ⋅ S i ( t ) – ( γ p ⋅ S i ( t ) ) + 
dt τn
(1 + ε ⋅ S (t)) ∑ i
dφ i ( t ) (3)
 = 1 ⋅ α [ Γ ⋅ G i ⋅ ( N ( t ) – N t ) – γ p ]
dt 2
1
G i = v g ⋅ a o ⋅ 

⎛ 1 + ⎛ 2i ( f i – f o )⎞ 2⎞

⎝ ⎝ Δf ⎠ ⎠
108
FABRY PEROT LASER
I ( t ) = I DC + I in ( t ) × I Pk (6)
Where I in ( t ) is the input signal current, I DC is the parameter Bias Current and
I Pk is the parameter Modulation peak current. If parameter Bias Current and
Modulation peak current have zero values, the internal current is given by I in ( t ) only.
E(t ) = ∑ P i exp ( j ⋅ Δω i ⋅ t + φ i )
i
with
ηo ⋅ h ⋅ vi ⋅ αm ⋅ V ⋅ Si
P i = 
Γ
109
FABRY PEROT LASER
The component also allows injection of external light coupled to the longitudinal
modes. The coupling constant is given by:
vg
K c = 
L ⋅ Rf
References
[1] Agrawal GP, Dutta NK. Semiconductor lasers, 2nd ed. New York: Van Nostrand Reinhold, 1993.
110
LED
LED
Ports
Parameters
Main
Simulation
Parameterized Parameterized — —
111
LED
Random numbers
Technical background
In this model, the mean of the optical power is a function of the modulation current
(input signal). The conversion of the current into optical power is described by the
responsivity of the LED:
i(t)
P = η ⋅ h ⋅ f ⋅ 
q
where η is the quantum efficiency
The modulated characteristics depend of the electron lifetime and the device of the
diode, and are modeled by the transfer function applied to the current:
1
H ( f ) = 
1 + j ⋅ 2 ⋅ π ⋅ f ⋅ ( τ n + τ rc )
Note: The noise bins signals are not produced by this modulator.
112
WHITE LIGHT SOURCE
Ports
Parameters
Main
Emission frequency
Simulation
113
WHITE LIGHT SOURCE
Noise
Random numbers
Technical background
The average output Power or Power spectral density and Frequency are parameters
that you specify. This model generates noise bins or sampled signals at the output
according to:
·
Ex ( t ) x x ( t ) + j·y x ( t )
= ⋅ P⁄4
Ey ( t ) r v
x y ( t ) + j yy ( t )
A Gaussian distribution has been assumed to describe the probability density function
for the real and imaginary part of Ex and Ey. P is the average power when PSD
parameter is false. If PSD is true, then P is calculated from the power spectral density
multiplied by the Sample rate.
114
PUMP LASER
Pump Laser
Ports
Parameters
Main
Emission frequency
Polarization
115
PUMP LASER
Simulation
Technical background
In the CW Laser case, average output Power is a parameter that you specify. This
model generates only parameterized signal at the output.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛⎜ 1 – k⎞⎟ ⋅ P
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
where the power splitting k and the phase difference θ are related to the parameters
Azimuth α and Ellipticity ε as follows:
2 k ( 1 – k ) cos ( θ )
tan ( 2α ) = 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
116
PUMP LASER ARRAY
Ports
Parameters
Main
Frequency
117
PUMP LASER ARRAY
Power
118
PUMP LASER ARRAY
Polarization
Simulation
119
PUMP LASER ARRAY
Notes:
120
CONTROLLED PUMP LASER
This component is a pump laser that can be controlled by an electrical analog signal.
It allows the design and simulation of automatic gain control schemes for optical
amplifiers, such as control loops for the pump laser current.
Ports
Parameters
Main
Emission frequency
121
CONTROLLED PUMP LASER
Control
Bias 0 — ]INF,+INF[
Polarization
Simulation
Technical background
The controlled pump laser designed for analog control of the output pump power. The
input signal is first scaled by the parameters Gain and Bias. If the value of the scaled
signal is less than the Maximum input current and greater than the Threshold current
the current is multiplied by the Slope efficiency. The model supports individual
samples for time driven simulation
122
CW LASER ARRAY
CW Laser Array
Ports
Parameters
Main
123
CW LASER ARRAY
Frequency
Power
124
CW LASER ARRAY
Polarization
Simulation
Parameterized Parameterized — — —
Noise
125
CW LASER ARRAY
Random numbers
126
CW LASER ARRAY ES
CW Laser Array ES
Ports
Parameters
Main
127
CW LASER ARRAY ES
Power
Polarization
128
CW LASER ARRAY ES
Simulation
Parameterized Parameterized — — —
Noise
Random numbers
Technical background
The CW Laser Array ES is equivalent to the conventional CW Laser Array
component. However, The CW Laser Array ES model is easier to set up for WDM
systems, because it only requires the initial laser emission frequency and the spacing.
The signal output power is the same for all the output signals.
129
CW LASER ARRAY ES
Notes:
130
CW LASER MEASURED
CW Laser Measured
Generates a continuous wave (CW) optical signal based on measurements. You can
enter parameters such as linewidth, side mode suppression, and relative intensity
noise (RIN).
Ports
Parameters
Main
Emission frequency
131
CW LASER MEASURED
Side Mode
RIN
132
CW LASER MEASURED
Polarization
Simulation
Parameterized Parameterized — — —
Noise
Noise bins spacing 100 GHz Hz, GHz, THz, [1, 1000]
nm
Determines noise bins spacing
133
CW LASER MEASURED
Random numbers
134
CW LASER MEASURED
Technical background
This model is similar to the CW Laser — however, it includes additional effects, such
as multiple side modes and RIN.
If the you enable the parameter Calculate side mode, the side mode will be generated
according to:
jϕ
E out ( t ) = P [ 1 + s cos ( 2πΔ f t ) + s cos ( – 2πΔ f t )e ]
where P is laser output power, s is the parameter Suppression ratio in linear scale,
and Δ f is defined by the parameter Separation.
If the parameter Independent side mode is enabled, the average signal power will be
greater than P, since it includes the contribution from the side mode. If this parameter
is disabled, the output power will be P. This means that the signal will be scaled in
order to give the same average power. The signal phase and polarization is calculated
in the same way as the CW laser.
The model can also works as a FabryPerot laser; in this case, the parameter Number
of side modes defines the number of modes of the laser. The normalized power for
each mode is calculated based on the power of the central mode and the power of the
first side mode [1], according to:
1
P n = 2
1 n
1 + ⎛  – 1⎞ ⎛ ⎞
⎝ P s ⎠ ⎝ M⎠
where M is the parameter Number of side modes, n is the index of each side mode
pair, and P s is calculated from the power of the first side mode:
1
P s = 
⎛ 1 – 1⎞ M 2 + 1
⎝s ⎠
If the parameter Include RIN is enabled, the model generates noise bins with
bandwidth and spacing that you define. The parameter RIN is the ratio of the mean
square optical intensity noise to the square of the average power [2][3]:
2
〈 ΔP 〉
RIN = 
2
dB ⁄ Hz
Pm
2
where 〈 ΔP 〉 is the meansquare optical intensity fluctuation at a specific frequency
2 2
and P m is the parameter Measured power. This models estimates 〈 ΔP 〉 based on the
parameters RIN and Measured power.
The signal phase and polarization is calculated in the same way as the CW laser,
where the laser phase noise is modeled using a Gaussian random variable for the
135
CW LASER MEASURED
phase difference between two successive time instants with zero mean and a
variance equal to 2π Δf , where Δf is the laser Linewidth.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛ 1 – k⎞ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ are calculated from the parameters
Azimuth α and Ellipticity ε :
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
References
[1] Agrawal, G.P. and Dutta, N.K., “Semiconductor Laser”, 2nd Edition, Van Nostrand Reinhold,
New York, N.Y., (1993).
[2] Lau, K. Y. and Yariv, A., "UltraHigh Speed Semiconductor Laser", J. Quant. Elect., 21, 121136,
(1985).
[3] Agrawal, G.P., FiberOptic Communication Systems, Second edition. John Wiley & Sons, Inc.,
N.Y., (1997).
136
DIRECTLY MODULATED LASER MEASURED
Directly modulated laser that allows you to specify the dynamic of the laser based on
measured parameters. You can also enter parameters such as linewidth, chirp, side
mode, suppression and relative intensity noise (RIN).
Ports
Parameters
Main
Emission frequency
137
DIRECTLY MODULATED LASER MEASURED
Measurements
Undershoot 30 % — [0,+INF[
Damping time leading edge 1/(Bit rate) * 0.5 s s, ms, ns, ps [0,+INF[
Damping time trailing edge 1/(Bit rate) * 0.5 s s, ms, ns, ps [0,+INF[
Resonant frequency leading edge (Bit rate) * 5 Hz Hz, MHz, GHz, [0,+INF[
THz
Frequency of the oscillations in the transition from
0 to 1
Resonant frequency trailing edge (Bit rate) * 5 Hz Hz, MHz, GHz, [0,+INF[
THz
Frequency of the oscillations in the transition from
1 to 0
138
DIRECTLY MODULATED LASER MEASURED
Side Mode
RIN
Chirp
139
DIRECTLY MODULATED LASER MEASURED
Polarization
Simulation
Noise
Noise bins spacing 100 GHz Hz, GHz, THz, [1, 1000]
nm
Determines noise bins spacing
Random numbers
140
DIRECTLY MODULATED LASER MEASURED
Technical background
This model is a different from the Laser Measured, where you can enter measured
parameters and the model calculates the rate equation parameter by using
sophisticated optimization routines. Here you can enter measured parameters that
describe the laser dynamics by building the laser output signal.
If the parameter Configuration is Digital, the range of the amplitude of the signal input
is normalized between 0 and 1. This means that this model converts the input signal
to a sequence of squared pulses.
The parameter Power is the steady state value of the output power at the 1 level. The
steadystate value for the power at the 0 level is calculated from the parameter
Extinction ratio:
Er = 10 log ( P 1 ⁄ P 0 )
where P1 is the parameter Power, Er is the parameter Extinction ratio, and P0 is the
steadystate power at the 0 level.
The measured parameters will be used to build P(t) (see Figure 1).
141
DIRECTLY MODULATED LASER MEASURED
If you enable the parameter Calculate side mode, the side mode is generated
according to:
jϕ
E out ( t ) = P ( t ) [ 1 + s cos ( 2πΔ f t ) + s cos ( – 2πΔ f t )e ]
where P is laser output power, s is the parameter Suppression ratio in linear scale,
and Δf is defined by the parameter Separation.
The model can also works as a FabryPerot laser; in this case, the parameter Number
of side modes defines the number of modes of the laser. The normalized power for
each mode is calculated based on the power of the central mode and the power of the
first side mode [1], according to:
1
P n = 2
1 n
1 + ⎛  – 1⎞ ⎛ ⎞
⎝ P s ⎠ ⎝ M⎠
where M is the parameter Number of side modes, n is the index of each side mode
pair, and P s is calculated from the power of the first side mode:
1
P s = 
⎛ 1 – 1⎞ M 2 + 1
⎝s ⎠
If the parameter Configuration is Analog, the model will use the parameters Threshold
current and Slope efficiency to scale the input signal, without normalization.Different
from the Digital, the Analog configuration supports individual samples for time driven
simulation.
If the parameter Include RIN is enabled, the model will generate noise bins with
bandwidth and spacing that you define. The parameter RIN is the ratio of the mean
square optical intensity noise to the square of the average power [2][3]:
2
〈 ΔP 〉
RIN = 
2
dB ⁄ Hz
Pm
2
where 〈 ΔP 〉 is the meansquare optical intensity fluctuation at a specific frequency
2
and P m is the parameter Measured power.
2
This model estimates 〈 ΔP 〉 based on the parameters RIN and Measured power.
142
DIRECTLY MODULATED LASER MEASURED
dϕ α d
 = e  InP ( t ) + κP ( t )
dt 2 dt
The signal phase and polarization is calculated in the same way as the CW laser,
where the laser phase noise is modeled using a Gaussian random variable for the
phase difference between two successive time instants with zero mean and a
variance equal to 2π Δf , where Δf is the laser Linewidth. The probability density
function is:
2
Δϕ
1 – 
4πΔfdt
f ( Δϕ ) =  ⋅ e
2π Δfdt
where Δϕ is the phase difference between two successive time instants and dt is the
time discretization.
The output is multiplied with a complex vector considering the state of polarization:
⎛ E X ( t )⎞ = ⎛ 1 – k⎞ ⋅ P ( t )
⎝ E Y ( t )⎠ ⎝ ke jθ ⎠
The power splitting k and the phase difference θ is calculated from the parameters
Azimuth α and Ellipticity ε :
k ( 1 – k ) cos ( θ )
tan ( 2α ) = 2 
1 – 2.k
sin ( 2ε ) = 2 k ( 1 – k ) sin ( θ )
References
[1] Agrawal, G.P. and Dutta, N.K., “Semiconductor Laser”, 2nd Edition, Van Nostrand Reinhold,
New York, N.Y., (1993).
[2] Lau, K. Y. and Yariv, A., "UltraHigh Speed Semiconductor Laser", J. Quant. Elect., 21, 121136,
(1985).
[3] Agrawal, G.P., FiberOptic Communication Systems, Second edition. John Wiley & Sons, Inc.,
N.Y., (1997).
143
DIRECTLY MODULATED LASER MEASURED
Notes:
144
VCSEL LASER
VCSEL Laser
Ports
Parameters
Main
Thermal
145
VCSEL LASER
Physical
Coefficient in 1/s
Meaurements
146
VCSEL LASER
Coefficients for the polynomial function of temperature for the 2.545e5 a1=A/C,
offset current curve
2.908e7  a2=A/C2,
2.531e10 a3=A/C3…
1.022e12
Coefficients for the polynomial function of temperature for the b1= V1/2/C,
currentvoltage curve
b2= V1/2/C2,
b3= V1/2/C3…
6.683e8 
4.296e9
LI curves filename LI
Temperature.dat
The filename with the measurements of the LI curves,
including the temperature dependence
IV curves filename IV
Temperature.dat
The filename with the measurements of the IV curves,
including the temperature dependence
The values loaded from the LI curves filename Col 2: C Col 2: [INF,+INF]
The values loaded from the IV curves filename Col 2: C Col 2: [INF,+INF]
147
VCSEL LASER
Numerical
Graphs
To 40 mA [0, +INF]
Simulation
Parameterized Parameterized
Noise
148
VCSEL LASER
Random numbers
149
VCSEL LASER
Graphs
Results
Voltage V
Quantum efficiency
Scaling factor W
a0 A
a1 A/C
a2 A/C^2
a3 A/C^3
a4 A/C^4
a5 A/C^5
a6 A/C^6
a7 A/C^7
a8 A/C^8
a9 A/C^9
b0 V^.5
b1 V^.5/C
b2 V^.5/C^2
150
VCSEL LASER
b4 V^.5/C^4
b5 V^.5/C^5
b6 V^.5/C^6
b7 V^.5/C^7
b8 V^.5/C^8
b9 V^.5/C^9
c0 V^.5
c1 V^.5/A
c2 V^.5/A^2
c3 V^.5/A^3
c4 V^.5/A^4
c5 V^.5/A^5
c6 V^.5/A^6
c7 V^.5/A^7
c8 V^.5/A^8
c9 V^.5/A^9
Technical Background
The modulation dynamics of the laser are modeled by coupled rate equations that
describe the relationship between the carrier density N(t), photon density S(t), and
between the optical phase Φ ( t ) and temperature T(t)[1][2].
dN ( t ) η i ( I ( t ) – I off ( t ) ) N ( t ) 1
 =  –  – g 0 ⋅ ( N ( t ) – N t ) ⋅  ⋅ S ( t ) (1)
dt q⋅V τn (1 + ε ⋅ S(t))
dS ( t ) 1 S(t) Γ ⋅ β ⋅ N(t)
 = Γ ⋅ g 0 ⋅ ( N ( t ) – N t ) ⋅  ⋅ S ( t ) –  +  (2)
dt (1 + ε ⋅ S(t)) τp τn
dφ ( t ) 1 1
 =  ⋅ α ⋅ Γ ⋅ g 0 ⋅ ( Nt – N t ) –  (3)
dt 2 τp
dT ( t ) 1
 =  ( T 0 + ( IV (I,T) – P 0 )R th – T ) (4)
dt τ th
151
VCSEL LASER
S ⋅ V ⋅ η0 ⋅ h ⋅ v
P 0 =  (5)
2 ⋅ Γτ p
1 dφ
Δv =  ⋅  (6)
2 ⋅ π dt
where
152
VCSEL LASER
By enabling the parameter Reduce parameters, the user can enter the alternative
parameters that will be used to calculate N t , η o and a o according to:
N0
N t =  (7)
V
G0 V
a 0 =  (8)
vg
2kτ p
η o =  (9)
hv
where
2 3 4 5 6 7 8 9
I off ( T ) = a 0 + a 1 T + a 2 T + a 3 T + a 4 T + a 5 T + a 6 T + a 7 T + a 8 T + a 9 T
When the parameter Parameter fitting is disabled, the component will calculate using
userdefined parameters. In this case, the user should provide all the parameters,
including the coefficient for the polynomial functions. The measured LI and IV curves
will not be used in the calculation.
153
VCSEL LASER
When the parameter Parameter fitting is enabled, the component will calculate new
parameters using the current parameters as a first guess, including the number and
the initial values for the polynomial coefficients.
First the component will calculate the coefficients for the IV curve, and then it will
calculate the coefficients for the offset current, the thermal impedance and the new
slope efficiency.
The maximum value of the input current is calculated from the current derivative of
the IV curve. However, the user should provide this value as an input parameter.
The parameters will be adjusted to reflect the new slope efficiency. The affected
parameters are the active layer volume and the quantum efficiency.
For each calculation, the component will also generate the peak power and voltage
results based on the bias and modulation peak current. These values can be used for
external parameter fitting if the user intends to use a different fitting engine.
The range for the current value should be the same for both files. If the range is not
the same, the parameterfitting engine will not converge to an optimum fitting.
For example, if the LI curve is provided from 0 to 40 mA, the IV curve must be also
provided from 0 to 40 mA.
The default parameters of the VCSEL are the same as in [2]. If the parameter Thermal
effects is disabled, the calculation will perform using the same equations as in [1],
without the thermal effects and the parameter fitting.
Parameters Bias current and Modulation peak current are scale factors applied to the
input electrical signal.
154
VCSEL LASER
I ( t ) = I DC + I in ( t ) × I Pk (4)
Where I in ( t ) is the input signal current, I DC is the parameter Bias Current and
I Pk is the parameter Modulation peak current. If parameter Bias Current and
Modulation peak current have zero values, the internal current is given by I in ( t ) only.
References
[1] J. C. Cartledge and G. S. Burley, "The Effect of the Laser Chirping on Lightwave System
Performance", J. Lightwave Technology, vol. 7, pp. 568573, March 1989.
[2] P. V. Mena, J. J. Morikuni, S. M. Kang, A. V. Harton and K. W. Wyatt, "A Simple RateEquation
Based Thermal VCSEL Model", J. Lightwave Technology, vol. 17, pp. 865872, May 1999.
155
VCSEL LASER
Notes:
156
SPATIAL CW LASER
Spatial CW Laser
This component is CW laser that includes transverse mode profiles in the optical
output. It is a subsystem built using the CW Laser and the Multimode Generator.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THZ, nm [10, 10000]
Emission frequency
Output power
Laser linewidth
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg [90, 90]
157
SPATIAL CW LASER
Spatial effects
Name and description Default value Default unit Units Value range
Power ratio array 1
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
158
SPATIAL CW LASER
Name and description Default value Default unit Units Value range
Sample rate Sample rate Hz Hz, GHz, THz [1, 1e+100]
Random Numbers
Technical Background
The layout of the Spatial CW Laser is presented in Figure 1. Refer to CW Laser and
Multimode Generator component documentation for the technical background of the
models.
159
SPATIAL CW LASER
Notes:
160
SPATIOTEMPORAL VCSEL
Spatiotemporal VCSEL
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THZ, nm [10, 10000]
Laser emission frequency
161
SPATIOTEMPORAL VCSEL
Thermal
Name and description Default value Default unit Units Value range
Thermal effects NO [YES, NO]
162
SPATIOTEMPORAL VCSEL
Geometrical
Name and description Default value Default unit Units Value range
Cavity length 9e005 cm [0, 1e+100]
Physical
Name and description Default value Default unit Units Value range
Group velocity 7137915666.667 cm/s [0, 1e+100]
163
SPATIOTEMPORAL VCSEL
Name and description Default value Default unit Units Value range
Bottom mirror reflectivity for sine 0.9985, 0.9985, [0, 1]
modes 0.9985, 0.9985,
0.9985, 0.9985,
0.9985
Internal loss for cosine modes 40, 40, 40, 40, 40, 1/cm [0, 1e+100]
40, 40
Internal loss for sine modes 40, 40, 40, 40, 40, 1/cm [0, 1e+100]
40, 40
Enhanced
Name and description Default value Default unit Units Value range
Parasitic effects NO [YES, NO]
164
SPATIOTEMPORAL VCSEL
Numerical
Graphs
Name and description Default value Default unit Units Value range
Calculate graphs NO [YES, NO]
To 40 mA [0, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
165
SPATIOTEMPORAL VCSEL
Name and description Default value Default unit Units Value range
Space width X Space width X um [1e100, 1e+100]
Noise
Name and description Default value Default unit Units Value range
Include noise YES [YES, NO]
Random Numbers
Graphs
Technical Background
This module simulates a spatiotemporal model of a VCSEL and is based on the
publications of Jungo et al [1][2][3][4]. It is an improved version, since it includes an
LP mode solver and parameters to control whether temperature, parasitic and
feedback effects are included in the calculation or not.
Parameters Bias current and Modulation peak current are scale factors applied to the
input electrical signal.
166
SPATIOTEMPORAL VCSEL
I ( t ) = I DC + I in ( t ) × I Pk (3)
Where I in ( t ) is the input signal current, I DC is the parameter Bias Current and
I Pk is the parameter Modulation peak current. If parameter Bias Current and
Modulation peak current have zero values, the internal current is given by I in ( t ) only.
Due to the complexity of this component, we only give the list of parameters. For
further information about the spatiotemporal model refer to the work of Jungo [1],
where the exact mathematical derivation and formulation of the core model as well as
of the advanced mechanisms can be found.
References
[1] Jungo, M., "Spatiotemporal VCSEL Model for Advanced Simulations of Optical Links,"in Series
in Quantum Electronics, vol. 30, edited by H. Baltes, P. Günter, U. Keller, F. K. Kneubühl, W.
Lukosz, H. Mechior, and M. W. Sigrist, 1st ed.Konstanz: HartungGorre Verlag, 2003
[2] Jungo, M.X.; Erni, D.; Bachtold, W., "VISTAS: a comprehensive systemoriented
spatiotemporal VCSEL model", IEEE Journal of Selected Topics in Quantum Electronics, pp.
939  948. Volume 9, Issue 3, MayJune 2003
[3] G. Sialm, D. Lenz, D. Erni, G. L. Bona, C. Kromer, M. X. Jungo, T. Morf, F. Ellinger, and H.
Jäckel, "Comparison of Simulation and Measurement of Dynamic FiberCoupling Effects for
HighSpeed Multimode VCSELs," J. Lightwave Technol. 23, 2318 (2005)
[4] M. Jungo; D. Erni; W. Baechtold, "D VCSEL model for investigation of dynamic fiber coupling
and spatially filtered noise”, IEEE Photonics Technology Letters, pp. 3  5, Volume 15, Issue 1,
Jan. 2003
167
SPATIOTEMPORAL VCSEL
Notes:
168
SPATIAL VCSEL
Spatial VCSEL
This component is VCSEL laser that includes transverse mode profiles in the optical
output. It is a subsystem built using the VCSEL laser and the Multimode Generator.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THZ, nm [10, 10000]
Emission frequency
Thermal
Name and description Default value Default unit Units Value range
Thermal effects YES [YES, NO]
169
SPATIAL VCSEL
Name and description Default value Default unit Units Value range
Thermal time constant 1e006 S [0, 1e+100]
Physical
Name and description Default value Default unit Units Value range
Reduce parameters YES [YES, NO]
Measurements
Name and description Default value Default unit Units Value range
Max input current 40 mA [0, 1e+100]
The maximum value for the signal input
current, it should match the maximum
value of the measurements
170
SPATIAL VCSEL
Name and description Default value Default unit Units Value range
a  Ioff(T) 1.246e3
2.531e10
1.022e12
b  V(T) 1
4.154e7
6.683e8
4.296e9
LI curves filename LI
Temperature.dat
The filename with the measurements of
the LI curves, including the temperature
dependence
IV curves filename IV
Temperature.dat
The filename with the measurements of
the IV curves, including the temperature
dependence
LI curves at different
temperatures (A C W)
The values loaded from the LI curves
filename
IV curves at different
temperatures (A C V)
The values loaded from the IV curves
filename
171
SPATIAL VCSEL
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1
172
SPATIAL VCSEL
Numerical
Graphs
Name and description Default value Default unit Units Value range
Calculate graphs NO [YES, NO]
To 40 mA [0, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
173
SPATIAL VCSEL
Noise
Name and description Default value Default unit Units Value range
Include noise YES [YES, NO]
Random Numbers
Graphs
Results
Voltage (V)
Quantum efficiency
174
SPATIAL VCSEL
a0 (A)
a1 (A/C)
a2 (A/C^2)
a3 (A/C^3)
a4 (A/C^4
a5 (A/C^5)
a6 (A/C^6)
a7 (A/C^7)
a8 (A/C^8)
a9 (A/C^9)
b0 (V^0.5)
b1 (V^0.5/C)
b2 (V^0.5/C^2)
b3 (V^0.5/C^3)
b4 (V^0.5/C^4)
b5 (V^0.5/C^5)
b6 (V^0.5/C^6)
b7 (V^0.5/C^7)
b8 (V^0.5/C^8)
b9 (V^0.5/C^9)
c0 (V^0.5)
c1 (V^0.5/A)
c2 (V^0.5/A^2)
c3 (V^0.5/A^3)
c4 (V^0.5/A^4)
c5 (V^0.5/A^5)
c6 (V^0.5/A^6)
c7 (V^0.5/A^7)
c8 (V^0.5/A^8)
175
SPATIAL VCSEL
Technical Background
The layout of the Spatial VCSEL is presented in Figure 1. Refer to VCSEL Laser and
Multimode Generator component documentation for the technical background of the
models.
176
SPATIAL LASER RATE EQUATIONS
This component is laser based on rate equations that includes transverse mode
profiles in the optical output. It is a subsystem built using the Laser Rate Equations
component and the Multimode Generator.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THz, nm [10,10000]
Emission frequency of the laser
177
SPATIAL LASER RATE EQUATIONS
Name and description Default value Default unit Units Value range
Threshold current 33.4572 mA — [0, 1000]
Physical
Name and description Default value Default unit Units Value range
Active layer volume 1.5e010 cm^3 [0, 0.001]
Spatial effects
Name and description Default value Default unit Units Value range
Power ratio array 1
178
SPATIAL LASER RATE EQUATIONS
Name and description Default value Default unit Units Value range
Pol. X m,n index array 00
Numerical
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
179
SPATIAL LASER RATE EQUATIONS
Name and description Default value Default unit Units Value range
Space width Y Space width Y um [1e100, 1e+100]
Noise
Name and description Default value Default unit Units Value range
Include noise YES [YES, NO]
Random Numbers
Technical Background
The layout of the Spatial Laser Rate Equations is presented in Figure 1. Refer to Laser
Rate Equations and Multimode Generator component documentation for the technical
background of the models.
180
SPATIAL LED
Spatial LED
This component is an LED that includes transverse mode profiles in the optical output.
It is a subsystem built using the LED component and the Multimode Generator.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THz, nm [10, 10000]
Emission frequency
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1
181
SPATIAL LED
Name and description Default value Default unit Units Value range
Mode type Laguerre Laguerre
Gaussian Gaussian,
Defines the output signal mode types Hermite
Gaussian
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
182
SPATIAL LED
Random numbers
Technical Background
The layout of the Spatial LED is presented in Figure 1. Refer to LED and Multimode
Generator component documentation for the technical background of the models.
183
SPATIAL LED
Notes:
184
SPATIAL LED
Transmitters Library
Optical Transmitters
• WDM Transmitter
• Optical Transmitter
• Optical Duobinary Transmitter
• Optical DPSK Transmitter
• Optical CSRZ Transmitter
• Optical QPSK Transmitter
• Optical DPQPSK Transmitter
• Spatial Optical Transmitter
185
SPATIAL LED
Notes:
186
WDM TRANSMITTER
WDM Transmitter
Ports
Parameters
Main
187
WDM TRANSMITTER
PRBS
MBits/s
GBits/s
Coding
188
WDM TRANSMITTER
Enhanced
Overshoot 30 % 
Percentage of overshoot during the
transition from 0 to 1 relative to the steady
state power
Undershoot 30 % 
Percentage of undershoot during the
transition from 1 to 0 relative to the steady
state power
Side Mode
189
WDM TRANSMITTER
RIN
Chirp
Polarization
190
WDM TRANSMITTER
Simulation
Noise
Random numbers
191
WDM TRANSMITTER
Technical background
WDM systems require multiple transmitters and different parameters for each one of
them. In addition, they also require different modulation schemes and formats. By
using multiple components, users can customize designs, but it is time consuming.
The WDM Transmitter encapsulates different components, allowing users to select
different modulation formats and schemes for multiple channels in one single
component. It is a transmitter array that allows for different modulation types and
schemes.
The block diagram for each WDM channel transmitter is shown below:
The first stage is the PRBS; the same engine used in the PseudoRandom Bit
Sequence Generator component is used in this stage. Parameters Bit rate, Order,
Number of leading and trailing zeros are used in the internal PseudoRandom Bit
Sequence Generator. A different seed will be used for each bit sequence for each
WDM channel. The operation and parameters of the PRBS component is described
in the technical background of the PseudoRandom Bit Sequence Generator.
The second stage is the Coding/Modulation; the parameter Modulation type has three
options: RZ, NRZ and Off. RZ and NRZ coding is generated by the engines of the RZ
Pulse Generator and NRZ Pulse Generator respectively. A CW operation of the
192
WDM TRANSMITTER
transmitter is possible by selecting Off as modulation type. The Duty cycle parameter
is used when modulation type RZ is selected. The operations and parameters of the
electrical pulse generators are described in the technical background of the RZ and
NRZ Pulse Generators.
The last stage is the optical source and modulation scheme; by using the parameter
Transmitter type the user can select between a external modulated laser scheme
(EML) or a directly modulated laser scheme (DML). The laser engine used in this
stage is the same used in the Directly Modulated Laser Measured component. The
operation and parameters of this component are described in the technical
background of the Directly Modulated Laser Measured.
193
WDM TRANSMITTER
Notes:
194
OPTICAL TRANSMITTER
Optical Transmitter
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 193.1 THz Hz, THz, nm [1, 10000]
Emission frequency
Output power
Laser linewidth
195
OPTICAL TRANSMITTER
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
Coding
Name and description Default value Default unit Units Value range
Modulation type NRZ   [Off, NRZ, RZ]
Enhanced
Name and description Default value Default unit Units Value range
Transmitter type EML   EML, DML
196
OPTICAL TRANSMITTER
Name and description Default value Default unit Units Value range
Overshoot 30 %  [0, 100]
Damping time leading edge 1/(Bit rate)*0.5 s s, ms, ns, ps [0, 1e+100]
Damping time trailing edge 1/(Bit rate)*0.5 s s, ms, ns, ps [0, 1e+100]
Resonant frequency leading (Bit rate)*5 Hz Hz, MHz, GHz, [0, 3e+015]
edge THz
Resonant frequency trailing edge (Bit rate)*5 Hz Hz, MHz, GHz, [0, 3e+015]
THz
Frequency of the oscillations in the
transition from high level to low level
Side Mode
Name and description Default value Default unit Units Value range
Calculate side mode False   True, False
197
OPTICAL TRANSMITTER
RIN
Name and description Default value Default unit Units Value range
Include RIN False   True, False
Chirp
Name and description Default value Default unit Units Value range
Alpha parameter 0   [100, 100]
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
198
OPTICAL TRANSMITTER
Name and description Default value Default unit Units Value range
Parameterized Parameterized   True, False
Noise
Name and description Default value Default unit Units Value range
Noise bandwidth Sample rate Hz Hz, GHz, THz, nm [0, 1e+100]
Noise bins spacing Sample rate Hz Hz, GHz, THz, nm [0, 1e+100]
Random numbers
Technical Background
Refer to WDM Transmitter for the technical background.
199
OPTICAL TRANSMITTER
Notes:
200
OPTICAL DUOBINARY TRANSMITTER
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 193.1 THz Hz, THz, nm [1, 10000]
Emission frequency
Output power
Laser linewidth
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
201
OPTICAL DUOBINARY TRANSMITTER
Name and description Default value Default unit Units Value range
Order log(Sequence   [0, 30]
length)/log(2)
Order of the PRBS
Coding
Name and description Default value Default unit Units Value range
Amplitude 2 a.u.  [1e+100,
1e+100]
Peaktopeak amplitude of the NRZ
pulse generator
Filter
Name and description Default value Default unit Units Value range
Filter type Bessel   [Butterworth,
Bessel]
Defines the filter type
202
OPTICAL DUOBINARY TRANSMITTER
Name and description Default value Default unit Units Value range
Order 4   [1, 100]
Modulator
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Simulation
Name and description Default value Default unit Units Value range
Enabled True  True, False
203
OPTICAL DUOBINARY TRANSMITTER
Name and description Default value Default unit Units Value range
Sample rate Sample rate Hz Hz, GHz, THz [1, 1e+100]
Random numbers
Technical Background
The layout representing the duobinary transmitter component is shown in Figure 1.
To generate the optical duobinary signal a CW laser source, a MachZehnder
modulator driven in a pushpull configuration to get a chirp free transmission, and a
NRZ pulse pattern generator were used. The NRZ duobinary signal was created
using a low pass Bessel/Butterworth filter; this signal then drives the MZ modulator.
In order to avoid recursive decoding in the receiver, a duobinary precoder was also
used. The duobinary precoder was composed of an exclusiveor gate with a delayed
feedback path.
204
OPTICAL DUOBINARY TRANSMITTER
205
OPTICAL DUOBINARY TRANSMITTER
Notes:
206
OPTICAL DPSK TRANSMITTER
This component simulates a single channel optical transmitter with Differential Phase
Shift Keying modulation.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 193.1 THz Hz, THz, nm [1, 10000]
Emission frequency
Output power
Laser linewidth
207
OPTICAL DPSK TRANSMITTER
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
Coding
Name and description Default value Default unit Units Value range
Duty cycle RZ33%   RZ33%, RZ
50%, RZ66%,
Duration of the high level bit NRZ
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
208
OPTICAL DPSK TRANSMITTER
Name and description Default value Default unit Units Value range
Parameterized Parameterized   True, False
Random numbers
Technical Background
The layout representing the Differential PhaseShift Keying transmitter component is
shown at Figure 1. The RZDPSK transmitter includes two modulators: one for phase
modulation of the data and one for amplitude modulation of the clock for RZ pulse
carving.
209
OPTICAL DPSK TRANSMITTER
The transmitter can simulate 3 DPSK signals: with 33%dutycycle RZ pulses, with
50%dutycycle RZ pulses, and with 66%dutycycle RZ pulses. Figure 2 (a), (b) and
(c) shows the correspondent spectra and time domain pulses respectively.
Figure 2 Spectra and time domain DPSK signals for (a) 33%dutycycle, (b) 50%dutycycle and (c)
66%dutycycle.
(a)
210
OPTICAL DPSK TRANSMITTER
(b)
(c)
211
OPTICAL DPSK TRANSMITTER
Notes:
212
OPTICAL CSRZ TRANSMITTER
This component simulates a single channel optical transmitter with an optical carrier
suppressed RZ signal.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 193.1 THz Hz, THz, nm [1, 10000]
Emission frequency
Output power
Laser linewidth
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
213
OPTICAL CSRZ TRANSMITTER
Name and description Default value Default unit Units Value range
Order log(Sequence   [0, 30]
length)/log(2)
Order of the PRBS
Coding
Name and description Default value Default unit Units Value range
Amplitude 1 a.u.  [1e+100,
1e+100]
Peaktopeak amplitude of the RZ pulse
generator
Modulator
214
OPTICAL CSRZ TRANSMITTER
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
Random numbers
215
OPTICAL CSRZ TRANSMITTER
Technical Background
The layout representing the CSRZ transmitter component is shown at the figure
below. The CSRZ signal is generated using a MZ modulator concatenated with a
phase modulator. The first modulator generates a RZ optical signal, and then a NRZ
electrical signal is applied to the phase modulator to generate an alternated phase in
the RZ signal.
216
OPTICAL QPSK TRANSMITTER
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
External laser False   True, False
Emission frequency
Output power
Laser linewidth
217
OPTICAL QPSK TRANSMITTER
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
Coding
Name and description Default value Default unit Units Value range
Gray code False   True, False
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
218
OPTICAL QPSK TRANSMITTER
Name and description Default value Default unit Units Value range
Parameterized Parameterized   True, False
Random numbers
Technical Background
The layout representing the optical coherent QPSK transmitter component is shown
in the figure below. The QPSK signal is generated by using MZ modulators to encode
the QPSK symbols onto an optical carrier. Each modulator branch modulates the in
phase (I) and quadrature components (Q) of a carrier.
219
OPTICAL QPSK TRANSMITTER
220
OPTICAL DPQPSK TRANSMITTER
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
External laser False   True, False
Emission frequency
Output power
Laser linewidth
221
OPTICAL DPQPSK TRANSMITTER
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
Coding
Name and description Default value Default unit Units Value range
Gray code False   True, False
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
222
OPTICAL DPQPSK TRANSMITTER
Random numbers
Technical Background
The layout representing the optical coherent dualpolarization QPSK transmitter
component is shown in the figure below. In this case, polarization multiplexing is used,
the laser output is split into two othogonal polarization components, which are
modulated separately by QPSK modulators (similar to the one shown in the QPSK
transmitter layout) and then combined using a polarization beam splitter (PBS).
223
OPTICAL DPQPSK TRANSMITTER
224
SPATIAL OPTICAL TRANSMITTER
This component is Optical transmitter that includes transverse mode profiles in the
optical output. It is a subsystem built using the WDM Transmitter Optical and the
Multimode Generator.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Frequency 850 nm Hz, THz, nm [1, 10000]
Emission frequency
Output power
Laser linewidth
225
SPATIAL OPTICAL TRANSMITTER
PRBS
Name and description Default value Default unit Units Value range
External PRBS False   True, False
Coding
Name and description Default value Default unit Units Value range
Modulation type NRZ   [Off, NRZ, RZ]
Enhanced
Name and description Default value Default unit Units Value range
Transmitter type EML   EML, DML
226
SPATIAL OPTICAL TRANSMITTER
Name and description Default value Default unit Units Value range
Overshoot 30 %  [0, 100]
Damping time leading edge 1/(Bit rate)*0.5 s s, ms, ns, ps [0, 1e+100]
Damping time trailing edge 1/(Bit rate)*0.5 s s, ms, ns, ps [0, 1e+100]
Resonant frequency leading (Bit rate)*5 Hz Hz, MHz, GHz, [0, 3e+015]
edge THz
Resonant frequency trailing edge (Bit rate)*5 Hz Hz, MHz, GHz, [0, 3e+015]
THz
Frequency of the oscillations in the
transition from high level to low level
Side Mode
Name and description Default value Default unit Units Value range
Calculate side mode False   True, False
227
SPATIAL OPTICAL TRANSMITTER
RIN
Name and description Default value Default unit Units Value range
Include RIN False   True, False
Chirp
Name and description Default value Default unit Units Value range
Alpha parameter 0   [100, 100]
Polarization
Name and description Default value Default unit Units Value range
Azimuth 0 deg  [90, 90]
Spatial Effects
Name and description Default value Default unit Units Value range
Power ratio array 1  
228
SPATIAL OPTICAL TRANSMITTER
Name and description Default value Default unit Units Value range
Mode type Laguerre   Laguerre
Gaussian Gaussian,
Defines the output signal mode types Hermite
Gaussian
Simulation
Name and description Default value Default unit Units Value range
Enabled True   True, False
229
SPATIAL OPTICAL TRANSMITTER
Name and description Default value Default unit Units Value range
Space width Y Space width Y um  [1e100, 1e+100]
Noise
Name and description Default value Default unit Units Value range
Noise bandwidth Sample rate Hz Hz, GHz, THz, nm [0, 1e+100]
Random numbers
Technical Background
The layout of the Spatial Optical Transmitter is presented in Figure 1. Refer to WDM
Transmitter and Multimode Generator component documentation for the Technical
Background of the models.
230
SPATIAL OPTICAL TRANSMITTER
231
SPATIAL OPTICAL TRANSMITTER
Notes:
232
SPATIAL OPTICAL TRANSMITTER
Transmitters Library
Modulators
Optical
• MachZehnder Modulator
• Electroabsorption Modulator
• Amplitude Modulator
• Phase Modulator
• Frequency Modulator
• Dual Drive MachZehnder Modulator Measured
• Electroabsorption Modulator Measured
• Single Drive MachZehnder Modulator Measured
• Dual Port Dual Drive MachZehnder Modulator Measured
• Lithium Niobate MachZehnder Modulator
233
SPATIAL OPTICAL TRANSMITTER
Notes:
234
MACHZEHNDER MODULATOR
MachZehnder Modulator
Ports
Parameters
Main
Simulation
Technical background
The MachZehnder modulator is an intensity modulator based on an interferometric
principle. It consists of two 3 dB couplers which are connected by two waveguides of
equal length (see Figure 1). By means of an electrooptic effect, an externally applied
voltage can be used to vary the refractive indices in the waveguide branches.
235
MACHZEHNDER MODULATOR
The different paths can lead to constructive and destructive interference at the output,
depending on the applied voltage. Then the output intensity can be modulated
according to the voltage.
where Δθ is the phase difference between the two branches and is defined as:
π
Δθ ( t ) =  ⋅ ( 0.5 – ER ⋅ ( Modulation ( t ) – 0.5 ) )
2
with
4 1
ER = 1 –  ⋅ arc tan ⎛ ⎞
π ⎝ extrat⎠
and
For parameterized and noise bins signals, the average power is calculated according
to the above.
236
ELECTROABSORPTION MODULATOR
Electroabsorption Modulator
Ports
Parameters
Main
Simulation
237
ELECTROABSORPTION MODULATOR
Technical background
In this model, the optical carrier is modulated externally by the electrical modulation
signal, (see Figure 1).
Figure 1 EA modulator
Assuming that the optical input signal is Ein, the following equation describes the
behavior of the model:
α
E out ( t ) = E in ( t ) ⋅ Mod ( t ) ⋅ exp ⎛ j  ⋅ ln ( Mod ( t ) )⎞
⎝ 2 ⎠
where Eout(t) is the output optical signal, α is the chirp factor, and Mod(t) is defined as
Mod ( t ) = ( 1 – MI ) + MI ⋅ modulation ( t )
where MI is the modulation index and modulation(t) is the electrical input signal. The
electrical input signal is normalized between 0 and 1.
For parameterized and noise bins signals, the average power is calculated according
to the above.
238
AMPLITUDE MODULATOR
Amplitude Modulator
Ports
Parameters
Main
Simulation
239
AMPLITUDE MODULATOR
Technical background
In this model, the optical carrier is modulated externally by the electrical modulation
signal. Assuming that the optical input signal is Ein, the following equations describe
the behavior of the model:
E out ( t ) = E in ( t ) ⋅ Mod ( t )
Mod ( t ) = ( 1 – MI ) + MI ⋅ modulation ( t )
where MI is the modulation index and modulation(t) is the electrical input signal. The
electrical input signal is normalized between 0 and 1.
For parameterized and noise bins signals, the average power is calculated according
to the above.
240
PHASE MODULATOR
Phase Modulator
Ports
Parameters
Main
Simulation
241
PHASE MODULATOR
Technical background
In this model, the electrical modulation signal imposes a phase modulation on an
optical carrier. Assuming that the optical input signal is Ein, the following equation
describes the behavior of the model.
where Eout(t) is the output optical signal, Δφ is the phase deviation, and modulation(t)
is the electrical input signal. The electrical input signal is normalized between 0 and 1.
The parameterized and noise bins signals are not affected by this modulator.
242
FREQUENCY MODULATOR
Frequency Modulator
Ports
Parameters
Main
Simulation
243
FREQUENCY MODULATOR
Technical background
In this model, the electrical modulation signal imposes a frequency modulation on an
optical carrier. Assuming that the optical input signal is Ein, the following equation
describes the behavior of the model:
t
⎛ ⎞
E out ( t ) = E in ( t ) ⋅ exp ⎜ j ⋅ 2π ∫ Δf ⋅ ( modulation ( τ ) – 0.5 ) dτ⎟
⎝ ⎠
0
where Eout(t) is the output optical signal, Δf is the frequency deviation, and
modulation ( τ ) is the electrical input signal. The electrical input signal is normalized
between 0 and 1.
The parameterized and noise bins signals are not affected by this modulator.
244
DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Ports
Parameters
Main
Simulation
245
DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Graphs
Technical background
In this model, you can specify the dependence of the measured absorption and phase
on applied voltage for a MachZehnder modulator. You can use the default
characteristics curves or choose to load from Filename.
For a modulator with the same input and output Ybranch splitting ratios, the output
signal is:
E0 Δα a ( V 1 ) Δα a ( V 2 )
E ( V 1 ,V 2 ) =  SR ⋅ exp ⎛⎝ – ⎛⎝  + j ⋅ Δβ ( V 1 )⎞⎠ L⎞⎠ + exp ⎛⎝ – ⎛⎝  + j ⋅ Δβ ( V 2 )⎞⎠ L – j ⋅ φ 0⎞⎠
1 + SR 2 2
E ( V 1 ,V 2 ) ≡ I ( V 1 ,V 2 ) ⋅ exp ( j ⋅ Φ ( V 1 ,V 2 ) )
Φ is the phase
V i ( i = 1, 2 ) is defined as:
where V bi is the bias voltage, V mod12 is the peaktopeak voltage, and v ( t ) is the
normalized modulation waveform with a peaktopeak amplitude of 1 and an average
value of 0. The electrical input signal can be normalized between 0.5 and 0.5.
246
DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Figure 1 Default characteristics of absorption and phase in the Dual MachZehnder model
247
DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
References
[1] Cartledge, J. C., “Combining selfphase modulation and optimum modulation conditions to
improve performance of 10 Gb/s transmission systems using MQW MachZehnder
modulators”, J. Light. Techn., 18, 647654, (2000).
248
ELECTROABSORPTION MODULATOR MEASURED
Ports
Parameters
Main
Simulation
249
ELECTROABSORPTION MODULATOR MEASURED
Graphs
Technical background
In this model, you can specify the dependence of the measured absorption and α 
parameter α m on the applied voltage for an EA modulator. You can use the default
characteristic curves or choose to load from file. In this case, the parameter Filename
is enabled.
In the case of the EA modulator, the output signal response to an applied voltage is:
1
E( V) = I ( V ) exp ⎛⎝ j  ∫ α m ( V ) d ln ( I ( V ) )⎞⎠ (1)
2
While Equation 1 is an accurate result, it is not in the most convenient form for
simulation purposes when empirical equations for α m ( V ) and I ( V ) are obtained
from a fitting to measured results. The determination of the argument of the
exponential function in Equation 1 requires function evaluation and integration.
The modulator output signal given by Equation 1 can also be written in the convenient
( 1 + jα ) ⁄ 2
form I using a voltagedependent parameter α r ( V ) as:
( 1 + jα r ( V ) ) ⁄ 2 (2)
E(V ) = I(V)
1
α r ( V ) =  ∫ α m ( V ) ⋅ dγ ( V ) (3)
γ(V)
250
ELECTROABSORPTION MODULATOR MEASURED
The default characteristics curves stored in the component, the dependence of the
measured absorption, and αparameter α m ( V ) on applied voltage, is illustrated in
Figure 1.
Figure 1 Dependence of the absorption and αm on the applied voltage for an MQWEAM
For this component, the electrical input signal can be normalized between 0.5 and
0.5. Then, the voltage applied to the modulator is given by:
(4)
V ( t ) = V b + V mod ⋅ v ( t )
where Vb is the bias voltage, Vmod is the peaktopeak voltage, and v(t) is the
normalized modulation waveform (electrical input signal) with a peaktopeak
amplitude of 1 and an average value of 0.
251
ELECTROABSORPTION MODULATOR MEASURED
Notes:
252
SINGLE DRIVE MACHZEHNDER MODULATOR MEASURED
Ports
Parameters
Main
253
SINGLE DRIVE MACHZEHNDER MODULATOR MEASURED
Simulation
Graphs
Technical background
In this model, you can specify the dependence of the measured absorption and phase
on applied voltage for a MachZehnder modulator. You can use the default
characteristics curves or choose to load from Filename.
For a modulator with the same input and output Ybranch splitting ratios, the output
signal is:
E0 Δα a ( V 1 ) Δα a ( V 2 )
E ( V 1 ,V 2 ) =  SR ⋅ exp ⎛⎝ – ⎛⎝  + j ⋅ Δβ ( V 1 )⎞⎠ L⎞⎠ + exp ⎛⎝ – ⎛⎝  + j ⋅ Δβ ( V 2 )⎞⎠ L – j ⋅ φ 0⎞⎠
1 + SR 2 2
E ( V 1 ,V 2 ) ≡ I ( V 1 ,V 2 ) ⋅ exp ( j ⋅ Φ ( V 1 ,V 2 ) )
Φ is the phase
254
SINGLE DRIVE MACHZEHNDER MODULATOR MEASURED
Figure 1 Default characteristics of absorption and phase in the Single MachZehnder mode
255
SINGLE DRIVE MACHZEHNDER MODULATOR MEASURED
References
[1] Cartledge, J. C., “Combining selfphase modulation and optimum modulation conditions to
improve performance of 10 Gb/s transmission systems using MQW MachZehnder
modulators”, J. Light. Techn., 18, 647654, (2000).
256
DUAL PORT DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Ports
Parameters
Main
257
DUAL PORT DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Simulation
Graphs
Technical background
In this model, you can specify the dependence of the measured absorption and phase
on applied voltage for a MachZehnder modulator. You can use the default
characteristics curves or choose to load from Filename.
For a modulator with the same input and output Ybranch splitting ratios, the output
signal is:
E0 Δα a ( V 1 ) Δα a ( V 2 )
E ( V 1 ,V 2 ) =  SR ⋅ exp ⎛ – ⎛  + j ⋅ Δβ ( V 1 )⎞ L⎞ + exp ⎛ – ⎛  + j ⋅ Δβ ( V 2 )⎞ L – j ⋅ φ 0⎞
1 + SR ⎝ ⎝ 2 ⎠ ⎠ ⎝ ⎝ 2 ⎠ ⎠
E ( V 1 ,V 2 ) ≡ I ( V 1 ,V 2 ) ⋅ exp ( j ⋅ Φ ( V 1 ,V 2 ) )
Φ is the phase
V i ( i = 1, 2 ) is defined as:
V i ( t ) = V bi ± V modi ⋅ v ( t ) for the normalized case
where V bi is the bias voltage, V modi is the peaktopeak voltage, and v ( t ) is the
normalized modulation waveform with a peaktopeak amplitude of 1 and an average
value of 0. The electrical input signal is normalized between 0.5 and 0.5.
258
DUAL PORT DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
Figure 1 Default characteristics of absorption and phase in the Dual MachZehnder model
259
DUAL PORT DUAL DRIVE MACHZEHNDER MODULATOR MEASURED
References
[1] Cartledge, J. C., “Combining selfphase modulation and optimum modulation conditions to
improve performance of 10 Gb/s transmission systems using MQW MachZehnder
modulators”, J. Light. Techn., 18, 647654, (2000).
260
LITHIUM NIOBATE MACHZEHNDER MODULATOR
Ports
Parameters
Main
261
LITHIUM NIOBATE MACHZEHNDER MODULATOR
Bandwidth Response
HF filename Filter.dat — —
Simulation
Technical background
The MachZehnder structure consists of an input optical branch, which splits the
incoming light into two arms, followed by two independent optical arms, which are
subsequently recombined by the output optical branch. Application of an electrical
signal to one of the optical arms controls the degree of interference at the output
optical branch and therefore controls the output intensity.
262
LITHIUM NIOBATE MACHZEHNDER MODULATOR
γ denotes the power splinting (combining) ration of arm two for the input (output,
respectively) Ybranch waveguide, and is given by:
1
γ = ⎛ 1 – ⎞ ⁄ 2
⎝ ε⎠ r
ExtRatio ⁄ 10
where ε r = 10 .
v bias1 and v bias2 , the DC bias voltages, are included separately as parameters due
to the possibility of the V πDC (Switching Bias Voltage) to be different from the
Switching RF Voltage.
If the Switching Bias Voltage is equal to the Switching RF Voltage, and the
Normalize Electrical Signal parameter is False, the bias voltage can be included in
the electrical signal.
The optical power and phase of the modulator output are determined in response to
the modulating voltage waveforms. The modulator transfer function relates the
effective drive voltage to the applied drive voltage. This component can also load the
modulator transfer function data from file or consider an ideal transfer function.
The file is formatted containing two items per line, the frequency and filter
measurement. The parameter File frequency unit determines the frequency unit of
the first item; it can be Hz or THz.
According to the parameter File format, the second item can be one value (Power or
Phase) or two values (Power and Phase or Real and Imag):
193.10 0
193.11 0.5
193.12 0.5
193.13 0
Power Phase
193.14 0 0
193.17 0 0
263
LITHIUM NIOBATE MACHZEHNDER MODULATOR
Real Imag
193.18 0 0
193.21 0 0
193.22 0
193.23 3.14
193.24 3.14
193.253 0
When the Normalize electrical signal parameter is True, the electrical signals of
port1 and port2 are normalized between 0.5 and 0.5. In this case, the amplitude of
each RF electrical signal considered in v 1 ( t ) and v 2 ( t ) will be the values in the
modulation voltage parameters divided by 2.
References
[1] Cartledge, J. C., Rolland, C., Lemerle, S., and Solheim, A., “Theoretical performance of 10 Gb/s
lightwave systems using a IIIV semiconductor MachZehnder modulator.”, IEEE Phot. Techn.
Letters., 6, 282284, (1994).
[2] Cartledge, J.C., "Performance of 10 Gb/s lightwave systems based on lithium niobate Mach
Zehnder modulators with asymmetric Ybranch waveguides". IEEE Phot. Techn. Letters., 7,
1090 1092, (1995).
264
LITHIUM NIOBATE MACHZEHNDER MODULATOR
Transmitters Library
Bit Sequence Generators
• PseudoRandom Bit Sequence Generator
• UserDefined Bit Sequence Generator
265
LITHIUM NIOBATE MACHZEHNDER MODULATOR
Notes:
266
PSEUDORANDOM BIT SEQUENCE GENERATOR
Ports
Parameters
Main
MBits/s
GBits/s
267
PSEUDORANDOM BIT SEQUENCE GENERATOR
Simulation
Random numbers
Technical background
This model generates a sequence of N bits:
where N = T w B r
NG = N – nl – nt
Tw is the global parameter Time window and Br is the parameter Bit rate.
The number of bits generated is N G . n l and n t are the Number of leading zeros and
the Number of trailing zeros.
Operation mode controls the algorithm used to generate the bit sequence:
• Probability: Random number generator is used, with parameter Mark probability
specifying the probability of ones in the sequence
• Order: PRBS generator[1] with Order k is used to generate a sequence with
period of 2k1
• Alternate: Alternate sequence of ones and zeros is generated
• Ones: A sequence of ones is generated
• Zeros: A sequence of zeros is generated
268
PSEUDORANDOM BIT SEQUENCE GENERATOR
References
[1] Press, W. H., Flannery, B. P., Teukolsky, S. A., and Vetterling, W. T., Numerical Recipes in C.
Cambridge University Press, (1991).
269
PSEUDORANDOM BIT SEQUENCE GENERATOR
Notes:
270
USERDEFINED BIT SEQUENCE GENERATOR
Ports
Parameters
Main
MBits/s
GBits/s
Number of leading zeros (Time window * 3 / 100) * Bit rate — [0, 1000]
Number of trailing zeros (Time window * 3 / 100) * Bit rate — [0, 1000]
271
USERDEFINED BIT SEQUENCE GENERATOR
Simulation
Technical background
You can enter the string Bit sequence or choose Load from file. In this, case the
parameter Filename is enabled.
All bit files are formatted containing one bit per line, e.g. the bit file representing the
sequence "01011..." has the following form:
N = TwBr
Tw is the global parameter Time window and Br is the parameter Bit rate. If the user
defined sequence is shorter than the N, the sequence will be repeated until the length
is equal to N.
272
Multimode Library
• Donut Transverse Mode Generator
• Hermite Transverse Mode Generator
• Laguerre Transverse Mode Generator
• Multimode Generator
• Measured Transverse Mode
• Mode ID Modifier
273
Notes:
274
DONUT TRANSVERSE MODE GENERATOR
This component attaches Donut transverse mode profiles to the input signal. It also
converts singlemode signals into multimode signals.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Power ratio array 1
275
DONUT TRANSVERSE MODE GENERATOR
Name and description Default value Default unit Units Value range
Pol. Y outer radius 5 um [1e100, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
The Donut [1] Transverse Mode Generator attaches mode profiles to the input signal
X and Y polarizations. A donut profile is attached to each polarization. Additionally,
singlemode inputs can be converted to a multimode signal scaled by a userdefined
power distribution.
The parameter Power ratio array is used to convert a singlemode signal into a
multimode signal. The size of the list is the number of signal modes, with timedomain
waveforms identical except for the power ratio factor. The sum of the power values is
normalized to “1” and used to scale the timedomain signals.
A Power ratio parameter of “1 2 3” will generate “3” modes. Each mode will have
power ratio equal to 1/6, 2/6 and 3/6, respectively.
The parameter Mode polarization defines how the spatial modes are attached to the
signal polarization. The user can select whether the mode profile is attached to only
one polarization (X or Y), or to both polarizations. If attached to both polarizations, it
can be the same for both (X=Y) or unique (X and Y).
The user can provide the list of mode indexes for each polarization, as well as the
inner and outer radius for the modes.
276
DONUT TRANSVERSE MODE GENERATOR
⎧ ⎧ cos ( m φ ), n ≥ 0
⎪⎨ r inner ≤ r ≤ r outer
ψ m ( r, φ ) = ⎨ ⎩ sin ( m φ ), n < 0 (1)
⎪
⎩ 0, r < r inner, r > r outer
where m is the azimuthal index, rinner is the inner radius and router is the outer radius
for each mode.
References
[1] Mahmoud, S.W.Z.; Wiedenmann, D.; Kicherer, M.; Unold, H.; Jager, R.; Michalzik, R.; Ebeling,
K.J. "Spatial investigation of transverse mode turnon dynamics in VCSELs", IEEE Photonics
Technology Letters, Volume: 13, Issue: 11, Nov. 2001 Pages: 1152  1154.
277
DONUT TRANSVERSE MODE GENERATOR
Notes:
278
HERMITE TRANSVERSE MODE GENERATOR
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Power ratio array 1
279
HERMITE TRANSVERSE MODE GENERATOR
Name and description Default value Default unit Units Value range
Pol. X inv. radius of curvature Y 0 1/um [0, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
The Hermite Transverse Mode Generator attaches mode profiles to the input signal X
and Y polarizations. A HermiteGaussian profile [1][2] is attached to each polarization.
Additionally, singlemode inputs can be converted to a multimode signal scaled by a
userdefined power distribution.
The parameter Power ratio array is used to convert a singlemode signal into a
multimode signal. The size of the list is the number of signal modes, with timedomain
waveforms identical except for the power ratio factor. The sum of the power values is
normalized to “1” and used to scale the timedomain signals.
280
HERMITE TRANSVERSE MODE GENERATOR
A Power ratio parameter of “1 2 3” will generate “3” modes. Each mode will have
power ratio equal to 1/6, 2/6 and 3/6, respectively.
The parameter Mode polarization defines how the spatial modes are attached to the
signal polarization. The user can select whether the mode profile is attached to only
one polarization (X or Y), or to both polarizations. If attached to both polarizations, it
can be the same for both (X=Y) or unique (X and Y).
The user can provide the list of mode indexes for each polarization, as well as the spot
size and the inverse of the radius of curvature for each mode for both X and Yaxis.
2x ⎛ x2 ⎞ ⎛ πx 2 ⎞ 2y ⎛ y2 ⎞ ⎛ πy 2 ⎞
ψ m, n ( r, ϕ ) = H m ⎛ ⎞ exp ⎜ – 2⎟ exp ⎜ j ⎟ H n ⎛ ⎞ exp ⎜ – 2⎟ exp ⎜ j ⎟
(1)
⎝ w ox ⎠ ⎝ w ox ⎠ ⎝ λR ox⎠ ⎝ w oy ⎠ ⎝ w oy ⎠ ⎝ λR oy⎠
where m and n represent the X and Y index that describe the mode dependencies for
the X and Yaxis. R is the radius of curvature and w0 is the spot size. Hm and Hn are
the Hermite polynomials.
References
[1] A. E. Siegman, Lasers, University Science Books, Sausalito, CA, 1986.
[2] A. Ghatak, K. Thyagarajan, Introduction to Fiber Optics, Cambridge University Press, New
York, NY, 1998.
281
HERMITE TRANSVERSE MODE GENERATOR
Notes:
282
LAGUERRE TRANSVERSE MODE GENERATOR
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Power ratio array 1
283
LAGUERRE TRANSVERSE MODE GENERATOR
Name and description Default value Default unit Units Value range
Pol. X inv. radius of curvature 0 1/um [0, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
The Laguerre Transverse Mode Generator attaches mode profiles to the input signal
X and Y polarizations. A LaguerreGaussian profile [1][2] is attached to each
polarization. Additionally, singlemode inputs can be converted to a multimode signal
scaled by a userdefined power distribution.
The parameter Power ratio array is used to convert a singlemode signal into a
multimode signal. The size of the list is the number of signal modes, with timedomain
waveforms identical except for the power ratio factor. The sum of the power values is
normalized to “1” and used to scale the timedomain signals.
A Power ratio parameter of “1 2 3” will generate “3” modes, each mode will have
power ratio equal to 1/6, 2/6 and 3/6, respectively.
The parameter Mode polarization defines how the spatial modes are attached to the
signal polarization. The user can select whether the mode profile is attached to only
one polarization (X or Y), or to both polarizations. If attached to both polarizations, it
can be the same for both (X=Y) or unique (X and Y).
284
LAGUERRE TRANSVERSE MODE GENERATOR
The user can provide the list of mode indexes for each polarization, as well as the spot
size and the inverse of the radius of curvature for each mode.
n
⎛ 2r 2 ⎞ 2 n ⎛ 2r 2 ⎞ ⎛ r2 ⎞ ⎛ πr 2 ⎞ ⎧ sin ( n ϕ ), n ≥ 0 (1)
ψ m, n ( r, ϕ ) = ⎜ 2⎟ L m ⎜ 2⎟ exp ⎜ 2⎟ exp ⎜ j ⎟ ⎨
⎝ wo ⎠ ⎝ wo ⎠ ⎝ wo ⎠ ⎝ λR o⎠ ⎩ cos ( n ϕ ), n < 0
where m and n represent the X and Y index that describe the azimuthal and radial
indexes, respectively. R is the radius of curvature and w0 is the spot size. Ln,m is the
Laguerre polynomial. If parameter Complex is enabled, a complex mode with sin and
cos terms will be created, otherwise the output is real and depends on the signal of
parameter n.
References
[1] A. E. Siegman, Lasers, University Science Books, Sausalito, CA, 1986.
[2] A. Ghatak, K. Thyagarajan, “Introduction to Fiber Optics”, Cambridge University Press, New
York, NY, 1998.
285
LAGUERRE TRANSVERSE MODE GENERATOR
Notes:
286
MULTIMODE GENERATOR
Multimode Generator
This component attaches transverse mode profiles to the input signal. It also converts
singlemode signals into multimode signals.
Ports
Parameters
Spatial effects
Name and description Default value Default unit Units Value range
Power ratio array 1
287
MULTIMODE GENERATOR
Name and description Default value Default unit Units Value range
Pol. X inv. radius of curvature 0 1/um [0, 1e+100]
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
The Multimode Generator attaches mode profiles to the input signal X and Y
polarizations. A HermiteGaussian or a LaguerreGaussian profile [1][2] is attached to
each polarization. Additionally, singlemode inputs can be converted to a multimode
signal scaled by a userdefined power distribution.
The parameter Power ratio array is used to convert a singlemode signal into a
multimode signal. The size of the list is the number of signal modes, with timedomain
waveforms identical except for the power ratio factor. The sum of the power values is
normalized to “1” and used to scale the timedomain signals.
A Power ratio parameter of “1 2 3” will generate “3” modes, each mode will have
power ratio equal to 1/6, 2/6 and 3/6, respectively.
The parameter Mode polarization defines how the spatial modes are attached to the
signal polarization. The user can select whether the mode profile is attached to only
one polarization (X or Y), or to both polarizations. If attached to both polarizations, it
can be the same for both (X=Y) or unique (X and Y).
288
MULTIMODE GENERATOR
The user can provide the list of mode indexes for each polarization, as well as the spot
size and the inverse of the radius of curvature for each mode.
Refer to the Laguerre Transverse Mode Generator component for the analytical
representation of the LaguerreGaussian profile.
For the HermiteGaussian profile, the Multimode Generator assumes the same
values for the spot size and radius of curvature for the X and Yaxis.
Refer to the Hermite Transverse Mode Generator component for the analytical
representation of the HermiteGaussian profile.
References
[1] A. E. Siegman, “Lasers”, University Science Books, Sausalito, CA, 1986.
[2] A. Ghatak, K. Thyagarajan, “Introduction to Fiber Optics”, Cambridge University Press, New
York, NY, 1998.
289
MULTIMODE GENERATOR
Notes:
290
MEASURED TRANSVERSE MODE
This component attaches measured transverse mode profiles to the input signal. The
measured profiles are loaded from a file using the BCF3DCX format. It also converts
singlemode signals into multimode signals
Ports
Parameters
Spatial effects
Name and description Default value Default unit Units Value range
Power ratio array 1
Pol. X files ““
Pol. Y files ““
291
MEASURED TRANSVERSE MODE
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
The measured transverse mode generator attaches mode profiles to the input signal
X and Y polarizations. A transverse mesh from a file is attached to each polarization,
additionally; singlemode inputs can be converted to a multimode signal scaled by a
user defined power distribution.
The parameter Power ratio array is used to convert a singlemode signal into a
multimode signal. The size of the list is the number of signal modes, with timedomain
waveforms identical except for the power ratio factor. The sum of the power values is
normalized to “1” and used to scale the timedomain signals.
A Power ratio parameter of “1 2 3” will generate “3” modes, each mode will have
power ratio equal to 1/6, 2/6 and 3/6, respectively.
The parameter Mode polarization defines how the spatial modes are attached to the
signal polarization. The user can select whether the mode profile is attached to only
one polarization (X or Y), or to both polarizations. If attached to both polarizations, it
can be the same for both (X=Y) or unique (X and Y).
The user can provide the list of filed for each polarization using the parameters Pol. X
files and Pol. Y files. For each power ratio a filename must be provided. Different from
other OptiSystem components, the measured transverse mode generator will reload
the files every time it calculates. This means the files must exist or an error message
will be generated during loading.
A Power ratio parameter of '1 2 3' will generate '3' modes and the parameter Pol. X
files should have three lines; each line will have the file name of a mode. For example:
Mode_X_1_1.f3d
Mode_X_2_1.f3d
Mode_X_3_1.f3d
The files should have the complex data file format BCF3DCX. Files that follow this
format are generated from the Save Transverse Mode component from OptiSystem
or the output files in BPM 3D.
292
MEASURED TRANSVERSE MODE
. . .
. . .
293
MEASURED TRANSVERSE MODE
BCF3DCX
100 100
1.000000E+001 1.100000E+001
4.582487025358980E004, 2.411965546811583E002
1.813879122411751E004, 2.322439514101689E002
8.864140535377826E004, 2.245463661588051E002
. . .
1.004141897700716E002, 7.709994296904761E003
9.736326254112302E003, 8.732395427319460E003
9.270032367315658E003, 9.686774052240091E003
294
MODE ID MODIFIER
Mode ID Modifier
This component changes the label identifier of a transverse mode. Transverse modes
with different labels identifiers are considered orthogonal.
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
User defined label NO [YES, NO]
Label ID ““
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Technical Background
This component changes the label identifier of a transverse mode. Transverse modes
with different labels identifiers are considered orthogonal. Multimode signals with
295
MODE ID MODIFIER
transverse modes with the same label identifier are not considered orthogonal. By
using this component to change the label identifier the modes are then considered
orthogonal.
296
Optical Fibers Library
• Optical fiber
• Optical fiber CWDM
• Bidirectional Optical Fiber
• Nonlinear Dispersive Fiber (Obsolete)
• Linear Multimode Fiber
• ParabolicIndex Multimode Fiber
• MeasuredIndex Multimode Fiber
297
Notes:
298
OPTICAL FIBER
Optical fiber
The optical fiber component simulates the propagation of an optical field in a single
mode fiber with the dispersive and nonlinear effects taken into account by a direct
numerical integration of the modified nonlinear Schrödinger (NLS) equation (when the
scalar case is considered) and a system of two, coupled NLS equations when the
polarization state of the signal is arbitrary. The optical sampled signals reside in a
single frequency band, hence the name total field [1]. The parameterized signals and
noise bins are only attenuated.
Ports
Parameters
Main
Fiber length
299
OPTICAL FIBER
Dispersion
300
OPTICAL FIBER
∂β 1 ∂D
D = , S =  (wavelength domain definition)
∂λ ∂λ
and
∂β 1 ∂β 2
β 2 = , β 3 =  (frequency domain definition)
∂ω ∂ω
of the dispersion parameters, but not to the argument of these functions, which is
always assumed to be the wavelength. All the parameters in the component
(including β 2 and β 3 ) are given as functions of wavelength (not frequency). This is
also the case when β 1 or β 2 are specified from a file  the first column of the file
contains wavelength values ( λ ) and the second column  the corresponding values
of β 1 ( λ ) or β 2 ( λ ) .
PMD
301
OPTICAL FIBER
Nonlinearities
302
OPTICAL FIBER
t
τ R1 = ( dImχ 1111 ( ω ) ⁄ dω ) ω = 0
α f = Re ( χ 1122 ( ω = 0 ) )
303
OPTICAL FIBER
Numerical
304
OPTICAL FIBER
Graphs
Note: The rest of the parameters in the Graphs tab of the component determine
which graphs are plotted after the simulation is complete.
305
OPTICAL FIBER
Simulation
Noise
Random numbers
306
OPTICAL FIBER
Technical Background
Scalar approach
Basic equation
When the optical field is assumed to maintain its polarization along the fiber length,
the evolution of a slowly varying electric field envelope can be described by a single
nonlinear Schrödinger (NLS) [2] equation (the scalar approach, Model type parameter
from the "Numerical" tab is set to "Scalar") of the form:
∂E β2 ( ω0 ) ∂2 E β3 ( ω0 ) ∂3 E ⎛ i ∂ ∂E ⎞
2
  –   = iγ ⎜ E 2 E +   ( E 2 E ) – ρτ R1 E ⎟
(1)
 + αE + i 
∂z 2 ∂T 2 6 ∂T 3 ⎝ ω 0 ∂T ∂T ⎠
307
OPTICAL FIBER
relations are used internally to convert between them and the commonly used
wavelength domain parameters D (dispersion) and S (dispersion slope).
dβ 1 2πc
D =  = – 
2 2
β
dλ λ (2)
λ 2 2 dD
β 3 = ⎛ ⎞ ( λ S + 2λD ), S = 
⎝ 2πc⎠ dλ
ω0 n2
γ =  (3)
cA eff
In Equation 3, n 2 is the nonlinear refractive index coefficient and A eff is the fiber
effective area. The first term in the righthand side in Equation 1 accounts for the self
phase modulation effect. It is responsible for the broadening of the pulse spectra and,
in the presence of anomalous GVD, for the formation of optical solitons (See "Self
phase modulation" and "Selfphase modulation and group velocity dispersion" from
the Tutorials). The second term in the righthand side of Equation 1 takes into account
the selfsteepening effect. It leads to an asymmetry in the SPMbroadened spectra of
ultrashort (femtosecond) pulses [2] and is responsible for the formation of optical
shocks (see "Selfsteepening" in the Tutorials). This effect will be taken into account
only if the "Full Raman response" parameter is set to False. The last term in
Equation 1 accounts for the intrapulse Raman scattering effect with the parameter
τ R1 being the parallel Raman selfshift time. The intrapulse Raman scattering is an
approximation to the actual Raman response of the material which is valid provided
that signal spectrum is narrow compared to the Ramangain spectrum. The τ R
parameter is related to the slope of the imaginary part of the Raman susceptibility
Im ( χ 1111 ( ω ) ) at zero frequency offset [2]. The parameter ρ is the fractional
contribution of the delayed response of the material to the total nonlinearity [2]. The
intrapulse Raman scattering effect is responsible for the selffrequency shift i.e.
energy transfer from higher to lower spectral components. It leads to a decay of higher
order solitons into its constituents (see "Intrapulse Raman scattering" in the Tutorials).
The intrapulse Raman scattering plays the most important role among the higher
order nonlinear effects [2].
(4)
P TOT B TOT L E < 9mWTHzMm ,
308
OPTICAL FIBER
where L E ≈ z ⁄ ( L amp α ) is the total effective length, α is the fiber loss, L amp is the
amplifier spacing, z the link length, P TOT is the total optical power, and B TOT is the
total optical bandwidth.
⎛ ∞ ⎞
∂E iβ 2 ( ω 0 ) ∂ 2 E β 3 ( ω 0 ) ∂ 3 E ⎜ 2 2 ⎟ (4a)
 + αE +   –   = iγ ( 1 – ρ ) E E + ρE ∫ h 1111 ( s ) E ( T – s ) ds
∂z 2 2 6 ∂T 3 ⎜ ⎟
∂T ⎝ ⎠
0
Numerical solution
In dimensionless form, Equation 1 reduces to:
2 3 2
∂U ∂ U 2 ∂ U ∂U ∂ 2 (5)
i  + D 2  + N 1 U U = iD 3  + N 2 U  – iN 3  ( U U ) – iAU ,
∂ξ ∂t
2
∂t
3 ∂t ∂t
2 3
T0 1 T0 1 τR (7)
LD = , L NL = , L D' = , s = , τ R' = , E = P 0 U, T = T 0 t, z = ξL D
β2 γP 0 β3 ω0 T0 To
In Equation 7,T 0 is the time window size and P 0 is the maximum (over the time
2
window) of the electric field intensity E ( z = 0, T ) .
309
OPTICAL FIBER
The symmetrized splitstep Fourier method [2, 4] is used to solve Equation 5. The
solution is advanced from ξ to ξ + h ( h is the stepsize, related to the value of the
NL 2
Max. nonlinear phase shift parameter ϕ max = max ( U h ) ) according to:
⎛ (ξ + h) ⎞
h ⎞ h
⎛
U ( ξ + h, t ) = exp ⎝ D̂⎠ exp ∫ N̂ ( ξ' ) dξ'⎟ exp ⎛  D̂⎞ U ( ξ, t ) ,
⎜ (8)
2 ⎜ ⎟ ⎝ 2 ⎠
⎝ ξ ⎠
2 3
∂ ∂ (9)
D̂ = iD 2 2 + D 3 3 – A
∂t ∂t
and
2 2
∂U ∂U ∂U
N̂ = iN 1 U – iN 2  – N 3 ⎛  + U∗  ⎞
2 (10)
∂t ⎝ ∂t ∂t ⎠
The different options available from the "Numerical" tab specify the details of the
implementation of Equation 8 and Equation 10 (see Figure 1). The simplest (and the
fastest) implementation corresponds to "Propagator type" set to "Exponential" and
"Calculation type" set to "Noniterative". In this case, the following approximation is
used:
ξ+h
∫
(11)
N̂ ( ξ' ) dz' ≈ hN̂ ( exp [ ( h ⁄ 2 )D̂ ] U ( ξ, t ) ) .
ξ
310
OPTICAL FIBER
According to Equation 11, the halfstep propagated field, with the nonlinear effects
ignored, is used in turn to evaluate the nonlinearity operator. The dispersion operator
is evaluated in the frequency domain according to:
h h
Ũ D ⎛ ξ + ⎞ = FFT exp ⎛  D̂ ( iω )⎞ FFT [ U ( ξ, t ) ] ,
–1 (12)
⎝ 2 ⎠ ⎝ 2 ⎠
where FFT means fast Fourier transform. If, in addition the "Step size" option is set
to "Constant" ("Propagator type", "Exponential", and "Calculation type" are set to
"Noniterative"), the number of operations per step decreases because the first and
the last Fourier transform for each step cancels each other out (dispersion operators
combine) (see Equation 13).
⎛(ξ + h) ⎞ ⎛ (ξ + h) ⎞
h h h h
U ( ξ + 2h, t ) = exp ⎛  D̂⎞ exp ⎜ ∫ N̂ ( ξ' ) dξ'⎟ exp ⎛  D̂⎞ exp ⎛  D̂⎞ exp ⎜ ∫ N̂ ( ξ' ) dξ'⎟ exp ⎛  D̂⎞ U ( ξ, t ) =
⎝2 ⎠ ⎜ ⎟ ⎝2 ⎠ ⎝2 ⎠ ⎜ ⎟ ⎝2 ⎠
⎝ ξ ⎠ ⎝ ξ ⎠
(13)
(ξ + h) (ξ + h)
⎛ ⎞ ⎛ ⎞
h h
exp ⎛  D̂⎞ exp ⎜⎜ ∫ N̂ ( ξ' ) dξ'⎟⎟ exp ( hD̂ ) exp ⎜⎜ ∫ N̂ ( ξ' ) dξ'⎟⎟ exp ⎛⎝  D̂⎞⎠ U ( ξ, t )
⎝2 ⎠ 2
⎝ ξ ⎠ ⎝ ξ ⎠
311
OPTICAL FIBER
When the "Propagator Type" is set to "RungeKutta 4th order" (or "RungeKutta 2nd
order") (RK4 or RK2), the exponent with the nonlinearity operator in Equation 8 is
replaced by the direct integration of the following system of coupled ordinary
differential equations:
⎛ ∂U
⎞ = N̂U (14)
⎝ ∂z ⎠ NL
by means of the standard RK4 (or RK2) routine (see example in [3]). The application
of the dispersion operator is the same.
ξ+h
h
∫
(15)
N̂ ( ξ' ) dξ' ≈  ( N ( ξ ) + N ( ξ + h ) )
2
ξ
312
OPTICAL FIBER
2
Figure 2 Evolution of E ( ξ, t = 0 ) for N=3 soliton over 15 soliton periods with different calculation
modes
NL
Note: In the three cases presented, ϕ max = 27.6mrad , constant step size.
A comparison between the "Iterative" and "Noniterative" approaches is presented in
Figure 2. Evolution of N=3 soliton over 15 soliton periods is presented. The "Step
size" is kept "Constant" with the "Max. nonlinear phase shift" parameter is equal to
27.6. mrad. The noniterative approach is the fastest but not accurate enough at this
step size. The development of spurious, numerical instability, which breaks the
periodicity of the soliton evolution [2], is evident at the end of the run. For the same
step size the iterative implementation of the splitstep Fourier method suppresses the
instability, thus improving the quality of the results, however this improvement is at the
expense of increased computation time.
The step size h in the component is determined through the value of the parameter
NL 2
ϕ max = γmax ( E )h . In the case of the constant step size calculation, it is
calculated once, using the input signal to obtain the maximum value of the intensity.
In the case of variable step size calculation such an evaluation is performed at each
step.
313
OPTICAL FIBER
NL NL
Figure 3 Variable step size, value of ϕ max is ϕ max = 50mrad
In Figure 3, the calculation presented in Figure 2 is repeated using variable step size.
This calculation takes longer in comparison to the "Noniterative" case presented in
Figure 2, but less than in the case where two iterations are used. Depending on the
behavior of the solution, variable step size calculation can take less time compared to
the constant step size, although the fixed step size calculation performs a smaller
number of operations per step (see Equation 13). In the presence of considerable
attenuation, the importance of nonlinear effects decreases along the fiber length,
which would permit the use of a larger step size. In this case, the use of variable step
size will reduce the computation time. The variable step size calculation is more
NL
flexible, because different tasks can be handled keeping the value of ϕ max constant.
For the case presented in Figure 3, this value is double the size of the one used in
Figure 2, but the results are even better (refer to compare with Figure 2,
"Noniterative").
The splitstep scheme used in the model is locally second order accurate which
3
means that the local error is proportional to the h . However, the global error (after N
3 2
steps) is proportional to Nh = Lh [22]. Thus, increasing the fiber length might
require decrease of the step size to maintain the same accuracy.
The use of FFT implies periodic boundary conditions. In some cases a part of the
pulse energy may spread eventually hitting the time window boundaries. When the
energy reaches one of the edges of the time window it automatically reenters from the
other edge perturbing the solution. This can be avoided using the absorbing type of
boundary conditions. To achieve this at each step the optical field is multiplied in the
time domain [10] by:
where t edge indicates the nearest edge. The effect of periodic and absorbing
boundary conditions is shown in Figure 4 where the results presented in Figure 3 from
314
OPTICAL FIBER
Figure 4 Periodic (left plot) and absorbing with filter steepness 0.05 (right plot) boundary conditions
315
OPTICAL FIBER
Vector approach
When the polarization state of the incident light is not preserved during its propagation
inside an optical fiber the scalar approach is no longer applicable and Equation 1 is
replaced by [2], [6]  [10]:
2 3
∂E X ∂E iβ ∂ E β ∂ E
 + β 1X X + 2 X – 3 X = iγ ( 1 – ρ ) ⎛ E X 2 + 2 E Y 2⎞ E X
∂z ∂t 2 ∂t 2 6 ∂t 3 ⎝ 3 ⎠
∞ ∞
2 2
+ iγρE X ∫ h1111 ( s ) EX ( t – s ) ds + ∫ h 1122 ( s ) E Y ( t – s ) ds
0 0
∞
2 3
∂E Y ∂E iβ ∂ E β ∂ E
 + β 1X Y + 2 Y – 3 Y = iγ ( 1 – ρ ) ⎛ E Y 2 + 2 E X 2⎞ E Y
∂z ∂t 2 ∂t 2 6 ∂t 3 ⎝ 3 ⎠
∞ ∞
2 2
+ iγρE Y ∫ h1111 ( s ) Eγ ( t – s ) ds + ∫ h 1122 ( s ) E X ( t – s ) ds
0 0
∞
+ iγρE X ∫ h 1212 ( s )E γ t – s E∗ X ( t – s ) ds
0
Equation 17, h ijkl ( t ) contains the Raman response functions [6], [18]. Their Fourier
transformations and Raman susceptibilities χ ijkl ( v ) , are shown in Figure 4.1. The
convolution integrals in Equation 17 are evaluated in the frequency domain, by
multiplying the spectra of the electric fields with the Raman susceptibilities and then
performing the inverse FFT.
316
OPTICAL FIBER
2 3
∂E ∂E iβ ∂ E β ∂ E
X + β 1X X + 2 X – 3 X =
∂z ∂t 2 ∂t 2 6 ∂t 3
2 2
2 1 + αf ∂ EX ∂ EY
+ ⎛⎝ ( 1 – ρ ) + ρ ⎞⎠ E Y – ρτ R1  – ρτ R2  E X
2 2
iγ E X
3 2 ∂t ∂t
τ R1 – τ R2 ∂ ( E X E Y∗ )
– iγρ   E Y
2 ∂t
(17a)
2 3
∂E ∂E iβ ∂ E β ∂ E
Y + β 1Y Y + 2 Y – 3 Y =
∂z ∂t 2 ∂t 2 6 ∂t 3
2 2
2 1 + αf 2 ∂ EY ∂ EX
+ ⎛⎝ ( 1 – ρ ) + σ ⎞⎠ E X – στ R1  – ρ τ R2  E Y
2
iγ E Y
3 2 ∂t ∂t
τ R1 – τ R2 ∂ ( E Y E X∗ )
– iγρ   E X
2 ∂t
Note: In the case of Equation 17 or Equation 17a, due to the orthogonal Raman
gain terms (the last sections in Equation 17 or Equation 17a), the "Exponential"
option for the "Propagator type" is not applicable. The component automatically
selects "Runge Kutta 2nd order" when the model type is set to "Vector", and the
Raman effect ("Intrapulse Raman scattering" or "Full Raman response" options
317
OPTICAL FIBER
2 3
∂u ∂u ∂ u ∂ u 2 2 2
i ⎛⎝  + δ ⎞⎠ + D 2 2 – iD 3 3 + N 1 ⎛⎝ u +  v ⎞⎠ u = 0
∂ξ ∂t ∂τ ∂τ 3
(18)
2 3
∂v ∂v ∂ v ∂ v 2 2 2
i ⎛⎝  + δ ⎞⎠ + D 2 2 – iD 3 3 + N 1 ⎛⎝ v +  u ⎞⎠ v = 0
∂ξ ∂t ∂τ ∂τ 3
The quantities β 1X and β 1Y are the inverse group velocities for the X and Y
polarization components respectively.
The "coarsestep method" [11] is used to simulate the PMD effects in the "Stochastic"
mode. The fiber is represented by a concatenation of trunks and the propagation of
light in each trunk is simulated by the splitstep Fourier method described in the
previous section. The lengths of the trunks are random numbers with a Gaussian
318
OPTICAL FIBER
distribution [12]. The average and the dispersion of this distribution are the "Scattering
section length" L scatt and "Scattering section dispersion" σ scatt parameters:
i 2
i 1 – ( L scatt – L scatt ) (19)
f ( L scatt ) =  exp 
2πσ scatt 2
2σ scatt
It is recommended [12] that the dispersion is 20% of the average value. The
birefringence of each trunk is given by [11] (see the related PMD examples in the
tutorials):
d DP
 ( Δβ ) =  (20)
dω i
L scatt
where D PMD is the PMD coefficient. The principal axes of the trunks are randomly
oriented with respect to each other (see Figure 4). To simulate the random mode
coupling at the end of each trunk the following transformation is applied [11], [13]:
In Equation 20, α and ϕ are random numbers uniformly distributed in the interval
[ 0, 2π ] .
Wavelength dependent parameters
The file that specifies the wavelength dependence of the parameters consists of two
columns with the left column being the wavelength in nanometers and the right
column containing the corresponding values of the parameters (see Table 1 ). The
sampling interval is not necessarily be constant. The parameter values must be given
in the units specified in the "Units" tab of the table.
319
OPTICAL FIBER
λ [ nm ] α [ dB ⁄ km ]
1400 0.31405
1402.5 0.30246
1405 0.29276
1407.5 0.28457
1410 0.27757
1412.5 0.27153
The values of the parameters in Equation 1 and Equation 17 are evaluated at the
reference wavelength.
Note: The reference wavelength must be within the wavelength interval covered
by the files for all the wavelength dependent parameters specified.
The reference wavelength can be either userspecified or "automatic". In the last case
the wavelength corresponding to the central frequency of the spectrum of the signal
is assumed by the component to be the reference wavelength. Linear interpolation is
used to calculate the values of the attenuation, effective area and n 2 parameters at
this wavelength. For the dispersion parameters the following procedure is used. The
wavelength dependence specified by the file is fitted internally using the fiveterm
Sellmeier formula [14]. The higherorder dispersion parameters are then obtained by
analytically differentiating this expression. If the option frequency domain parameter
is unchecked, the file may give either the group delay β 1 ( λ ) or dispersion D ( λ )
(depending on the choice made in the "Dispersion file format" tab), and if the
frequency domain parameters option is selected, either β 1 ( λ ) or β 2 ( λ ) can be
supplied, again determined by the value of the "Dispersion file format" parameter. If
the wavelength dependence of the group delay is given by the user, two successive
differentiations are applied to its Sellmeier fit. Differentiating the analytical fit instead
of using a direct numerical differentiation of the data provides the advantage of being
able to produce reasonable results even in the case where the supplied data is noisy
(see Appendix 1).
Note: The accuracy of the Sellmeier fit depends on the type of the fiber. This is
shown in Figure 6, where the results obtained for dispersion flattened and
dispersion shifted fibers are shown.
320
OPTICAL FIBER
Figure 6 Comparison between the original dispersion data and their fits for two fiber types
To avoid the aliasing phenomena (see e.g. [3]), the sample rate is chosen to be at
least three times bigger (Figure 7) than the bandwidth occupied by the simulated
channels (see e.g. [15]).
Any frequency component outside the frequency range (FcSR/2, Fc+SR/2), where
SR is the sample rate and Fc is the reference frequency is falsely translated (aliased)
into that range by the very act of discrete sampling [3]. If the sample rate is bigger than
the bandwidth occupied by the WDM channels (so it can accommodate all the
channels) but less than three times that value in the presence of nonlinear effect the
fourwave mixing products resulting from the nonlinear interaction between the
channels (spurious waves [16]) will be aliased. In [16], to minimize the amount of
321
OPTICAL FIBER
aliased power the requirement that the value of the power spectrum at the boundary
of the available spectral range be 40 dB of its peak value is used.
The longitudinal step size depends on the importance of the nonlinear effects for the
particular simulation. If all the nonlinear effects are disabled step size equal to the
fiber length will be used. The increase of the impact of nonlinearity will require
decrease of the step size (decrease of the value of the max. nonlinear phase shift
parameter) to maintain the same accuracy.
NL NL
Figure 8 Output spectra corresponding to ϕ max = 50mrad and ϕ max = 3mrad
Values in the order of a few milliradians (one [15] and three [17])) are used with this
parameter in a WDM system simulation. The effect of an improperly chosen step size
is shown in Figure 8, where the output spectra corresponding to an interaction of two
Gaussian pulses with carrier wavelengths one nm spaced are shown (see "Cross
phase modulation" from the Tutorials). While the correct result that the fourwave
mixing products (or spurious waves) should disappear when the pulses are no longer
overlapped (in the absence of any loss and gain [16]) is reached when the stepsize
is small enough, in the opposite case, the spurious frequencies present in the output
spectra are still evident. The improperly chosen step size (too big) tends to
exaggerate the four wave mixing products (see [22] and references therein).
322
OPTICAL FIBER
Appendix 1
Dispersion fitting according to the Sellmeier formula
When the option "Dispersion from file” is selected, the dispersion data are internally
fitted according to the fiveterm Sellmeier formula [14], namely:
–4 –2 2 4 (1)A
τ = c1 λ + c2 λ + c3 + c4 λ + c5 λ
where τ is the group delay (per unit fiber length) or, respectively:
dτ –5 –3 3 (2)A
D =  = c 1'λ + c 2'λ + c 4'λ + c 5'λ
dλ
where D is the dispersion [ps/nm/km]. The user supplies data either for the
dispersion or the group delay that are then fitted according to Equation 2A or
Equation 1A, and the slope and/or dispersion are calculated by differentiating
Equation 1A and Equation 2A analytically.
N
–5 –3 3 2
∑ ( c1λ i + c 2λ i + c 4λ i + c 5λ i – D i ) = min (3)A
Q =
i=1
∂Q
 = 0, i = 1…4 , (4)A
∂c i
323
OPTICAL FIBER
– 10 –8 –4 –2 –5
∑ λi ∑ λi ∑ λi ∑ λi C1 ∑ Diλ i
–8 –6 –2 –3
∑ λi ∑ λi ∑ λi N C2
=
∑ Diλ i (5)A
–4 –2 2 4 C4
∑ λi ∑ λi ∑ λi ∑ λi ∑ Diλ i
–2 4 6 C5 3
∑ λi N ∑ λi ∑ λi ∑ Diλ i
In the case when the user supplies a group delay data file, Equation 1A is used and
Equation 5A transforms into Equation 6A.
The fitting procedure is useful when/if noisy data is supplied by the user, as the
following example shows. Figure 1A shows dispersionversuswavelength
dependence of SMF28 and the corresponding "exact" results for dispersion
parameters are displayed below the graph.
–8 –6 –4 –2 –4
∑ λi ∑ λi ∑ λi ∑ λi N
C1
∑ τiλ i
–6 –4 –2 2 –2
∑ λi ∑ λi ∑ λi N ∑ λi C2 ∑ τiλ i
–4 –2 2 4 (6)A
∑ λi ∑ λi N ∑ λi ∑ λi C3 = ∑ τi
–2 2 4 6 C4 2
∑ λi N ∑ λi ∑ λi ∑ λi C5
∑ τiλ i
2 4 6 8 4
N ∑ λi ∑ λi ∑ λi ∑ λi ∑ τiλ i
324
OPTICAL FIBER
To assess the influence of noise on the results from the calculation some noise is
added to the data presented in Figure 1A with the resulting graph presented in Figure
2A. Supplying the data from Figure 2A to the Nonlinear Dispersive Fiber Total Field
component gives the results for the dispersion parameters presented under Figure
2A.
325
OPTICAL FIBER
326
OPTICAL FIBER
Appendix 2
Optical fiber data
SMF28
The SMF28 model used in OptiSystem has the following characteristics:
Figure 1 Attenuation
327
OPTICAL FIBER
328
OPTICAL FIBER
+D NZDSF model
The +D NZDSF model used in OptiSystem has the following characteristics:
Figure 5 Attenuation
329
OPTICAL FIBER
330
OPTICAL FIBER
D NZDSF model
The D NZDSF model used in OptiSystem has the following characteristics:
Figure 9 Attenuation
331
OPTICAL FIBER
332
OPTICAL FIBER
CDF (Standard)
The DCF model used in OptiSystem has the following characteristics:
Figure 13 Attenuation
333
OPTICAL FIBER
334
OPTICAL FIBER
References
[1] G. P. Agrawal, "Applications of nonlinear fiber optics", Academic press, 3rd edition, 2001.
[2] G. P. Agrawal, "Nonlinear fiber optics", Academic press, 3rd edition, 2001.
[3] W. H. Press, et al., "Numerical Recipes: The Art of Scientific Computing", 2nd Edition,
Cambridge University Press, 1992.
[4] M. Lax, J. H. Batteh and G. P. Agrawal, Journ. Appl. Phys. 52 , 109, (1981).
[5] F. Matera and M. Settembre, Journ. Lightwave Technol. 14, 1 (1996).
[6] R. W. Hellwarth, Prog. Quant. Electr. 5, 1 (1977).
[7] E. A. Golovchenko and A. N. Pilipetskii, JOSA B, 11, 92 (1994).
[8] P. T. Dinda, G. Millot, and S. Wabnitz JOSA B, 15, 1433 (1998).
[9] C. R. Menyuk, Opt. Lett., 12, p. 614 (1987).
[10] C. R. Menyuk, JOSA B, 5, p. 392(1988).
[11] D. Marcuse, C. R. Menyuk and P. K. A. Wai JLT, vol. 15, No. 9, pp. 1735 (1997).
[12] C. H. Prola Jr., J. A. Pereira da Silva, A. O. Dal Forno, R. Passy, J. P. Von der Weid, and N.
Gisin IEEE Phot. Technol. Letters, 9, No. 6, 842 (1997).
[13] P. K. A. Wai, C. R. Menyuk, and H. H. Chen , Opt. Lett. 16 1231 (1991).
[14] L. G. Cohen, Journ. Lightwave Technol. 3, 958, (1985).
[15] M. I. Hayee and A. E. Willner, IEEE Phot. Technol. Lett. 11, No. 8, (1999).
[16] D. Marcuse, A. R. Chraplyvy, and R. W. Tkach, Journ. Lightwave Technol, 9, 121 (1991).
[17] R. W. Tkach, A. R. Chraplyvy, F. Forghieri, A. H. Gnauck, and R. M. Derosier, Journ. Lightwave
Technol, 13, 841 (1995).
[18] P. Tchofo Dinda, G. Millot, and S. Wabnitz, JOSA B, 15, 1433, (1998).
[19] R.H.Stolen, J.P.Gordon, W.J. Tomlinson and H.A. Haus, JOSA B, 6, 1159 (1989).
[20] C.R.Menyuk, M.N.Islam and J.P.Gordon, Optics Letters, 16 566, (1991).
[21] K.J. Blow and D. Wood, IEEE J. Quant. Electr., 25, 2665, (1989).
[22] O. Sinkin, R. Holzlohner, J. Zweck and C. R. Menyuk, Journ Lightwave Technol. 21, 61 (2003).
335
OPTICAL FIBER
Notes:
336
OPTICAL FIBER CWDM
Ports
337
OPTICAL FIBER CWDM
Parameters
Main
Fiber length
338
OPTICAL FIBER CWDM
Attenuation vs wavelength — — — —
If "Attenuation data type" is set to "From file" this
field specifies the file containing the attenuation
data. In this case the attenuation effect is
wavelength dependent for all types of signals 
values corresponding to the central frequency of
each signal will be calculated by linear
interpolation and used internally.
Dispersion
339
OPTICAL FIBER CWDM
PMD
The notation "Frequency domain parameters" refers to the alternative definitions
∂β 1 ∂D
D = , S =  (wavelength domain definition) and
∂λ ∂λ
∂β 1 ∂β 2
β 2 = , β 3 =  (frequency domain definition).
∂ω ∂ω
However, the format of the file specifying the wavelength dependence of the
dispersion is the following: the first column of the file contains wavelength values (λ) ,
and the second column, the corresponding values of β 1 ( λ ) . Consequently,
"Frequency domain parameters" is meaningless when the dispersion is specified
from a file.
340
OPTICAL FIBER CWDM
In the case when more than one sampled signal (separated channels) is propagating
in the fiber, the PMDinduced penalties will be the same for all channels. Different
penalties (and Qfactors) will be obtained if all the channels are merged (total field
approach). Birefringence and random mode coupling does not affect parameterized
signals and noise bins.
Nonlinearities
341
OPTICAL FIBER CWDM
n2 vs wavelength — — — —
If the "n2 data type" is set to "From file" then this tab
specifies the file containing the nonlinear index of
refraction wavelength data.
342
OPTICAL FIBER CWDM
343
OPTICAL FIBER CWDM
Numerical
344
OPTICAL FIBER CWDM
Graphs.
345
OPTICAL FIBER CWDM
Simulation
Noise
346
OPTICAL FIBER CWDM
Random numbers
Technical Background
Scalar approach
Signal propagation equations with Interband Raman scattering
When the optical field is assumed to maintain its polarization along the fiber length (so
called scalar approach, Model type parameter from the "Numerical" tab is set to
"Scalar") the evolution of the slowly varying electric field envelopes { E i } of a set of
sampled signals (SS), powers { P l } of another set of parameterized signals (PS) and
powers { N m } of a third set of noise bins (NB) is governed by the set (1) of equations.
The subsystem (1a) consists of Number of SS (the total count of sampled signals)
coupled nonlinear Schrödinger (NLS) [1], [2], (1b) contains Number of PS equations
(the total count of PS) and (1c)  Number of NB (the total count of NB) equations.
347
OPTICAL FIBER CWDM
2 3
∂E i ∂E iβ 2 ( ω i ) ∂ E i β 3 ( ω i ) ∂ E i
 + ( β l ( ω i ) – β l ( ω 0 ) ) i + α ( ω i )E i + 
  –   =
∂z ∂T 2 ∂T
2 6 ∂T 3
Number of SS Number of PS
2 2
∑ ∑
(1a)
(2 – ρ) Ek – ( 1 – ρ ) Ei + ( 2 – ρ ) +
iγ i k=1 l=1 Ei
Number of SS Number of PS
( SS ) 2 ( PS )
ρ ∑ R ik Ek + ρ ∑ R il Pl
k=1 l=1
⎛ Number of PS ⎞
⎜ ( PP ) ⎟
⎜ ∑ Rlh Ph + ⎟
dP l ⎜ h=1 ⎟ (1b)
 = – 2α l P l + 2ργ l P l Im ⎜ ⎟
dz ⎜ Number of SS Time window
⎟
1
⎜  ( SP ) 2 ⎟
⎜ Time window
 ∑ Rli ∫ Ei t dt⎟
⎝ i=l 0 ⎠
⎛ Number of PS ⎞
⎜ ( PP ) ⎟
⎜ ∑ Rmh Ph + ⎟
dN m ⎜ h=1 ⎟ (1c)
 = – 2α m N m + 2ργ m N m Im ⎜ ⎟
dz ⎜ Number of SS Time window
⎟
1
⎜  ( SN ) 2 ⎟
⎜ Time window ∑ R mi ∫ i E t d t
⎟
⎝ i=l 0 ⎠
348
OPTICAL FIBER CWDM
⎧
⎪ R
= ⎨ χ 1111 ( f i – f k ), i ≠ k 1 ≤ i ≤ Number of SS, 1 ≤ k ≤ Number of SS
( SS ) (2a)
R ik
⎪ 0, i=k
⎩
⎧ R
( PS ) ⎪ χ ( f – f ), f ≠ f
R il = ⎨ 1111 i l i l 1 ≤ i ≤ Number of SS, 1 ≤ l ≤ Number of PS (2b)
⎪ 0, fi = fl
⎩
⎧
⎪ R
= ⎨ χ 1111 ( f l – f h ), l ≠ h 1 ≤ l ≤ Number of PS, 1 ≤ h ≤ Number of PS
( PP ) (2c)
R lh
⎪ 0, l=h
⎩
⎧ R
( SP ) ⎪ χ ( f – f ), f ≠ f
R li = ⎨ 1111 l i l i 1 ≤ l ≤ Number of PS, 1 ≤ i ≤ Number of SS (2d)
⎪ 0, fl = fi
⎩
⎧ R
( PN ) ⎪ χ ( f – f ), f m ≠ f h
R mh = ⎨ 1111 m h 1 ≤ m ≤ Number of NB, 1 ≤ h ≤ Number of PS (2e)
⎪ 0, fm = fh
⎩
349
OPTICAL FIBER CWDM
⎧ R
( SN ) ⎪ χ ( f – f ), f ≠ f
R mi = ⎨ 1111 m i m i 1 ≤ m ≤ Number of NB, 1 ≤ h ≤ Number of SS (2f)
⎪ 0, fm = fi
⎩
Raman susceptibility for fused quartz is shown in Figure 1. It should be noted that
R R ∗
χ 1111 ( – ω ) = ( χ 1111 ( ω ) ) , where "*" means complex conjugation.
350
OPTICAL FIBER CWDM
The nonlinear coefficients for every SS, NB, or PS in (1) are defined according to:
ωj n2 ( ωj )
γ j =  (3)
cA eff ( ω j )
The meaning of the terms on the lefthand side of the subsystem (1a) is the same as
in the total field approach fiber model (see the technical description of this
component). The first two terms in the right hand side of (1a) give the SPM and XPM
contributions of the remaining sampled signals. The third term is the XPM contribution
of the PS. The fourth and the fifth term describe the SRS induced interactions
between the i th sampled signal and rest of the sampled signals and with the
parameterized signals, respectively.
Subsystems (1b) and (1c) describe the power balance of the set of PS and NB
respectively. These are obtained by replacing the NLS equations for NB and PS with
the timeaveraged versions of their power conservation laws. In the absence of
attenuation the total number of photons is conserved as (1) shows. The first terms in
the righthand sides of (1b) and (1c) take into account the attenuation effects. The
second and the third terms in the righthand side of (1b) describe the SRS induced
power transfer between the l th PS and the rest of the PS and between the l th PS
and the SS respectively. The second and the third terms in the righthand side of (1c)
are responsible for the SRSinduced interactions between noise bins and PS and
noise bins and SS. Note that in describing the interactions through SRS between NB
and SS and NB and PS the power of the noise bins is neglected with respect to that
of PS and SS  i.e. all the NB are treated as a weak "probe". They change their power
due to the interactions with SS and PS, however the amount of power transferred from
SS and PS to NB is neglected with respect to the power of SS and NB. This
approximation is valid, provided the power of NB remains much smaller compared to
that of SS and NB. With multiple SS present in the fiber the SRS effect is represented
through interband Raman scattering. This is an approximation to the full expression
for the Raman polarization [1],[2] that is valid provided that the frequency separation
between the interacting signals is large enough compared to their individual
bandwidths.
In the opposite case (frequency separation between the signals comparable with their
individual spectral bandwidth) total field approach can be implemented by turning on
the option "Merge sampled bands". In this case the system (1a) is replaced by the
following single NLS Equation 4 and (1b) and (1c) remain unchanged. In Equation 4,
the Raman response function h 1111 ( t ) is the Fourier transform of the Raman
susceptibilities shown in Figure 1. Total field approach however should be used with
some care. At first, in this case, (single sampled band) XPM and four wave mixing
351
OPTICAL FIBER CWDM
effects are included automatically in the simulation and turning on or off the "XPM"
parameter in the "Nonlinearities" tab will have no effect on the results.
2
∂E iβ 2 ( ω 0 ) ∂ E i β 3 ( ω 0 ) ∂ 3 E
 + αE + 
  –   =
∂z 2 ∂T
2 6 ∂T 3
⎛ ∞ ⎞
⎜(1 – ρ) E + ρ h
2 2
+⎟⎟
⎜ ∫ 1111 ( s ) E ( T – τ ) ds (4)
⎜ 0 ⎟
iγ ⎜ ⎟E
Number of PS
⎜ ⎟
⎜ ( PS ) ⎟
⎜
+ρ ∑ R 1l Pl ⎟
⎝ l=1 ⎠
Figure 2
352
OPTICAL FIBER CWDM
Figure 3 Total field approach implemented with improper choice of sample rate. The output probe power is
0.931 mW.
Figure 4 The correct result is obtained when the bandwidth is high enough. Output probe power is
1.377mW. The slight difference in the output probe power could be attributed to FWM.
The following example shows the importance of the proper choice of numerical
parameters. Figure 2 shows the layout. The input consists of a strong (1 W power)
pump wave at 193 THz and a weak (1 mW) probe wave at 192.5 THz. "Merge
sampled bands" parameter of the optical fiber component is enabled, which means
that total field approach will be used. Attenuation effect is disabled and we use
"Constant" step size with the "Maximum nonlinear phase shift" parameter equal to
5 mrad. Raman effect is enabled The rest of the setup of the optical fiber component
is the default one. Since total field approach will be used enabling or disabling the
"XPM" parameter will have no effect on the results.
Figure 3 and Figure 4 show the obtained results together with the global parameters
of the layout in each case. Figure 5 gives the result treating the two waves as
separated channels  "Merge sampled bands" parameter is set to FALSE in the optical
fiber component. In the case presented in Figure 3 the simulated bandwidth is too
353
OPTICAL FIBER CWDM
small to accommodate the FWM mixing products of both waves and hence they are
aliased (see e.g. [8]). This false translation of the frequency of the wave (known as
aliasing) can put the a weak FWM product in the closed spectral vicinity of the pump
which will trigger a strong FWM (or modulation instability since the signal wavelengths
are in the anomalous GVD regime, which is also a kind of FWM) if the frequency
separation is small, and consequently, the coherence length is large. The result is an
entirely unphysical generation of new frequency components. Note that probe
attenuation is obtained instead of probe amplification. Figure 4 gives the correct result
since no aliasing occurs. This is achieved by having the simulated bandwidth (or
equivalently the sample rate) high enough to accommodate the three times the input
signal bandwidth.
Figure 5 Simulation in which both signals are treated as separated channels. Output probe power is
1.371 mW.
A comparison with Figure 5 (obtained treating the pump and the probe wave as two
separate sampled bands) which gives the same output power for the probe wave as
the total field approach with the sample rate correctly chosen shows that in this case
FWM effects are quite small. Besides, treating the signals as separate frequency
bands leads a significant reduction of the simulation time.
It should be kept in mind however that while in the case of total field approach, all the
parameters (dispersion, attenuation, etc.) are evaluated just once  at the reference
frequency, here (when multiple SS are considered) a set of parameters is evaluated
for each sampled signal  at the center frequency of the corresponding signal. The
meaning of the reference frequency (and reference wavelength) is the following: The
subsystem (1a) is written in a frame moving with group velocity corresponding to the
reference wavelength  no other signal parameters are evaluated at this frequency.
The reference wavelength can be either userspecified or "automatic", which
corresponds to the averaged frequency of the center frequencies of all SS and PS.
354
OPTICAL FIBER CWDM
1 2 (5)
β 1 ( ω ) – β 1 ( ω 0 ) = β 2 ( ω 0 ) ( ω – ω 0 ) +  β 3 ( ω 0 ) ( ω – ω 0 )
2
Evaluating Equation 5 and its first and second derivatives with respect to ω at the
signal frequencies { ω 1 } gives the sets of parameters:
{ β 2 ( ω 1 ) – β 1 ( ω 0 ) }, { β 2 ( ω i ) } and { β 3 ( ω i ) } .
It should be kept in mind however that with multiple sampled signals present,
specifying nonzero β 2 and β 3 (or D and S) and disabling in the same time the
"Group velocity dispersion" and "Third order dispersion" will result in
{ β 2 ( ω i ) = 0, ∀i }, { β 3 ( ω i ) = 0, ∀i } , but { β 1 ( ω i ) ≠ β 1 ( ω j ), if i ≠ j } ,
which means that no GVD induced pulse broadening will be observed but pulses with
different center frequencies will propagate with different group velocities. In contrary,
if all the sampled signals are merged to form a single frequency band disabling the
GVD effects will not only disable pulse broadening, but also will set the group velocity
constant for the entire sampled band considered.
If "Dispersion data type" is set to "From file" the data set specified by the file is
Sellmeier fitted than dispersion parameters are calculated by analytically
differentiating the fit. The file specifying the dispersion data must provide the
dependence of group delay [ps/km] on the wavelength [nm]. For this reason
"Frequency domain parameters" is disabled when "Dispersion data type" is set to
"From file".
355
OPTICAL FIBER CWDM
k = 1, k ≠ i 0
Number of PS ∞
– i ( ω i – ω l )τ
+ iγ i ρ ∑ ∫ h1111 ( τ )Ei ( T – τ )e dτ
l=1 0
∞ Number of SS
2
+ iγ i ρE i ( T ) ∫ h 1111 ( τ ) ∑ E k ( T – τ ) dτ
0 k=1
dP l ⎛ Number of PS ⎞
 = – 2α l P l + 2ργ l P l Im ⎜ ∑ R
( PP ) ⎟
P h⎟ +
dz ⎜ lh
⎝ h=1 ⎠
(6b)
Number of SS T.W. ⎧ ∞ ∗⎫
2ργ l P l ⎪ – i ( ω i – ω l )τ ⎪

T.W.
 ∑ ∫ Im ⎨ E i ( t )
⎪
∫ h1111 ( τ )Ei ( t – τ ) ( e dτ ) ⎬dt
⎪
i=1 0 ⎩ 0 ⎭
356
OPTICAL FIBER CWDM
dN m ⎛ Number of PS ⎞
 = – 2α m N m + 2ργ m N m Im ⎜ ( PN ) ⎟
∑ Rmh Ph⎟ +
dz ⎜
⎝ h=1 ⎠
(6c)
Number of SS T.W. ⎧ ∞ ∗⎫
2ργ m N m ⎪ – i ( ω i – ω m )τ ⎪

T.W. ∑ ∫ Im ⎨⎪ Ei ( t ) ∫ h1111 ( τ )Ei ( t – τ ) ( e dτ ) ⎬dt
⎪
i=1 0 ⎩ 0 ⎭
In Equation (6), the time window size is denoted by T.W., and the star symbol means
complex conjugation. The first three terms in the R.H.S of Equation (6a) are the SPM
and XPM caused by the rest of the sampled signals and the parameterized signals on
the i th sampled signal. The fourth term is responsible for the SRS induced energy
exchange between the i th sampled signal and all the other sampled signals. The
fifth term takes into account the energy exchange between the i th sampled signal
and all the parameterized signals. The last (sixth) term describes the SPM and XPM
stemming from the delayed nonlinear response of the material. This effects can be
turned on and off by the "Molecular SPM and XPM" parameter. The fourth and fifth
terms (responsible for the SRS induced energy exchange between the sampled
signals and the parameterized signals, respectively, are simultaneously switched on
by setting the "Complete Raman response" parameter to TRUE.
The sets contained in Equations (6b) and (6c) describe the evolution with propagation
of the parameterized signals powers and noise bins powers respectively. The physical
meaning of the terms in the RHS of Equations (6b) and (6c) is the following: The first
terms take into account the attenuation. The second terms describe the energy
exchange with parameterized signals due to SRS effect. These two terms are
included in the simulation by switching on the "Complete Raman response"
parameter. The last terms in the sets of Equations (6b) and (6c) describe the
interaction of parameterized signals with sampled signals and of that of noise bins
with sampled signals respectively.
Using
∞
– i 〈 ω i – ω m〉 τ
h 1111 ( ω i – ω m ) = ∫ h1111 ( τ )e dτ ,
0
the set of Equations (6) reduces to its simplified version, the set of Equations (1).
357
OPTICAL FIBER CWDM
Vector approach
Signal propagation equations
When the polarization state of the incident sampled signals is not preserved during its
propagation inside the optical fiber the scalar approach is no longer applicable and (1)
is replaced by ("Model type" parameter must set to "Vector").
2
∂E iX ∂E iX iβ 2 ( ω i ) ∂ E iX _
 + ( β 1X ( ω i ) – β 1 ( ω 0 ) ) 
 + α ( ω i )E iX + 
 
∂z ∂T 2 ∂T
2
(7a)
Number of SS Number of SS
β 3 ( ω i ) ∂ 3 E iX 2 2 2 2
6
  = iγ i 2

∂T
3 ∑ E kX – E iX + 
3 ∑ E kY E iX
k=1 k=1
2
∂E iY ∂E iY iβ 2 ( ω i ) ∂ E iY _
 + ( β 1Y ( ω i ) – β 1 ( ω 0 ) ) 
  + α ( ω i )E iY + 
 
∂z ∂T 2 ∂T
2
(7b)
3 Number of SS Number of SS
β 3 ( ω i ) ∂ E iY 2 2 2 2

6
  = iγ i 2
∂T
3 ∑ E kY – E iY + 
3 ∑ E kX E iY
k=1 k=1
SRS is disabled automatically when the vector model is selected and noise bins and
PS are just attenuated. The nonlinear terms in (7a) and (7b) contain SPM, XPM
between parallel polarization components, and XPM between orthogonal polarization
components. If the parameter "XPM" is set to TRUE, both XPM contributions
(between parallel and between orthogonal polarization components) will be included.
If "XPM" is set to FALSE, only the nonlinear contributions of SPM will included in the
model. Note that the group delays are different for the two polarization components of
the same sampled band which takes into account the birefringence. The birefringence
can be two types: "Deterministic" and "Stochastic". In the first case, the birefringence
is assumed constant and no energy exchange between the two polarization
components occurs. In the second case, ("Stochastic" birefringence) random mode
coupling is also enabled, which gives the possibility to simulate PMD (see the
technical description of Nonlinear Dispersive Fiber Total Field for the details of the
358
OPTICAL FIBER CWDM
PMD simulator). It should be kept in mind however that when the signals are
represented as multiple sampled bands PMD impairments will be identical for all
WDM channels. To obtain the frequency dependence of the penalties (or Qfactors)
total field approach must be implemented by setting "Merge sample bands" to true.
Numerical solution
The symmetrized noniterative splitstep Fourier method [1] (see the technical
description of Nonlinear Dispersive Fiber Total Field) is used to solve Equations (1a),
(6a), and (7a, b).
⎛z + h ⎞
h ⎞ h
E ( z + h, T ) = exp D̂ exp ⎜ ∫ N̂ ( z' ) dz'⎟ exp ⎛  D̂⎞ E ( z, t )
⎛ (8)
⎝2 ⎠ ⎜ ⎟ ⎝2 ⎠
⎝ z ⎠
where the D̂ is the dispersion and N̂ are the nonlinearity operators [1],[9]. Dispersion
operator is applied in the frequency domain using FFT. The approximation:
z+h
∫
(9)
N̂ ( z' ) dz' ≈ hN̂ ( exp [ ( h ⁄ 2 )D̂ ]E ( z, t ) )
z
359
OPTICAL FIBER CWDM
is used. When the "Step size" parameter is set to "Constant" (7) can be simplified
according to:
⎛z + h ⎞ ⎛z + h ⎞
h ⎞ h ⎞ h ⎞ ⎜ h
E ( z + 2h, t ) = exp ⎝ D̂⎠ exp ⎜ ∫ N̂ ( z' ) dz'⎟ exp ⎝ D̂⎠ exp ⎝ D̂⎠ ⎜ ∫ N̂ ( z' ) dz'⎟⎟ exp ⎛⎝  D̂⎞⎠ E ( z, t ) =
⎛ ⎜ ⎟ ⎛ ⎛
2 2 2 2
⎝ z ⎠ ⎝ z ⎠
(10)
z+h z+h
⎛ ⎞ ⎛ ⎞
h h
exp ⎛⎝  D̂⎞⎠ exp ⎜⎜ ∫ N̂ ( z' ) dz'⎟ exp ( hD̂ ) exp ⎜ ∫ N̂ ( z' ) dz'⎟⎟ exp ⎛⎝  D̂⎞⎠ E ( z, t )
2 ⎟ ⎜ 2
⎝ z ⎠ ⎝ z ⎠
When the system (6) is solved (the parameter "Complete Raman response" is set to
TRUE), the second order RungeKutta scheme is used to apply the nonlinearity
operator.
References
[1] G. P. Agrawal, "Applications of nonlinear fiber optics", Academic press, 3rd edition, 2001.
[2] G. P. Agrawal, "Nonlinear fiber optics", Academic press, 3rd edition, 2001.
[3] R. W. Hellwarth, Prog. Quant. Electr. 5, 1 (1977).
[4] P. Tchofo Dinda, G. Millot, and S. Wabnitz, JOSA B, 15, 1433, (1998).
[5] R.H.Stolen, J.P.Gordon, W.J. Tomlinson and H.A. Haus, JOSA B, 6, 1159 (1989).
[6] C.R.Menyuk, M.N.Islam and J.P.Gordon, Optics Letters, 16 566, (1991).
[7] K.J. Blow and D. Wood, IEEE J. Quant. Electr., 25, 2665, (1989).
[8] W. H. Press, et al., "Numerical Recipes: The Art of Scientific Computing", 2nd Edition,
Cambridge University Press, 1992.
[9] M. Lax, J. H. Batteh and G. P. Agrawal, Journ. Appl. Phys. 52 , 109, (1981).
360
BIDIRECTIONAL OPTICAL FIBER
361
BIDIRECTIONAL OPTICAL FIBER
Ports
Parameters
Main
Fiber length
362
BIDIRECTIONAL OPTICAL FIBER
Dispersion
363
BIDIRECTIONAL OPTICAL FIBER
364
BIDIRECTIONAL OPTICAL FIBER
PMD
In the case when more than one sampled signal (separated channels) is propagating
in the fiber, the PMDinduced penalties will be the same for all channels.
Different penalties (and Qfactors) will be obtained if all the channels are merged (total
filed approach). Birefringence and random mode coupling do not affect parameterized
signals and noise bins.
Nonlinearities
365
BIDIRECTIONAL OPTICAL FIBER
n2 vs. wavelength
If the "n2 data type" is set to "From file",
this tab specifies the file containing the
nonlinear index of refraction wavelength
data.
366
BIDIRECTIONAL OPTICAL FIBER
367
BIDIRECTIONAL OPTICAL FIBER
Enhanced
368
BIDIRECTIONAL OPTICAL FIBER
Numerical
369
BIDIRECTIONAL OPTICAL FIBER
370
BIDIRECTIONAL OPTICAL FIBER
Graphs
371
BIDIRECTIONAL OPTICAL FIBER
Simulation
Noise
Name and description Default value Default value Units Value range
Convert noise bins False [True, False]
Random numbers
Name and description Default value Default value Units Value range
Generate random seed True [True, False]
Technical Background
Numerical Solution
To model the bidirectional signal propagation in a fiber, an algorithm that takes two
numerical steps is used [1].
• In the first step, the equations describing the signal propagation in the forward
and backward direction are solved by an iterative method (Power analysis) and
the power distribution along the fiber is calculated.
372
BIDIRECTIONAL OPTICAL FIBER
• In the second step, the signals are propagated using the nonlinear Schrödinger
equation to describe the dynamic interactions between the copropagating
signals.
Power Analysis
The equations that describe the interactions between signals propagating in the
forward direction and backward direction and describe the generation of optical noise
due the Raman and Rayleigh scattering are defined by [2]:
373
BIDIRECTIONAL OPTICAL FIBER
where
v i, v j are frequencies
h is Plank's constant
In these equations, the following physical effects were taken into account:
a) pumptopump, signaltosignal and pumptosignal Raman interactions
b) spontaneous Raman emission and its temperature dependency
c) stimulated Raman scattering
d) pump depletions due to Raman energy transfer
e) highorder stokes generation
f) multiple Rayleigh backscattering
g) fiber loss
h) spontaneous emission noise
In the first solution step in this component, the equations (1) (forward and backward)
are solved through direct integration. In direct integration, the signal launch
configuration defines the boundary conditions in both ends of the fiber.
The convergence of the model is checked in two directions: forward and backward.
The forward direction is from input port 1 to output port 1. The backward direction is
from input port 2 to output port 2.
374
BIDIRECTIONAL OPTICAL FIBER
When a file with the normalized Raman gain is entered, it must be given values for
the Raman gain peak and Raman gain reference pump. These values are used to
calculate the Raman gain used in the simulation according the following formula:
where
where
The modeling of SBS used here is based on reference [7] and it can not be used
together with Raman amplification.
After the calculation of the power distribution along the fiber for the signals,
spontaneous emission and Rayleigh scattering, the dynamic interaction between the
copropagating signals are analyzed using the nonlinear Schrödinger equations
375
BIDIRECTIONAL OPTICAL FIBER
Scalar approach
Signal propagation equations with Raman scattering
In the scalar approach, the optical field maintains its polarization along the fiber
length. The Model type parameter from the "Numerical" tab is set to Scalar.
In this case, the following set (4) of equations governs the evolution of the slowly
varying electric field envelopes (Ei).
These envelopes are a set of sampled signals (SS), powers (Pl) of another set of
parameterized signals (PS), and powers (Nm) of a third set of noise bins (NB).
376
BIDIRECTIONAL OPTICAL FIBER
377
BIDIRECTIONAL OPTICAL FIBER
Raman susceptibility for fused quartz is shown in Figure 1. It should be noted that "*"
means complex conjugation.
In Equation (4a), E i = E i (z,T) is the electric field envelope of the ith sampled
signal.
378
BIDIRECTIONAL OPTICAL FIBER
The derivatives of the propagation constant of the fiber mode β ( ω ) , with respect to
n n
frequency β n = ∂ β ( ω ) ⁄ ( ∂ω ) n = 1, 2 are the first order β 2 and second order
β 3 group velocity dispersion (GVD) parameters and are evaluated at the center
frequencies { ω i } of the sampled signals.
The nonlinear coefficients for every SS, NB or PS in (4) are defined according to
The meaning of the terms in the lefthand side of the subsystem (4a) is the same as
in the total field approach fiber model (see the technical description of this
component).
The first two terms in the right side of (4a) give the SPM and XPM contributions of the
remaining sampled signals. The third term is the XPM contribution of the PS. The
fourth and the fifth terms describe the SRSinduced interactions between the ith
sampled signal and rest of the sampled signals and with the parameterized signals,
respectively.
Subsystems (4b) and (4c) describe the power balance of the set of PS and NB,
respectively. These are obtained by replacing the NLS equations for NB and PS with
the timeaveraged versions of their power conservation laws.
In the absence of attenuation, the total number of photons is conserved as (4) shows.
The first terms in the right sides of (4b) and (4c) take into account the attenuation
effects. The second and the third terms in the right side of (4b) describe the SRS
induced power transfer between the lth PS and the rest of the PS and between the l
th PS and the SS, respectively.
The second and the third terms in the right side of (4c) are responsible for the SRS
induced interactions between noise bins and PS and noise bins and SS. With multiple
SS present in the fiber, the SRS effect is represented through interband Raman
scattering.
This is an approximation of the full expression for the Raman polarization [3,4] that is
valid if the frequency separation between the interacting signals is large enough
compared to their individual bandwidths.
When the frequency separation between the signals is comparable with their
individual spectral bandwidth, the total field approach can be implemented by turning
on the option "Merge sampled bands".
In this case, the system (4a) is replaced by the following single NLS equation (7) and
(4b) and (4c) remain unchanged.
379
BIDIRECTIONAL OPTICAL FIBER
In equation (7), the Raman response function h 1111 ( t ) is the Fourier transform of
the Raman susceptibilities shown in Figure 1.
Total field approach however should be used with some care. At first in this case
(single sampled band), XPM and four wave mixing effects are included automatically
in the simulation and turning on or off the XPM parameter in the "Nonlinearities" tab
will have no effect on the results.
Keep in mind that in the total field approach, all the parameters (such as dispersion
and attenuation) are evaluated just once  at the reference frequency.
In this case, when multiple SS are considered, a set of parameters is evaluated for
each sampled signal  at the center frequency of the corresponding signal.
The meaning of the reference frequency (and reference wavelength) is the following:
The subsystem (4a) is written in a frame moving with group velocity corresponding to
the reference wavelength. That is, no other signal parameters are evaluated at this
frequency.
Evaluating (8) and its first and second derivatives with respect to ω at the signal
frequencies { ω i } gives the sets of parameters β 1 ( ω ) – β 1 ( ω 0 ) }, { β 2 ( ω i ) }
and { β 3 ( ω i ) } .
It should be kept in mind, however, that with multiple sampled signals present,
specifying nonzero β 2 and β 3 (or D and S) and at the same time disabling the
"Group velocity dispersion" and "Third order dispersion", will result in
{ β 2 ( ω i ) = 0 ,∀i } , { β 3 ( ω i ) = 0 ,∀i } but { β 1 ( ω i ) ≠ β 1 ( ω j ), ifi ≠ j } .
This means that no GVDinduced pulse broadening will be observed but pulses with
different center frequencies will propagate with different group velocities.
380
BIDIRECTIONAL OPTICAL FIBER
To the contrary, if all the sampled signals are merged to form a single frequency band,
disabling the GVD effects will not only disable pulse broadening, but it also will set the
group velocity constant for the entire sampled band considered.
If "Dispersion data type" is set to "From file", the data set specified by the file is
Sellmeier fitted. The dispersion parameters are calculated by analytically
differentiating the fit.
The file specifying the dispersion data must provide the dependence of group delay
[ps/km] on the wavelength [nm]. For this reason, "Frequency domain parameters" is
disabled when "Dispersion data type" is set to "From file".
Vector approach
When the polarization state of the incident sampled signals is not preserved during its
propagation inside the optical fiber, the scalar approach is no longer applicable. A
vector model is then selected and solved.
The vector model is similar to the model presented in the Optical Fiber WDM (see
Optical Fiber WDM Technical Background). In the same way, Raman scattering is not
applied.
References
[1] J. Ko; S. Kim; J. Lee; S. Won; Y. S. Kim; J. Jeong, "Estimation of performance degradation of
bidirectional WDM transmission systems due to Rayleigh backscattering and ASE noises using
numerical and analytical models", IEEE J. of Lightwave Technology, Vol.: 21 , Issue: 4 , April
2003, Pag.:938  946
[2] M. Karasek, M. Menif, "Protection of surviving channels in pumpcontrolled gainlocked Raman
fibre amplifier", Optics Communications 210 (2002) 5765.
[3] G. P. Agrawal, "Applications of nonlinear fiber optics", Academic press, 3rd edition, 2001.
[4] G. P. Agrawal, "Nonlinear fiber optics", Academic press, 3rd edition, 2001.
[5] R. W. Hellwarth, Prog. Quant. Electr. 5, 1 (1977).
[6] P. Tchofo Dinda, G. Millot, and S. Wabnitz, JOSA B, 15, 1433, (1998).
[7] A. backa, G. Jacobsen, and B. Tromborg, "Dynamic Stimulated Brillouin Scattering Analysis,"
J. Lightwave Technol. 18, 416 (2000)
381
BIDIRECTIONAL OPTICAL FIBER
Notes:
382
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
This component is an obsolete version that is included with OptiSystem for backwards
compatibility purposes  It was replaced by the Optical Fiber component.
Ports
Parameters
Main
Fiber length
383
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Dispersion
384
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Birefringence
385
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Nonlinearities
386
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
387
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Simulation
Variable = Main
Channel Current
Nonlinear
length/Number of
Steps
388
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
3D graphics selection
Waveform ON — [ON,OFF]
Graphs
389
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Parameters—Detailed descriptions
In the following section, the parameters descriptions are further elaborated. There are
descriptions of features pertaining to multiple parameters, and also extended
descriptions of individual parameters.
Note: Many parameters pertaining to the NDF can be defined as either constant
or wavelength dependent/from file values. The first option is used usually for rapid
development of simple designs. If a parameter is wavelength dependent
(arb. curve ) you have to prepare a text file with (Wavelength
ParameterValue) data pairs, and create the parameter in the appropriate
Component properties dialog box. This option is recommended for detailed,
quantitatively precise designs. Many parameters of the NDF, such as losses,
dispersion, and effective fiber area, can be defined in both ways  as constants or
curves loaded from a file. When a parameter is defined as a curve, the format of
the text file is as follows:
Wavelength_1 ParameterValue_1
Wavelength_2 ParameterValue_2
Wavelength_3 ParameterValue_3
......
Wavelength_N ParameterValue_N
The units of wavelength are nanometers ( nm ). The units and the value ranges of the
parameter values are the same as those of the respective 'constant' parameters.
For example, when a loss spectrum is loaded from file it might look like:
1500 1.99E01
1525 1.92E01
1550 1.89E01
1575 1.93E01
1600 2.05E01
or:
1500 0.199
1525 0.192
1550 0.189
1575 0.193
1600 0.205
390
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
An arbitrary number of points (file lines) are permitted, except 0 (empty file). The
column separator can be an arbitrary number (except 0) of either spaces or tabs. The
files are opened using the standard Windows "File Open" dialog box.
Technical background
Origin of the nonlinearity
At high optical intensities for intense electromagnetic fields, the dielectric medium
behaves as a nonlinear medium. This is also the case for the fiber material. Under the
influence of intense electromagnetic fields, the motion of bound electrons becomes
an harmonic and, as a result, the induced polarization P from the electric dipoles
becomes nonlinear function of the electric field E:
OptiSystem currently supports several different models specialized for different signal
representations and/or combinations of parameters.
Model Ia
This model has been derived for the separated channels signal representation. It
also accounts explicitly for the nonlinear interactions and mixing of the orthogonal
polarization modes in an SM fiber. It is a system of 2N coupled modified nonlinear
Schrödinger equations (NLSE).
391
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
For Sampled signals, the following effects are accounted for: XPM, XPM of
orthogonally polarized modes, Raman, FWM, and SSFS.
Whereas, for Parameterized signals and ASE noise bins, we account for Raman and
FWM.
There are 3 types of polarization evolution that could be taken into account:
Averaged polarizations
In this case, the effect of the Kerr nonlinearity is averaged over the Poincaré sphere,
and is taken into account with a coefficient value of 8/9. The effect of nonlinear PMD
[2] is not taken into account.
The intrapulse Raman scattering (or Raman Self Shifting) effect, which leads to
soliton self frequency shift, has to be considered for very short optical pulses with
duration ~ picosecond or smaller.
2 3
∂A ix ∂A ∂ A ix 1 ∂ A ix 1
 ± β 1 ix + i β 2 
 –  β 3  +  α i A ix =
∂z ix ∂t 2 i ∂t 2 6 i ∂t 3 2
N
f ijkl
i ∑ Mγ x, μ, v, ρ δ ( ω k + ω l – ω j – ω i )  A jμ∗ A kv A lρ exp ( iΔβz ) +
f ii
j, k, l = 1
j, k, l ≠i
μ = x
v, ρ = x , y
1
 iγA iy 2 A ix∗ exp ( – 2iΔβ xy z ) –
3
N N
j n f ij 2 i n f ij 2
i ∑ gR gR ( ωj – ω i )  A jx A ix – ig R
f ii ∑ g R ( ω i – ω j )  A jx A ix
f ii
j = 1 j = 1
j ≠1 j ≠1
ωj > ωi ωj < ωi
392
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
N
2 f ij 2
iγ A ix A ix + 21γ ∑ fii Ajx Aix +
j = 1
j ≠1
N
1 2 2 f ij 2
3
iγ A iy A ix +  iγ
3 ∑ fii Ajy A ix +
j = 1
j ≠1
2
∂ A ix
iγT R  A ix
∂t
(1)
where Aix, Aiy are the slowly varying complex electric field amplitudes of the radiation in the
respective x/y polarization mode of the i’th WDM channel,
2πcβ 2i
D = – 
2

λ
αi is the loss coefficient for the respective carrier frequency of the channel
n
g R is the normalized Raman gain function taken from reference [1], Figure 8.1 on
page 300.
393
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
∞ ∞
2
∫ ∫ ( F ( x ,y ) ) dx dy
–∞ –∞

∞ ∞

Aeff =
4
∫ ∫ F ( x ,y ) dx dy
–∞ –∞
(3)
where F(x,y) is the modal field distribution of the fiber mode.
∞ ∞
2 2
∫ ∫ F i ( x ,y ) F j ( x ,y ) dx dy
–∞ –∞


fii = ∞ ∞ ∞ ∞
2 2
∫ ∫ F i ( x ,y ) dx dy ∫ ∫ F j ( x ,y ) dx dy
–∞ –∞ –∞ –∞
(4)
M is the multiplicity factor. Its value is 2 if all three waves are different — otherwise,
its value is 1.
〈 F i∗ F j∗ F k F l〉


2 2 2 2 1⁄2
[ 〈 Fi 〉 〈 Fj 〉 〈 Fk 〉 〈 Fl 〉 ]
fijkl=
394
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
(5)
where the angle brackets denote integration over the transverse coordinates x and y.
Also
Δβ = ( ω k n k + ω l n l – ω j n j – ω i n i ) ⁄ c
(6)
where
Δβ xy = β y – β x
(7)
are the propagation constant mismatches of the processes of FWM and (PFWM) and TR ~ 5
fsec is the slope of the Raman gain curve.
Model Ib
Similar to Model Ia, but disregards the polarization evolution of the signal and uses
the average power of the two polarization modes. It consists of a system of only N
coupled modified nonlinear Schrödinger equations (NLSE) with correspondingly
adjusted nonlinear coefficients.
Model I
Derived for the case of the total field signal representation. All sampled signals are
in a single frequency band. This is the basic method used for modeling WDM
systems.
It also accounts explicitly for the mixing of the orthogonal polarization modes in an SM
fiber. It is a system of two coupled modified nonlinear Schrödinger equations (NLSE).
It works with all types of signals: Sampled, Parameterized and ASE noise bins. For
parameterized and ASE noise bins, only linear losses are taken into account.
'Total field approach' automatically accounts the XPM and FWM effects. There is no
possibility to switch off these effects.
'Total field approach for both polarizations' will additionally account for PXPM of
orthogonally polarized signals' and PFWM of orthogonally polarized signals'.
The model for the case of one polarization has the following form:
2 3 2
∂A
 ± β 1 ∂A
 + i β 2 ∂
A 1 ∂
A 1 2 ∂A
– β 3 3 +  αA x = iγ A A – iγT R  A
∂z ∂t 2 ∂t 2 6 ∂t 2 ∂t
(8)
395
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
All the parameters in the above equation have been explained, along with the
Model Ia.
Numerical Methods
The three models (Model la, Model lb, and Model l) are solved by a scalar or
vectorial version of the splitstep Fourier transform method:
∂A
 = [ D + N ]A
∂z
(9)
Fixed
In this case the step size is simply Δz = L ⁄ N , where L is the length of the fiber
and N is the user defined number of steps.
396
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
L NL ( z ) = 1 ⁄ γP ( z ) .
In this way, the possible changes in L NL ( z ) due to loss or gain are taken into
account.
The term indicates the channel used in the calculations above. When the separate
channels signal representation is used, it is either the channel with the highest power
or the central channel. If we use only one continuous spectral band, as in the total field
signal representation, there can be only one main channel.
References
[1] Agrawal, G.P., “Nonlinear Fiber Optics, 3rd Edition”, Academic Press, 2001.
[2] Marcuse, D., Menyuk, C.R., and Wai, P.K.H., "Application of the Manakov  PMD Equation to
Studies of Signal Propagation in Optical Fibers with Randomly Varying Birefringence", Journ.
Light. Technol.,15, 17351746 (1997).
[3] Tchofo Dinda, P., Milot, G., and Wabnitz, S. "Polarization Switching and Suppression of
Stimulated Raman Scattering in Birefringent Optical Fibers", JOSA B, 15, 14331441 (1998).
397
NONLINEAR DISPERSIVE FIBER (OBSOLETE)
Notes:
398
LINEAR MULTIMODE FIBER
This component is a multimode fiber. The component has two modes of operation.
The fist one assumes the fiber has sufficient mode mixing due to imperfections or
splices; in this case the modal transfer function approaches a Gaussian function. The
second one allows the user to load measured modal delays and powercoupling
coefficients. The component also includes first and secondorder chromatic
dispersion.
Ports
Parameters
Main
Fiber attenuation
Cutback factor
399
LINEAR MULTIMODE FIBER
Propagation delay
Delay skew
Chromatic dispersion
400
LINEAR MULTIMODE FIBER
Simulation
Noise
Technical background
The optical fiber has three dominant dispersion mechanisms, modal, and firstorder
chromatic and secondorder chromatic. We can assume that modal dispersion and
chromatic dispersion mechanisms act independently and can be treated
separately[1][2]. Propagation delay and Delay skew parameters are applied to the
signal output after the dispersion effects.
Modal dispersion
Personick has shown that if a multimode fiber has sufficient mode mixing due to
imperfections or splices, in this case the modal transfer function approaches a
Gaussian function [3][4][5]
2 2
ω σ
–  (1)
2
HM ( ω ) = e
401
LINEAR MULTIMODE FIBER
where ω is the angular baseband frequency and σ is the RMS impulse response
width.
In this model, the modal dispersion is characterized by the 6 dB half of the optical
power frequency:
2 ⋅ 1n ( 2 ) 1 γ
B 6dB =  = B M ⎛ ⎞
(2)
2⋅π⋅σ ⎝ L⎠
where B M is defined by the parameter Modal bandwidth and L is the fiber parameter
Length. γ is the cutback factor, that takes into account the mode coupling, mixing and
concatenation effects.
⎛ ⎞
⎜ 1n ( 2 ) ⋅ f
2 ⎟
⎜ – 
γ 2⎟
 (3)
⎜ ⎛ B ⋅ ⎛ 1⎞ ⎞ ⎟
⎝ ⎝ M ⎝ L⎠ ⎠ ⎠
HM ( f ) = e
If the Measured modal delays is enable, the modal dispersion is calculated from
measured modal delays and power coupling coefficients from parameter Filename.
The file format for the modal delay and power coupling coefficients file is the following:
Each file contains three columns. The first column contains the order of each mode
group supported by the fiber. The second column contains the average modal delay
of each mode group, in ns. The third column contains the powercoupling coefficients,
which indicate the relative excitation of each mode group. The modal delay is relative
to the parameter Reference length. The output signal is calculated from the impulse
response of the fiber in time domain. If the parameter Frequency response is enabled,
the output signal will be calculated in the frequency domain, in this case, the transfer
function of the fiber is calculated according to
– j2πfτ m (4)
HM ( f ) = ∑ ( Pm e )
m
Where P m is the power coupling coefficient and the τ m is the modal delay for mode
m. The principle of this model is described in detail in [8].
402
LINEAR MULTIMODE FIBER
Chromatic dispersion
Since most of the injectionlasers used in optical fiber communications have a
Gaussian line shape [1][2][6][7], we can use the solution for the chromatic transfer
function for a perfect Gaussian linewidth case:
2
( ω ⁄ ω1 )
– 

1 2 ( 1 + iω ⁄ ω 2 ) (5)
H D ( ω ) = 
1⁄2
e
( 1 + iω ⁄ ω 2 )
2 –1
ω 1 = [ σ λ ( S + 2 D ⁄ λ r )L ]
(6)
–1
ω1 = –( σλ D L )
The parameter Use Sellmeier approximations defines whether you will enter D and
S directly, or if they will be calculated from the Sellmeier approximations [2]:
4
S0 ⎛ λ 0⎞
D =  ⎜ λ r – 3⎟
4⎝ λ ⎠ r (7)
4
S0 ⎛ λ 0⎞
S =  ⎜ 1 + 3 4⎟
4⎝ λr ⎠
403
LINEAR MULTIMODE FIBER
References
[1] C. Yabre, "Comprehensive Theory of Dispersion in GradedIndex Optical Fibers", Journal of
Lightwave Technology, Vol. 18, No. 2, pp. 166176, February 2000.
[2] G.D. Brown, "Bandwidth and Rise Time Calculations for Digital multimode FiberOptical Data
Links", Journal of Lightwave Technology, Vol. 10, No. 5, pp. 672678, May 1992.
[3] S.D.Personick "Baseband Linearity and Equalization in Fiber Optic Digital Communication
Systems", The Bell System Technical Journal, pp. 11741194, September 1973.
[4] D.G.Duff, "ComputerAided Design Of Digital Lightwave Systems", IEEE Journal on Selected
Areas in Communications, Vol. SAC2, No. 1, pp. 171185, January 1984.
[5] D.O.Harris, J.R. Jones "Baud Rate Response: Characterizing Modal Dispersion for Digital Fiber
Optic Systems", Journal of Lightwave Technology, Vol. 6, No. 5, pp. 668677, May 1988.
[6] J.L.Gimlett, N,K,Cheung "Dispersion Penalty Analysis for LED/SingleMode Fiber
Transmission Systems", Journal of Lightwave Technology, Vol. LT4, No. 9, pp. 13811391,
September 1986.
[7] T. Pfeiffer, M. Witte, B. Deppisch "HighSpeed Transmission of BroadBand Thermal Light
Pulses Over Dispersion Fibers", IEEE Photonic Technology Letters, Vol. 11, No. 3, pp. 385387,
March 1999.
[8] M. Webster et al., “A statistical analysis of conditioned launch for Gigabit Ethernet links using
multimode fiber”, Journal of Lightwave Technology, Vol. 17, No. 9, pp. 15321541, September
1999.
404
PARABOLICINDEX MULTIMODE FIBER
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Length 1 km m, km [0, 100000]
Fiber Profile
Name and description Default value Default unit Units Value range
Core radius 25 um [1, 100]
405
PARABOLICINDEX MULTIMODE FIBER
Name and description Default value Default unit Units Value range
Refractive index step 1 % [0.01, 10]
Chromatic Dispersion
Name and description Default value Default unit Units Value range
Include chromatic dispersion NO [YES, NO]
Numerical
Name and description Default value Default unit Units Value range
User defined wavelength NO [YES, NO]
406
PARABOLICINDEX MULTIMODE FIBER
Name and description Default value Default unit Units Value range
Solver wavelength 820 nm [100, 2000]
Report
The summary of fiber attributes,
including number of modes, coupling
coefficients and delays
407
PARABOLICINDEX MULTIMODE FIBER
Graphs
Name and description Default value Default unit Units Value range
Calculate graphs NO [YES, NO]
LP(m,n) 0, 1
Simulation
Name and description Default value Default unit Units Value range
Enabled YES [YES, NO]
Graphs
408
PARABOLICINDEX MULTIMODE FIBER
Power coupling and modal delay Modal delay (s) Power coupling
Technical Background
This component is a multimode fiber with parabolic refractive index (Figure 1). The
parabolic profile is described analytically as [1]:
where n1 is the parameter Refractive peak index at the fiber center, n2 is the refractive
index in the cladding, Δ is the parameter Refractive index step, a is the parameter
Core radius and (ba) is the parameter Clad radius.
The radial distance from the fiber center r is discretized using the parameter Number
of radial steps.
409
PARABOLICINDEX MULTIMODE FIBER
The main result of the fiber calculation is the spatial profile, coupling coefficients and
the time delay for each mode.The signal center frequency for the mode solver
depends on the center frequency of the input signal. The user can force the mode
solver to work at a user defined wavelength by enabling parameter User defined
wavelength.
Additionally, the user can provide a file with the modal attenuation. The modal
attenuation file format is a list with the m and n mode index and the attenuation in
dB/km for polarizations X and Y:
The final solution for the output field of the combined temporal and spatial properties
of the fiber for N number of modes is shown below:
where Ein is the signal input field, ci is the coupling coefficient between the fiber
modes and the spatial profile if the input field and Ei is the fiber mode for each index i.
410
PARABOLICINDEX MULTIMODE FIBER
If the power of (ci.Ein) is below the parameter Min. signal power, the signal i is
discarded.
The component has an analytical mode solver that will calculate the LP(m,n) modes.
The parameter LP(m,n) max. defines the maximum order for the radial and azimuthal
indexes m and n when searching for fiber modes. The analytical solution for the field
in the core, for each m and n index is [1]:
where Ea,0 is a scaling factor for the boundary conditions in the core/clad fiber
interface. L is the Laguerre polynomial function, and k0 and ρ are given by:
where λ 0 is the center wavelength. The solution in the clad is given by:
where Eb,0 is a scaling factor for the boundary conditions in the clad/core fiber
interface, K is the modified Bessel function. The propagation constant β m, n is
calculated accordingly to:
There are two main results of this calculation. They are the time delay associated with
each mode and the coupling coefficient between the input spatial fields and each of
the spatial fiber modes. The propagation constant β is used to calculate the time
delay per mode:
411
PARABOLICINDEX MULTIMODE FIBER
where L is the fiber length. The coupling coefficient is calculated according to:
where Ei is the spatial profile for each m,n mode, including the sin and cosine factors,
and Ein is the spatial input field.
The user can also generate a constant mode power distribution (MPD). In this case
the coefficients will be constant. Enabling the parameter Generate overfilled launch
can generate an overfilled launch mode.
After the calculation, the parameter Report will have a list of the modes, coupling
coefficients and delays for each mode and polarization.
By enabling the parameter Calculate graphs, the user can see the results from the
analytical mode solver. The results can include the mode index number for the
calculated modes, the effective index, delays, power coupling, refractive index
profiles, and spatial and radial profiles for the individual and overfilled mode.
The fiber model also includes the chromatic dispersion effects. If chromatic dispersion
is enabled, the user can specify the value of the dispersion and dispersion slope, as
well as Sellmeier.
References
[1] A. Ghatak, K. Thyagarajan, “Introduction to Fiber Optics”, Cambridge University Press, New
York, NY, 1998.
[2] G.D. Brown, "Bandwidth and Rise Time Calculations for Digital multimode FiberOptic Data
Links", Journal of Lightwave Technology, VOL. 10, NO 5, May 1992, pp. 672678.
412
MEASUREDINDEX MULTIMODE FIBER
Ports
Parameters
Main
Name and description Default value Default unit Units Value range
Length 1 km m, km [0, 100000]
Fiber Profile
Name and description Default value Default unit Units Value range
OptiFiber file format NO [YES, NO]
Filename Index.txt
413
MEASUREDINDEX MULTIMODE FIBER
Material Properties
Name and description Default value Default unit Units Value range
Include material properties NO [YES, NO]
3.0% B2O3
doped silica,
3.1% germania
doped silica,
9.1% P2O5
doped silica,