Professional Documents
Culture Documents
Models Reference
Phone: 914-923-2164
Fax: 914-923-2169
E-mail: info@rsoftdesign.com
Web: www.rsoftdesign.com
OptSim
1989 – 2010 RSoft Design Group. – All rights reserved.
www.rsoftdesign.com
Build OS0521010
This manual contains confidential and proprietary information regarding
products of RSoft Design Group. The manual is intended solely for use by
authorized OptSim users, who have in effect a License for the OptSim product
(under the License Agreement conditions) and the information contained
herein may not be used or disclosed to others except by written permission of
RSoft Design Group. No part of this document may be reproduced or
transmitted in any form or by any means, mechanical or electronic, including
photocopying and recording, for any purpose, without the express written
permission of RSoft Design Group.
Information in this document is subject to change without notice and does
not represent a commitment on the part of RSoft Design Group. The software
described in this document is furnished under a License Agreement and may
be used or copied only in accordance with the terms of the agreement.
Software copying is against the law except as specifically allowed in the
license agreement.
Trademarks
Artifex, OptSim, ModeSYS, NetworkDesigner are trademarks of RSoft Design
Group and its subsidiaries.
All other products and companies mentioned herein are trademarks of their
respective owners.
Contents
1 Introduction.......................................................................................................... 1
1.1 How to Utilize the Components............................................................................................3
1.1.1 Connection ...........................................................................................................5
1.2 Multi Input-Output Components ..........................................................................................5
1.3 Logical Signal Levels ..........................................................................................................5
1.4 Center Frequency Requirements ..........................................................................................6
1.5 Time Domain Samples........................................................................................................7
1.6 Electric Clock Signal Information .........................................................................................8
3 Electrical Sources................................................................................................ 55
3.1 Electrical Waveform Generator ..........................................................................................57
3.1.1 Bias Signal Generator...........................................................................................57
3.1.2 Sinusoidal Wave Signal Generator .........................................................................58
3.1.3 Sawtooth Wave Signal Generator...........................................................................59
3.1.4 Square Wave Signal Generator ..............................................................................59
3.1.5 Clock Signal Generator .........................................................................................60
3.2 Ideal Clock Generator with Continuous Time Resolution........................................................62
3.3 Impulse Generator...........................................................................................................63
3.4 Clock Signal Voltage-Controlled Generator ..........................................................................64
3.5 Voltage Controlled Oscillator .............................................................................................66
3.6 Timing Pulse Generation on Zero Crossing ..........................................................................67
3.7 Electrical Noise Generator.................................................................................................68
3.7.1 Gaussian-distributed Noise Signal Generator ...........................................................68
3.7.2 Rayleigh-distributed Noise Signal Generator............................................................69
3.7.3 Uniform-distributed Noise Signal Generator ............................................................69
3.7.4 Video-Like distributed Noise Signal Generator .........................................................69
5 Record&Playback ................................................................................................ 77
5.1 Optical Playback ..............................................................................................................78
5.2 Optical Recorder ..............................................................................................................80
5.3 Electrical Playback ...........................................................................................................81
5.4 Electrical Recorder ...........................................................................................................82
5.5 Logical Playback ..............................................................................................................83
5.6 Logical Recorder ..............................................................................................................85
7 Fibers .......................................................................................................... 95
7.1 Optical Fiber Link.............................................................................................................96
7.1.1 Loss Section...................................................................................................... 100
7.1.2 Dispersion Section ............................................................................................. 102
7.1.3 Grammar for Fiber Files...................................................................................... 103
7.1.4 Statistical Dispersion Section .............................................................................. 106
7.1.5 Advanced Effects Section.................................................................................... 108
7.1.6 Raman Effect Section ......................................................................................... 110
7.1.7 SBS Section ...................................................................................................... 118
7.2 References.................................................................................................................... 124
14 RF Mo-Demodulators........................................................................................ 213
14.1 Analog Amplitude Modulator............................................................................................ 216
14.2 Coherent Amplitude Demodulator .................................................................................... 217
14.3 Frequency Modulator...................................................................................................... 218
14.4 Frequency Demodulator.................................................................................................. 219
14.5 Analog Phase Modulator ................................................................................................. 220
14.6 Phase Demodulator........................................................................................................ 221
14.7 Phase Detector .............................................................................................................. 221
14.8 Electrical Envelope......................................................................................................... 222
14.9 PSK Digital Modulator..................................................................................................... 223
14.10 PSK Modulator with Residual Carrier ........................................................................... 224
14.11 PSK Coherent Demodulator........................................................................................ 225
14.12 Offset-PSK Modulator Parallel Input ............................................................................ 226
14.13 Offset-PSK Demodulator Parallel Input ........................................................................ 227
14.14 Differential PSK Modulator ......................................................................................... 228
14.15 Differential PSK Demodulator ..................................................................................... 229
14.16 M-QAM Modulator ..................................................................................................... 230
14.17 M-QAM Demodulator................................................................................................. 231
14.18 Duobinary MSK Modulator.......................................................................................... 232
14.19 Duobinary MSK Demodulator ..................................................................................... 233
14.20 FSK Modulator with Continuous Phase ......................................................................... 234
14.21 FSK Demodulator with Continuous Phase..................................................................... 235
14.22 Binary Pulse Amplitude Demodulator........................................................................... 236
14.23 Differential MSK Demodulator .................................................................................... 237
Index..................................................................................................................... 531
OptSim has a wide and growing library of components, which are included for
your convenience in the component palette, at the left of the design area.
Several useful description and details on components are reported into the
following sections:
section 1.1 describes how to insert a component and its parameters in a
design
section 1.2 describes how to manage connections for components having
multiple inputs and/or outputs
section 1.3 describes the meaning of the "level" of a logical signal
section 1.4 reports information on the central frequency of a subcarrier
encoded electrical signal
section 1.5 reports information related to the number of samples used to
represent a digital signal
section 1.6 reports information on a particular class of components that
are used to generate timing signals
With the LIB button you can access to the secondary library, which contains
less frequently used components, and all user-defined components. These
components allow you to extend OptSim simulation capability to:
complex modulation formats such as PSK, DPSK, FSK, QAM, as typically
used in CATV applications
carrier and clock recovery for these formats
complex digital and analog signal processing
generation of several signals formats, both deterministic (sinusoidal,
square and sawtooth wave)
systems simulation such as Voltage Controlled Oscillators, Phase Locked
Loops and many others
Introduction
Note
LIB components do not have any control on the parameter ranges. You
should use these components paying attention to the parameter ranges,
which are specified in the component Help pages. While incorrect parameter
values for components belonging to the main library are detected as soon as
the parameters are inserted into a dialog window, they may generate runtime
error when this happens for LIB components.
2. Double click on the icon; the following dialog window will appear showing
all the currently available LIB components divided into categories (Figure
1.2).
3. Click on the chosen Category and then on the chosen component in
Blocks; a new dialog window will appear (Figure 1.3) showing all the
component parameters.
1.1.1 Connection
Different connection types are drawn in different colors depending on the kind
of components they are joined to:
red for optical connections i.e. ideal fiber (thin line) and realistic fiber
(thick line)
blue for electrical connections
black for logical connections
Electrical signal
representation
1 GHz 1/2 f
Simulation
Bandwidth
0 .8
DELT = [ps]
BWVBS
Usually, sampling related problems are completely transparent, because time
related parameters (such as the period of a square wave) are always
requested in terms of picoseconds, and not in terms of number of samples.
Unfortunately, some LIB components need to have information about timing
in terms of number of samples.
Please, use the previous formula to derive sampling time and the following
one to obtain timing information in terms of number of samples:
Note
Clock signals are standard electrical signals, and can thus be visualized by
any electrical measurement component.
2 Optical Sources
Lasers
Pulse Sources
Noise Sources
2.1 Laser
This component simulates a laser. The output state of polarization is
aligned with the geometrical x-axis, so as to be compatible with the other
components such as the Polarization Rotator (section 19.8) and the
Polarization Modulator (section 8.5).
Three models have been implemented:
• a simple model considering only the phase noise (CW Lorentzian
Laser)
• a realistic model based on rate equation integration (Rate Equations
Laser)
• a realistic model based on rate equation integration for Separate
Confinement Heterostructure Multi Quantum Well lasers (SCH-MQW)
where physical parameters of the laser can be obtained with a fitting
procedure over experimentally measured curves.
SPT Behavior
Laser sources are considered as if they generated a single tone at the
nominal center emission frequency of the source. Therefore, in the optical
spectrum a single line is placed. Its level is equal to the defined laser output
power. Linewidth is neglected if the CW Lorentzian Laser is selected. For the
Rate Equations Laser and the Custom MQW laser, their output power could be
found only using time-domain simulation. Hence, for SPT simulations, the
user is explicitly requested to supply the output power. If you don’t know the
average output power of the laser we suggest you perform a VBS simulation
of the device only, measure the output power, and then employ the
measured value as SPT parameter for the laser.
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the three laser models, CW
Lorentzian, Rate Equation, Custom MQW.
Laser center emission frequency [nm] or [THz]: the emission frequency
of the source. Note that for the Rate Equations and Custom MQW models this
parameter sets the laser emission frequency at threshold. The actual
emission frequency will then be a function of the injected current. Note that a
steady-state blue shift is expected, since thermal effects are neglected and
only adiabatic chirp is simulated. For the Rate Equations model, the shift
amount is indicated in the parameter dialog box. Allowed values: min flowSPT;
max fupSPT; default f0,VBS.
Status [-]: turns on "1" or turns off "0" the optical source. Allowed values: 0
or 1; default 1.
Parameters
CW power [dBm] or [mW]: the output power of the laser. Allowed values:
min: -; max: -; default: 0 dBm.
Linewidth Sets the laser linewidth.
FWHM [MHz]: the power spectrum width at half maximum. Allowed values:
min: 0; max: -; default: 10.
full width at -20 dB [MHz]: the power spectrum width at –20 dB from
maximum. Allowed values: min: 0; max: -; default: relative to FWHM at 10
MHz.
Laser Phase: sets the laser phase at the start of the simulation.
Random: the initial laser phase is randomly set. This option is useful
when several lasers are used in the same simulation, like in WDM
systems, in order to ensure a statistical randomization of the initial
phases of the laser sources.
Deterministic: the initial laser phase is set by the Initial Laser Phase
value.
Initial Laser Phase [rad]: sets the laser phase at the start of the
simulation. Allowed values: min: 0; max: 2π; default: 0 (if the Deterministic
option is selected).
Laser Noise Bandwidth: sets the laser noise bandwidth.
Ideal: infinite bandwidth phase noise.
Realistic: bandwidth-limited phase noise. Phase noise is filtered using a
two poles filter, therefore Ralaxation Oscillation Peak Frequency and
Ralaxation Oscillation Peak Overshoot options are requested.
Ralaxation Oscillation Peak Frequency [GHz]: sets the resonant
frequency of laser phase noise. Allowed values: min: >0; max: -; default: 5
(if the Realistic option is selected).
c2 2 c 3 c
Pout , NL (t ) = Pout (t ) + Pout (t ) + 3 Pout (t ) + c4 Pout4 (t ) + 5 Pout5 (t ) (1)
2 6 24 120
where Pout,NL(t) becomes the actual power output by the model and the cj's
are user-supplied parameters.
The coefficients cj's in (1) may be used to determine the amount of CSO and
CTB that can be expected from the laser in multi-subcarrier transmission. In
the following we refer to CTB and CSO induced by the cj's alone but note that
a certain amount of non-linearity stemming from the standard rate equation
model will already be present and is not accounted for by the cj's. You may
want to first assess through simulation the amount of non-linearity generated
by the rate equations alone and then, if necessary, add further non-linearity
through the cj's.
To quantitatively show how the cj's effect the P/I curve, we first re-write
expression (1) substituting c1Iin(t) = c1[I(t) - Ith(t)] for Pout(t), where c1 is the
slope of the P/I curve as it results from the rate equations alone , I(t) is the
injection current, Ith is the threshold current and Iin(t) is the amount of
injection current above threshold:
c2 2 2 c3 3 3 c4 4 4 c5 5 5
Pout , NL (I in ) = c1 I in + c1 I in + c1 I in + c1 I in + c1 I in (2)
2 6 24 120
When using subcarrier modulation, the laser is generally biased at a certain
Ibias. We will call the corresponding amount of bias current above threshold as
Iin,bias = Ibias(t) - Ith(t). Note that c1 is available as a read-only parameter in
the laser dialog window, under the name “P-I Slope [mW/mA]” (see
Advanced Laser Parameters, below).
In [2], page 488, formulae (5) and (6) express CSO and CTB for multi-
subcarrier laser modulation. We report them here for convenience:
2 L'2 2
L' ' L0 m
CSO idB = 10 log10 (3)
Ci 2
4 L'3 2
L' ' ' L20 m 2
CTBidB = 10 log10 (4)
Ci 3
In these expressions, Ci2 and Ci3 represent the second and third-order inter-
modulation products, respectively, present at a certain frequency. They are
functions of the frequency values of the comb of subcarriers input to the
lasers and of the probed frequency. Please refer to theory for their
expressions.
The definitions of the other terms appearing above, and their expressions
after (2), are as follows:
c2 2 2 c c c
L0 = Pout , NL (I in ,bias ) = c1 I in ,bias + c1 I in ,bias + 3 c13 I in3 ,bias + 4 c14 I in4 ,bias + 5 c15 I in5 ,bias (5)
2 6 24 120
∂Pout , NL (I in ) c3 3 2 c c
L' = = c1 + c2 c12 I in ,bias + c1 I in ,bias + 4 c14 I in3 ,bias + 5 c15 I in4 ,bias (6)
∂I in I in = Iin ,bias
2 6 24
∂ 2 Pout , NL (I in ) c4 4 2 c
L' ' = = c2 c12 + c3c13 I in ,bias + c1 I in ,bias + 5 c15 I in3 ,bias (7)
∂ 2 I in 2 6
I in = I in ,bias
∂ 3 Pout , NL (I in ) c5 5 2
L' ' ' = = c3c13 + c4 c14 I in ,bias + c1 I in ,bias (8)
∂ 3 I in 2
I in = Iin ,bias
Finally, m is the modulation index. Assuming that all subcarriers have indeed
the same m, its definition is:
Limitations
The main limitations of this model lie in the assumption of a uniform
longitudinal distribution of both carriers and field and in the modeling of gain
as a linear function of the carrier density. No modeling specific to MQW (Multi
Quantum Well) structures has been added. Yet, this model is a good
compromise between complexity and realism and can be effectively used in
most practical cases. The model is also based on a single-longitudinal mode
assumption. The simulation of lasers that may have multiple longitudinal
modes is therefore inaccurate. Fabry-Perot lasers can be simulated with this
model only for those structures or operating regimes where a single
longitudinal mode is clearly dominant.
Current parasitic and current leakages are not directly accounted for. Modern
lasers suffer relatively little from these impairments. In general, the typical
residual effects of electrical parasitic can be easily modeled in OptSim placing
a single or double-pole electrical filter on the laser injection current. Current
leakage may be responsible for some non-linearity in the P/I curve. It can be
fairly accurately accounted for by means of the non-linearity coefficients cj's.
This is more practical and efficient than implementing a leakage model that
needs microscopic parameters of the internal structure, that are generally
difficult to characterize and widely variable.
A further limitation is that thermal effects are not taken into account.
However, the typical time-scale of complex Optical System simulations is tens
to hundreds of ns. Over this time-scale, thermal effects would not have time
to develop, even if implemented in the code.
Inputs Outputs
Electrical: the injection current [A] Optical: the generated optical field
Electrical: the instantaneous output
power [mW]
Electrical: the instantaneous field
frequency [GHz]
1 vg 1
= ln (11)
τ out L r
where vg is the lasing mode group velocity, L is the cavity length and r is the
field facet reflectivity (note: the facet power reflection coefficient would be
|r|2). For a DFB laser the calculations to derive τout are more complex due to
the action of the grating. Please see [3] for reference. Allowed values: min:
0; max: -.
Carrier Density at Transparency n0 [m-3]: the carrier density at which a
perfect balance between stimulated emission and stimulated absorption
occurs. Allowed values: min: 0; max -.
Gain Constant a [m2]: the linear slope of the optical gain g versus the
carrier density n. The expression of the optical gain, without gain saturation,
is: g = a(n-n0). Allowed values: min: 0; max: -.
Optical Gain vs. Photon Number Slope GP [s-1]: accounts for non-linear
gain saturation. The expression of the complete optical gain, expressed as a
rate of difference between stimulated emission and stimulated absorption, is:
G = Γv g a(n − n0 ) − GP P (12)
τ e = (Anr + Bn + Cn 2 ) [s ]
−1
(13)
µg = cv
g
where c is the speed of light in vacuum and vg is the field group velocity in
the cavity. Allowed values: min 0; max -.
Non-Linearity Order 2 [W-1], Order 3 [W-2], Order 4 [W-3], Order 5 [W-4]: the c2, c3,
Coefficients c4, c5, coefficients of the series expansion of the non-linear mapping law (1)
between the output power as generated by the rate equations and the actual
output power of the laser simulation model. It can be used to impose a
controlled amount of non-linearity on the laser P/I curve.
Warning
Make sure that the dimensions of the cj’s are according to the specification
above. The specification assumes that Pout,NL in equation (1) above has the
dimension of [W].
dN b I N b N b N w
= − − +
dt q τ cap τ n τ esc
dN w N b N w N w N − Nt
= − − − g op w S
dt τ cap τ esc τ n 1+ ε pS
dS N − Nt S N
= g op w S− +β w
dt 1+ ε pS τp τn
dφ α
= g op ( N w − N wr ) + g ob ( N b − N br )
dt 2
S (t )ηhv
p(t ) =
τp
where:
Γg 0
g op =
Vw
Γε
εp =
Vw
gb
g ob = (1 − Γ)
Vb
The meaning of the rate equation parameters is the following:
1. g0 [m3/s] is the differential gain in the active region, i.e. in the well (gain
slope)
2. Γ [-] is the optical confinement factor
3. Vw [m3] is the volume of the active region
4. ε [m3] is the gain saturation parameter
5. Vb [m3] is the volume of the SCH region
Γg 0
6. g op = [s-1] is the “normalized gain slope”
Vw
Γε
7. εp = [-] is the “normalized gain saturation factor”
Vw
8. S [-] is the number of photons in the cavity
9. τesc [s-1] is the thermionic emission lifetime
10. τcap [s-1] is the carrier capture time
11. Nt [-] is the carrier number at transparency
12. τp [s] is the photon lifetime
13. β [-] is the spontaneous emission factor
14. I(t) [A] is the injected laser current
15. q [C] is the electron charge
16. α [-] is the linewidth enhancement factor
gb
17. g ob = (1 − Γ) [s-1] is the composed parameter where
Vs
gb = (2πν0/µeff)∗dnb/dNb with nb the SCH refractive index; ν0 the laser
emission frequency; µeff the effective group index of the lasing mode
18. φ(t) [-] is the phase of the laser field
19. τn [s-1] is the carrier lifetime. In order to take into account thermal effects
we use the following expression: τn = τn0 e-KtI where τn0 [s-1] is the carrier
lifetime at the reference temperature; Kt is a thermal constant
20. η [-] is the composed parameter Kcη0 where η0 is the quantum differential
efficiency and Kc takes into account other loss factors between the laser
output and measurement instrument input
21. h [Js] is Planck’s constant
22. τn0 [s] is the bimolecular recombination lifetime at a reference
temperature
23. Kt [A-1] is the thermal constant, this parameter includes thermal effects
on the AM response model, since some lasers exhibit a strong nonlinear
dependence on temperature [6].
The parameters β, gop, Nt, τn0, τcap, τesc, τp, εp, η, α, gob, Kt are obtained by
fitting the data from a measurement data file that contains the measured
curves from a realistic laser component. This file must be written following
the grammar rules described in section 2.1.4.
The curves stored in the file may be:
small-signal amplitude modulation (AM) curves
power vs current (P-I) curve
small-signal frequency modulation (FM) curves
Relative Intensity Noise (RIN) curves
Fiber Transfer Function (FIB) curves (AM Curves with Dispersive Fiber)
Laser Linewidth (LINE) curve
Not all the curves listed above must necessarily be present in the data file.
AM and P-I curves are mandatory while the other are recommended (but not
mandatory). Please note that to fit the values of gob, α and β parameters the
following rules must be observed.
if you do not supply FM Curves or FIB curves you must supply explicitly in
the data file the value of the gob parameter
if you do not supply FM or FIB or LINE curves you must supply explicitly
the value of the α parameter
if you do not supply FM or FIB or RIN or LINE curves you must supply
explicitly the value of the β parameter
See section 2.1.4 for the grammar rules of the data file.
When a parameter is explicitly written in the data file the corresponding
fitting procedure will not be performed. The explicit value will be used in the
time domain integration of the laser rate equations.
In order to save time, if in a project there are many Custom MQW Laser
components with the same measurement data file, the fitting procedure is
executed only once.
Since fitting results are only weakly dependent on laser emission frequency, if
in a project there are two or more lasers with the same measurement data
file but with different emission frequencies, then the fitting procedure is
redone only if the frequency difference between one laser and another is
greater than 3 THz.
Fitting is done on the following curves.
AM Curves
The AM curves are the small signal Amplitude Modulation response of the
laser [1,2] obtained at different bias currents. Such currents must be
distributed uniformly in the range Ith-Imax. We suggest 4 curves at least. For
example, if the threshold current Ith is 16 mA and the maximum current Imax
is 90 mA you can take the 4 curves at 20, 40, 60 and 80 mA.
The data to be fitted are frequencies in GHz and the corresponding AM
response in dB (20·log10(|HAM|)). There is no need for phase information. We
suggest also you take about 60-250 points for each curve.
AM curves are fitted on the following analytical model, which is found by
linearizing the rate equations at a given operating bias current. The curves
are normalized to the first sample found in the file (for each curve).
1
H IM (ω , I b )
( jωτ RC + 1)
H AM = (1)
H IM (0, I b )
where
β
Qn +
ηhv0 τn
H IM (ω , I b ) =
qτ cτ p 1 1 β 1
( jω + + )Q p (Qn + ) + R(ω ) jω − Q p +
τc τn τn τ p
1 1 1 1 1
R(ω ) = jω + + ⋅ jω + + + Qn −
τc τn τe τn τ cτ e
g op S 0
Qn =
1 + ε p S0
g op ( N ω 0 − N t )
Qp =
(1 + ε p S 0 ) 2
S0 is the number of photons in stationary conditions:
− b − b 2 − 4 ac
S0 =
2a
with
1 τ c τ c τ c
a=− 1 + + ε + 1 + τ n g 0
τ p τ e τ n τ n
b=
Ib
(τ n g op + β spε p ) − 1 + τ c + τ c N t g op + 1 + 1 + τ c β sp N t g op
q τ e τ n τ p τn
Ib
c = β sp
q
while Νω0 is
Ib τ c
+ 1 + N t Q n
q τ n
Nω0 =
τc τc 1 τc
1 + + + 1 + Q n
τe τn τn τn
P-I Curve
The P-I curve is the power vs. current characteristic of the laser. The fitting is
based on the model
S 0ηhv
P0 =
τp
The power values must be in mW and the current value in mA. Note that the
attenuation present between the laser chip and the measurement equipment
is accounted for through the parameter η that need not be supplied by the
user. The measurement can be taken on either temperature-stabilized or
uncooled lasers. The static increase in temperature caused by the higher bias
current is taken into account in the bestfit procedure. At run-time, however,
thermal effects are not included in the simulation and the model is in fact
“adiabatic”.
RIN Curves
The RIN curves are the Relative Intensity Noise spectra of the laser [6].
These curves are not mandatory but they help estimating the spontaneous
emission factor β correctly. See the beginning of this section for more details.
Bestfit is done against an approximated RIN expression that takes into
account the diffusion coefficient Dss only. This approximation is reasonable
because the other noise diffusion coefficients do not appreciably contribute to
the RIN.
2
1 1 1 1 1
jω + + jω + + + Qn −
2 Dss τ c τ n τe τn τ cτ e
RIN (ω ) = 2
S 0 1 1 β 1
( jω + + )Q p (Qn + ) + R(ω ) jω − Q p +
τc τn τn τ p
βN w0 S0
Dss =
τn
The data to be fitted are frequencies in GHz and the corresponding RIN
curves in 1/Hz in dB (10·log10(RIN)).
FM Curves
The FM curves are the small-signal Frequency Modulation response of the
laser. [8]. We encourage you to supply this data even if you only had a few
measured points per curve. In fact, even just one point per curve supplies
important information.
The FM curves should not be measured in the low frequency range (<100
MHz) where dynamic thermal effects are important. These effects cause a
typical dip in the laser FM response. Since dynamic thermal phenomena are
not taken into account in the fitting procedure they could cause convergence
problem. FM curves are fitted on the following analytical model.
δv(ω ) αg opτ p
H FM (ω ) ≡ = R(ω ) H IM (ω , I b ) +
δi(ω ) 4πηhv
τc
τ p g ob τe
+ R(ω ) H IM (ω , I b ) +
2πηhv τ
1 + c + jωτ c
τn
τc
g ob q
+ (2)
2π τ c
1 + + jωτ c
τn
where HIM(ω,Ib) and R(ω) are defined in the “AM Curves” section.
The data to be fitted are frequencies in GHz and the corresponding HFM
response in dB (20·log10(|HFM|)) where HFM is in GHz/mA.
2S0 α
H IMfiber = cos(ω 2 F ) − g op A +
jω 2
A 1 1 sin(ω 2 F )
+ g ob + (3)
τ e qH IM (ω ) jω + +
1 1
τ c τ n
L λ2 D
F= (4)
4πc
where L is the fiber length, λ the wavelength, c the light speed and D fiber
dispersion.
1
jω − Q p +
τp
A=
β
Qn +
τn
The data to be fitted are the frequencies in GHz and the corresponding HIMfiber
responses in dB (20·log10(|HIMfiber|)). The measurements must not include the
notches of the function, you must collect the data up to the first notch
(excluded).
Laser Linewidth
Fitting is also done on laser linewidth [6]. One or more linewidth values,
taken at different bias currents, can be supplied. The analytical model that is
bestfit to the given data is the following.
1
∆ν = S φ (ω )
2π
where Sφ(ω) is frequency noise spectral density obtained from
Sφ (ω ) = Sφ 1 + Sφ 2 + Sφ 3 + Sφ 4 (5)
where
2
QpQ X 2
Sφ 1 = 2 DNwNw 1 − (6)
kB k 2
Qp X 2 QpQ
Sφ 2 = −4 DNwS 1 − (7)
k 2 B kB
Q p2 X 2
Sφ 3 = 2 DSS (8)
k 2B2
Sφ 4 = 2 Dφφ (9)
with
1 1 1
k = jω − + + + Qn (10)
1 1 τe τn
jω + + τ eτ c
τc τn
QQ p 1
B = jω − Q p + + (11)
k τp
α g ob
X = g op + (12)
2 1 1
τ e jω + +
τc τn
β
Q = Qn + (13)
τn
with DSS, DNwNw, DNwS, Dφφ diffusion coefficients related to the noise terms.
DSS = RspS0
Dφφ = Rsp/4S0
DSφ = DNwφ = 0
DNwNw = RspS0 + Nw0/τn
DSNw = -RspS0
with
Rsp = βNw0/τn
Linewidths must be supplied in MHz, at a given bias current specified in mA.
Inputs Outputs
Electrical: the injection current [A] Optical: the generated optical field
Note
The file must be saved in the same directory where the project is saved.
Grammar
Notation used is EBNF (Extended Backus Naur Form)
[ item ] stands for optional item occurring only once
{ item } stands for optional item occurring any number of times
( item_a | item_b) stands for items that can occur as alternative
Bold stands for language keywords
File_Header ::=
{beta beta}
{gob gob}
{alpha disp}
Measure_Header ::=
MeasureType (AM | PI | FM | RIN | FIB | LINE)
PointsNum num
I current
{Lenght length
Disp disp}
Description
grammar_version ::= 1
File Header The File_Header rule contains those parameters, either gob, α or β, that must
be explicitly supplied by the user when the user does not provide those
measurement curves that are needed to obtain them through the fitting
procedure (see below for an explanation of when this happens).
In the following table the quantities used for each measure are listed:
Measure Keyword X value [Unit] Y value [Unit] Mandatory
AM curves AM Frequency [GHz] 20·log10(HAM) [dB] Yes
PI curve PI current [mA] Power [mW] Yes
RIN curves RIN Frequency [GHz] 10·log10(RIN) [dB{1/Hz}] NO
FM curves FM Frequency [GHz] 20·log10(HFM) [dB] NO
FIB curves FIB Frequency [GHz] 20·log10(HImfiber) [dB] NO
LINE curve LINE Current [mA] Linewidth [MHz] NO
Example
OptSimOli 1
gob 45125.7
StartMeasures
MeasureType AM
PointsNum 201
I 20
0.1980 0
0.2970 -0.0605
...
MeasureType AM
PointsNum 201
I 40
0.1980 0
0.2970 -0.0039
...
MeasureType PI
PointsNum 101
0 0
0.8000 0.0415
8000 0.0992
...
2.1.5 VCSEL
This block models a vertical-cavity surface-emitting laser (VCSEL) directly
modulated with an electrical signal. It uses the electrical current injected into
the laser’s optical cavity to solve the laser rate equations for the optical
output. Important VCSEL behaviors such as spatial hole burning, lateral
carrier diffusion, thermally dependent gain, and thermal carrier leakage are
all accounted for.
The driving source consists of the electrical signal input into the model. The
VCSEL cavity is modeled via a simplified current-voltage (IV) relationship and
spatially independent VCSEL rate equations.
Driving Source
The VCSEL is driven by the electrical signal at its input.
The input signal is assumed to come from an ideal current source. In other
words, the laser is assumed to be undergoing direct-drive modulation.
VCSEL Cavity
Both electrical and optical effects are modeled within the VCSEL cavity.
Electrical
The electrical model of the VCSEL cavity is that of a simplified diode IV
relationship, consisting of a series resistance Rd (Rd) and turn-on voltage Von
(Von). During solution of the cavity current I, the model ensures that
negative currents are effectively limited to zero.
Rate Equations
At the core of the VCSEL block are spatially independent semiconductor laser
rate equations, which determine the optical output in response to the cavity
current I [10]. Relative intensity noise is modeled via a constant value RIN
(RIN), and the optical emission frequency is set by λ (wavelength). The
model rate equations are based on the following single-mode spatially
dependent equations [10]:
r r r
∂N (r , t ) ηi I (r , t ) N (r , t ) I (N , T )
2
r r Leff 2 r
= − − G (r , t )S (t )ψ (r ) + ∇ N (r , t ) − l (1)
∂t q τn τn q
∂ S (t ) S (t ) β 1 v 1 v v
=− + ⋅ ⋅ ∫ N (r , t )dv + ⋅ ∫ G (r , t ) S (t )ψ (r )dv (2)
∂t τ p τn V V V
V
where σ1 is the first nonzero root of J1(x), and R is the active-layer effective
radius. If we further assume a uniform current distribution and linear gain,
we can eliminate the explicit spatial dependence from (1) and (2), thereby
obtaining the following spatially independent VCSEL rate equations:
dN0 ηi I N 0 G (T ) ⋅ [γ 00 ( N 0 − Nt (T )) − γ 01 N1 ] I (N ,T )
= − − ⋅S − l 0 (4)
dt q τn 1+ εS q
dS S β N 0 G (T ) ⋅ [γ 00 ( N 0 − N t (T )) − γ 01 N1 ]
=− + + ⋅S (6)
dt τp τn 1+ εS
Pout = k f S (7)
In (4)-(7), the spatial dependence of the gain is now accounted for via the
overlap coefficients γ00 (gam00), γ01 (gam01), φ100 (phi100), and φ101
(phi101), while diffusive effects are taken care of via hdiff (hdiff), which is
2
equal to (σ1Leff / R ) . The thermal dependence of the gain is taken into
account via the thermally dependent gain constant G(T) and transparency
number Nt(T), while gain saturation is modeled via ε (e), the gain saturation
factor. Furthermore, the leakage Il is now a function of N0 (the average
carrier number), as opposed to N. Finally, the photon number S is converted
to an output power via the output-power coupling coefficient kf (kf).
Mode-Carrier Overlap
If the transverse mode profile ψI is normalized such that
∞
2
⋅ ∫ ψ (r )r ⋅ dr = 1 (8)
R2 0
R
2
2 ∫ 0 i
γ 0i = ⋅ J (σ r / R ) ⋅ ψ (r )r ⋅ dr (9)
R 0
R
2
φ10i = ⋅ ∫ J 0 (σ i r / R ) ⋅ J 0 (σ1r / R) ⋅ ψ (r )r ⋅ dr (10)
R 2 J 02 (σ1 ) 0
a g 0 + ag1T + ag 2T 2
G (T ) = Go ⋅ (11)
bg 0 + bg1T + bg 2T 2
where Go (Go) is a gain constant, Ntr (Ntr) is a transparency number, and ag0-
ag2 (ag0-ag2), bg0-bg2 (bg0-bg2), and cn0-cn2 (cn0-cn2) are fitting
parameters. Generally, the gain constant will be peaked about some optimal
temperature value, as a result of the temperature-dependent mismatch
between lasing wavelength and gain peak. An example of the gain constant
based on the model’s default values is illustrated in Fig. 4. The transparency
number generally increases with temperature.
Figure 2.1 Sample thermally dependent gain constant based on default model
values
− a + a N + a2 N 0T − a3 / N 0
Il ( N 0 , T ) = I lo ⋅ exp 0 1 0 (13)
T
where Ilo (Ilo) is the leakage current factor, and a0-a3 (a0-a3) are fitting
parameters. This expression accounts for the interdependence of the carrier
number and temperature in determining the total leakage current. An
example of (13) for various temperatures is illustrated in Fig. 5.
Figure 2.2 Sample thermal leakage current based on default model values
dT
T = To + ( ItotV − Pout ) ⋅ Rth − τ th ⋅ (14)
dt
where To (To) is the ambient temperature, Itot is the total current flowing
through the VCSEL, Rth (Rth) is the device thermal impedance, and τth (tth) is
the thermal time constant.
Polarization
The laser emits an output field polarized along the X axis, with no
corresponding Y-polarized component. The polarization of the field can be
changed using the polarization rotator component.
Properties
Inputs Outputs
Electrical: the injection current [A] Optical: the generated optical field
In this section you can choose a predefined laser model clicking on the model
name in the Available lasers list. The list shows all the laser model files
(with extension obp) currently stored in the library directories or in the
current project directory. With New you can define a new laser model having
the actual parameters of the laser.
The parameter values of all the sections (Laser Behaviour, Physical,
Empirical, Advanced, Spectral Estimation) depend on the laser model
selected. If you modify a value when a predefined laser model is selected, the
selection goes automatically to This laser only, then all the values can be
modified and saved as a new library model.
The Parameter values of the Behaviour section are calculated as a function of
the parameters supplied in all the other sections. Many of these quantities
can be calculated only at a steady state operating point. Therefore you are
requested to input values in the Test Current Value field and in the Test
Temperature Value field. Please note that to reflect any changes in the
laser input parameters, you have to click on the Update button.
Center emission frequency [nm] or [THz]: the referring emission
frequency of the source. This value is used for the calculation of the most of
the system-relevant quantities in the behaviour section.
Insert Test Current Value [mA]: the input value of injection current that
sets the operating point for the calculation of the most of the system-relevant
quantities. It must be greater than the Threshold Current, otherwise
meaningless data will be displayed when you click on the Update button.
Allowed values: min Threshold Current; max -.
Threshold Current [mA]: the calculated laser threshold current.
Linear Output Power [mW]: it is the output power into a fiber pigtail at the
Test Current Value. It is assumed that the power produced by the laser is
equally split between the two output facets.
P-I Slope [mW/mA]: the slope of the Power/Current curve as measured at
the output of a fiber pigtail, under the same assumptions as defined above for
the Linear Output Power.
Turn-On Delay [ns]: the time needed for the carrier density to reach the
threshold density, using the injection current specified in Test Current
Value. Note that the turn-on delay only says how long it takes for the laser
to reach threshold gain conditions, but not how long it takes for the laser to
reach steady state. After turn-on, a series of oscillations ensues, that may
last for several ns.
Overlap calculation: Sets whether the overlap has to be calculated from the
overlap parameter or assigned equal to the overlap coefficients; Allowed
values: No; Yes; default No.
Overlap parameter: Allowed values: min 0.01; max 10; default 1.
Overlap coefficient for N0 and S rate equations, gam00 [-]: Allowed
values: min -; max -; default 1.
Overlap coefficient for N0 and S rate equations, gam01 [-]: Allowed
values: min -; max -; default 0.37978.
During SPT Simulation the Rate Equations are not solved and the laser is
treated as a light source emitting at its nominal Laser Center Emission
Frequency (see above), with a certain Average Output Power, which must be
supplied by the user (see directly below).
Average output power for Spectral Propagation [dBm] or [mW]:
accounts for the mean laser output power in SPT simulation. Allowed values:
min 0 mW; max -.
SPT Behavior
Pulse generators are considered as they generate a single tone at the nominal
center emission frequency of the source. Therefore, in the optical spectrum a
single row is placed. Its level is equal to the average output power. The
spectrum of the different pulse shapes is not taken into account. The
generated spectrum is always a single-row one, independent of the choice on
the pulse-shape. This approximation is justified by the SPT simulation
technique, which considers channels in terms of average power and spectral
position only.
Parameters
Name: the optional name that appears in the component label.
Laser Emission Frequency [nm] or [THz]: the emission frequency of the
laser that is modulated to generate the chosen pulse train. Allowed values:
min flowSPT; max fupSPT; default f0,SPT.
Source Linewidth [MHz]: the spectrum width at half maximum, of the laser
if it was operated in CW. Allowed values: min: 0; max: -; default: 0.
Signal Shape: opens the related dialog windows for the three models,
Raised Cosine, SuperGaussian, Soliton.
Laser Phase: sets the laser phase at the start of the simulation.
Random: the initial laser phase is randomly set. This option is useful
when several lasers are used in the same model, like in WDM systems, in
order to ensure a statistical distribution for initial phases of laser sources.
Deterministic: the initial laser phase is set by the Initial Laser Phase
value.
Initial Laser Phase [rad]: set the laser phase at the start of the simulation.
Allowed values: min: 0; max: 2π; default: 0 (if the Deterministic option is
selected).
Status [-]: turns on "1" or turns off "0" the optical source. Allowed values: 0
or 1; default 1.
Optical Sources
2.2.1 Soliton
This model implements a soliton source, i.e. a source whose pulses have the
following shape:
T jCT 2
U (T ) = A sech exp 2
(1)
T0 2T0
where C is the chirp parameter of the pulse and T0 is related to the TFWHM
(Full Width Half Maximum time) of the pulse by:
TFWHM
T0 =
(
2 ln 1 + 2 ) (2)
- Optical
Parameters
Time between two pulses [ps]: the pulse rate. Allowed values: min 0;
max Tsim (Total Simulated Time Span simulation parameter); default Rb-1.
Delay of the pulse train [ps]: a delay relative to the beginning of the
simulation. During this delay pulses are not outputted. Allowed values: min 0;
max Tsim; default 0.
TFWHM [ps]: it is the Full Width Half Maximum time. It is the time between
two points at a half of the peak power. Allowed values: min 0; max Tsim;
default 30.
T0 [ps]: it is the parameter used in the analytical expression. Allowed values:
min 0; max Tsim; default 17.019.
Peak Power [mW]: the peak power of the source. Allowed values: min 0;
max -; default 10.
Average Power [mW]: it is the average power of the pulse train. Allowed
values: min 0; max -; default 1.70.
Chirp Factor: Allowed values: min -; max -; default 0.
Note that the pulse generators have been implemented using the chirp factor
C that relates the phase to a function of time. This is the definition used in
[5], and it is different from the chirp parameters α defined for the
modulators. The α parameter refers to the definition of phase that depends
on the power, as it is defined in [4]. Note that the meaning of the two
definitions is completely different, therefore, the impact on system
performance is also different. For a first order Gaussian pulse only α = -C.
Comparing the implemented formulas with the ones presented in [5], you
may note a difference in the sign of C. The reason is that in [5] the
eigenfunction for the analytic signal representation is e-jωT. Therefore, the
instant frequency is
1 ∂Φ (T )
f =− ,
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
U (T ) = Af (T )e jΦ (T ) (1)
U (T ) = A f (T )e jΦ (T ) (2)
T (1 + α )
0 , T > L
2
1 T − T T (1 − α ) T (1 + α )
f (T ) = 1 − sin π L
, L
≤T ≤ L (3)
2 αTL 2 2
T (1 − α )
1 , T < L
2
where
TS
TL = (4)
(1 + α )
α is the roll-off parameter and TS the duration of a bit.
A supplemental delay parameter allows to choose when enable an optical
output: no optical power is present at the output before the delay time.
Inputs Outputs
- Optical
Parameters
Resulting Raised Cosine Shape is in: sets what to shapes, the Amplitude
or the Power of the signal.
Time between two pulses [ps]: the pulse rate. Allowed values: min 0;
max Tsim (Total Simulated Time Span simulation parameter); default Rb-1.
Delay of the pulse train [ps]: a delay relative to the beginning of the
simulation. During this delay pulses are not outputted. Allowed values: min 0;
max Tsim; default 0.
Peak Power [mW]: the peak power of the source. Allowed values: min 0;
max -; default 10.
Average Power [mW]: it is the average power of the pulse train. Allowed
values: min 0; max -; default 6.66.
Roll-off: Roll-off parameter of the pulses, see (4). Allowed values: min 0;
max 1; default 0.5.
Chirp Factor [rad/mW]: Chirp factor that relates pulse phase [rad] and
power [mW] with the following law. Allowed values: min -; max -; default 0.
2
T
T− S
C 2
φ (T ) = (5)
2 TL
Note that the pulse generators have been implemented using the chirp factor
C that relates the phase to a function of time. This is the definition used in
[5], and it is different from the chirp parameters α defined for the
modulators. The α parameter refers to the definition of phase that depends
on the power, as it is defined in [4]. Note that the meaning of the two
definitions is completely different, therefore, the impact on system
performance is also different. For a first order Gaussian pulse only α = -C.
Comparing the implemented formulas with the ones presented in [5], you
may note a difference in the sign of C. The reason is that in [5] the
eigenfunction for the analytic signal representation is e-jωT. Therefore, the
instant frequency is
1 ∂Φ (T )
f =− ,
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
2.2.3 Supergaussian
This model implements a supergaussian source, i.e. a source whose pulses
have the following shape:
1 m jC T 2 m
U (T ) = A exp − (1)
2 T0
where C is the chirp parameter of the pulse, m is the order and T0 is related
to the TFWHM (Full Width Half Maximum time) of the pulse by:
TFWHM
T0 = (2)
22 m ln 2
The pulse has a peak power set to A2.
A supplemental delay parameter is allowed to choose with enabled optical
output: no optical power is present at the output before the delay time.
Inputs Outputs
- Optical
Parameters
Time between two pulses [ps]: the pulse rate. Allowed values: min 0;
max Tsim (Total Simulated Time Span simulation parameter); default Rb-1.
Delay of the pulse train [ps]: a delay relative to the beginning of the
simulation. During this delay pulses are not outputted. Allowed values: min 0;
max Tsim; default 0.
TFWHM [ps]: it is the Full Width Half Maximum time. It is the time between
two points at a half of the peak power. Allowed values: min 0; max Tsim;
default 30.
T0 [ps]: it is the parameter used in the analytical expression. Allowed values:
min 0; max Tsim; default 18.01.
Peak Power [mW]: the peak power of the source. Allowed values: min 0;
max -; default 10.
Average Power [mW]: it is the average power of the pulse train. Allowed
values: min 0; max -; default 1. 59.
Supergaussian Order: order of the supergaussian function. Allowed values:
min 1; max -; default 1.
Chirp factor: Allowed values: min 0; max -; default 0.
Note that the pulse generators have been implemented using the chirp factor
C that relates the phase to a function of time. This is the definition used in
[5], and it is different from the chirp parameters α defined for the
modulators. The α parameter refers to the definition of phase that depends
on the power, as it is defined in [4]. Note that the meaning of the two
definitions is completely different, therefore, the impact on system
performance is also different. For a first order Gaussian pulse only α = -C.
Comparing the implemented formulas with the ones presented in [5], you
may note a difference in the sign of C. The reason is that in [5] the
eigenfunction for the analytic signal representation is e-jωT. Therefore, the
instant frequency is
1 ∂Φ (T )
f =− ,
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
SPT Behavior
The optical white noise generator generates an optical noise with well-defined
flat spectral characteristics. During the SPT simulation, it generates a flat
optical spectrum whose level corresponds to the noise level defined.
Inputs Outputs
- Optical
Parameters
Name: the optional name that appears in the component label.
One-Sided Noise Spectral Density [dB{mW/THz}]: the power spectral
density of the generated noise; the notation "one-sided" means that like all
the OptSim optical signal, also the optical noise is considered as an analytic
signal spectrally distributed around the simulation central frequency, and the
"mirrored" negative frequency are not considered. Allowed values: min -;
max -; default 0.
Note
The notation dB{mW/THz} means that the power spectral density N0,lin
expressed in mW/THz must be converted to the requested unit to N0,dB , that
is the number to introduce in the dialog, in the following way:
Status [-]: turns on "1" or turns off "0" the optical source. Allowed values: 0
or 1; default 1.
Power Spectral
Density
[mW/THz]
Bw (simulation bandwidth)
N0,lin
Figure 2.1 Power spectral density of the Optical White Noise Generator
References
[1] Govind P. Agrawal and Niloy K. Dutta, Semiconductor lasers. New York :
Van Nostrand Reinhold, copyr. 1993.
[2] Jan Lipson et al., High fidelity lightwave transmission of multiple AM-VSB
NTSC signals, IEEE Trans. on Microwave Theory and Techniques, vol. 38,
no. 5, May 1990.
[3] Max Ming-Kang Liu, Principles and Applications of Optical
Communications. Irwin Book Team, London, 1996.
[4] F.Koyama and K. Iga, Frequency chirping in external modulators, J.
Lightwave Technol., vol.6, no.1, pp.87-33, Feb.1988.
[5] Govind P. Agrawal, Nonlinear Fiber Optics, 2nd edition, Academic Press,
San Diego, 1995
[6] M. M. Freire, H. J. A. da Silva, Estimation of multiple-quantum well laser
parameters for simulation of dispersion suppoted transmission systems at
20 Gbit/s, IEE Pcoceedings of Optoelectronics, Vol. 146, No. 3, pp. 93-98,
1999.
[7] M.M. Freire, L.N. Vicente, H.J.A. da Silva, Nonlinear Least Square
Estimation of MQW Laser Parameters from IM Response Measurements
[8] R. F. S. Ribeiro, J. R. F. da Rocha, H. J. A. da Silva, FM Response of
Quantum-Well Lasers Taking into Account Carrier Transport Effects, IEEE
Photonics Tecnology Letters, vol. 7, no. 8, pp. 857-859, 1995.
[9] R.C. Srinivasan, J.C. Cartlege, On Using Fiber Transfer Function to
Characterize Laser Chirp and Fiber Dispersion, IEEE Photonics Tecnology
Letters, vol. 7, no. 11, pp. 1327-1329, 1995.
[10] P. V. Mena, J. J. Morikuni, S.-M. Kang, A. V. Harton, and K. W.
Wyatt, “A comprehensive circuit-level model of vertical-cavity surface-
emitting lasers,” Journal of Lightwave Technology, 17, 2612 (1999).
[11] P. V. Mena, J. J. Morikuni, and K. W. Wyatt, “Compact
representations of mode overlap for circuit-level VCSEL models,”
IEEE/LEOS Annual Meeting Conference Proceedings, 234 (2000).
[12] J. W. Scott, R. S. Geels, S. W. Corzine, and L. A. Coldren,
“Modeling temperature effects and spatial hole burning to optimize
vertical-cavity surface-emitting laser performance,” IEEE Journal of
Quantum Electronics, 29, 1295 (1993).
[13] N. Bewtra, D. A. Suda, G. L. Tan, F. Chatenoud, and J. M. Xu,
“Modeling of quantum-well lasers with electro-opto-thermal interaction,”
IEEE Journal of Selected Topics in Quantum Electronics, 1, 331 (1995).
[14] M. Born and E. Wolf, Principles of Optics, 7th. Ed. (Cambridge University
Press, Cambridge, 1999).
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the five waveform shapes, Bias,
Sinusoidal, Sawtooth, Square and Clock.
- Electrical
Eout(t)
Level
Parameters
Level [AU]: it is the constant level value of the signal to be generated.
Allowed values: min -; max -; default 1.
Eout (t ) = A ⋅ cos (2 π f t + ϕ )
Inputs Outputs
- Electrical
Eout(t)
T=1/f
A cos (ϕ)
A
Parameters
Frequency [GHz]: it is the sinusoid frequency f. Allowed values: min 0; max
BWVBS /2; default 10.
Amplitude [AU]: it is the sinusoid peak amplitude A. Allowed values: min -;
max -; default 1.
Phase [rad]: it is the sinusoid starting phase ϕ, expressed in radians.
Allowed values: min 0; max 2π; default 0.
- Electrical
Eout(t)
A=T∗Slope
t
Edge
T=1/BitRate
Parameters
Bit Rate Nominal [Gbit/s]: it is the bit-rate of the sawtooth signal. Allowed values:
min 0; max BWVBS/2; default 10.
Simulated [Gbit/s]: it is the actual simulated bit-rate with the current VBS
bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it is the number of time-sample per bit referred at the
actual Simulated bit-rate.
Period Edge with respect to Simulation Start Time [bit fraction]: it is
the shift of the bit edge respect to the simulation starting time. This shift is
given as a fraction of the bit duration, Edge/T in Figure 3.3. Allowed values:
min 0; max 1; default 0.
Slope [AU/ns]: it is the ramp slope of the sawtooth signal. In Figure 3.3 is
shown the relationship between maximum signal amplitude and slope.
Allowed values: min -; max -; default 1.
- Electrical
Eout(t)
Level 1st part
T=1/BitRate
t
nd
Edge T1st Level 2 part
Figure 3.4 Square wave signal generator: generated signal with parameters
Parameters
Bit Rate Nominal [Gbit/s]: it is the bit-rate of the generated square wave signal.
Allowed values: min 0; max BWVBS/2; default 10.
Simulated [Gbit/s]: it is the actual simulated reference bit-rate with the
current VBS bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it is the number of time-sample per bit referred at the
actual Simulated bit-rate.
Period Edge with respect to Simulation Start Time [bit fraction]: it is
the shift of the bit edge respect to the simulation starting time. This shift is
given as a fraction of the bit duration, Edge/T in Figure 3.4. Allowed values:
min 0; max 1; default 0.
Duty Cycle [%]: it is the duty cycle of the square wave. Duty cycle is
defined as the ratio, expressed in percentage, between the duration of the
first part of the bit and the total bit duration: T1st/T in Figure 3.4. Allowed
values: min 0; max 100; default 50.
Level for the 1st part of the period: it is the signal level for the first part
of the bit duration. Allowed values: min -; max -; default 0.
Level for the 2nd part of the period: it is the signal level for the second
part of the bit duration. Allowed values: min -; max -; default 1.
- Electrical
Eout(t)
T=1/BitRate
1
Edge
Parameters
Bit Rate Nominal [Gbit/s]: it is the bit-rate of the generated clock signal. Allowed
values: min 0; max BWVBS/2; default 10.
Simulated [Gbit/s]: it is the actual simulated reference bit-rate with the
current VBS bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it is the number of time-sample per bit referred at the
actual Simulated bit-rate.
Period Edge with respect to Simulation Start Time [bit fraction]: it is
the shift of the pulse respect to the simulation starting time. The shift is given
as a fraction of the bit duration, Edge/T in Figure 3.5. Allowed values: min 0;
max 1; default 0.
Parameters
Rb [THz]: it is the clock rate.
TSYNC [ps]: the instant within the symbol interval at which the clock pulse
must be generated, if it is not a multiple of DELT the pulse will be generated
the following sample with the value indicated above. Allowed values: min 0;
max 1/Rb; default -.
Parameters
TSI [ps]: the interval between two impulses. Allowed values: min 0 (single
impulse); max -; default -.
AMP: the amplitude of the generated signal.
TDEL [ps]: the starting instant [ps] of the delta train; or the instant of the
single impulse generation (when TSI = 0). The time is given in number of
samples (see section 1.5 for further information). Allowed values: min 0; max
-; default -.
ZTE
DELAY Delay = TCSI
ZTD
DELAY Delay = TS/2
The rest period of the VCO is equal to TS; the VCO is controlled by the input
base-band signal XC and the sensitivity of the VCO is defined by the
parameter A.
The signal produced by the VCO is used to generate the timing signal ZTT
when it crosses the zero axis with positive slope; on average, one timing
pulse is produced every NS samples.
In addition to the natural timing pulse ZTT, three other timing pulses are
generated; one of them, ZTD, is delayed by TS/2, the other two placed
symmetrically with respect of ZTD at a distance equal to ±(TS/2 - TCSI). The
time location is the following.
TCSI time
TS/2
TS-TCSI
Inputs Outputs
Parameters
TS [ps]: the average period TS between the pulses.
TCSI [ps]: it is the time offset. Allowed values: min 0; max TS/2; default 0.
A [-]: it is the VCO sensitivity.
Note
This component is used, in particular, to generate all the timing signals used
in the Early-Late-Gate Symbol Synchronizers ELGSY1 (section 23.28) and
ELGSY2 (section 23.29), and in the In-phase Mid-phase Symbol
Synchronizers IMPSY1 (section 23.30) and IMPSY2 (section 23.31).
Parameters
F0 [THz]: the center frequency of the oscillator.
K1: it is the output amplitude factor; the amplitude of the output signal is
K1⋅ 2 .
K2: it is the VCO sensitivity; the phase of the output signal is PHASO +
∫
K2⋅ X .
Ein: input base-band signal Eclk: the output electrical clock (see
section 1.6)
Parameters
LSLOPE: the approaching slope of the input signal. If it is set to +1, the
timing pulse is generated when the input signal crosses the zero axis with
positive slope; if it is set to -1, the timing pulse is generated when the input
signal crosses the zero axis with negative slope. Allowed values: min -1; max
+1; default -; zero is excluded.
OFFSET: delay between the effective zero crossing and the generated output
timing pulse. Allowed values: min 0; max -; default 0.
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the four models, Gaussian,
Rayleigh, Uniform and Video Source.
( x − Mean )2
f ( x ) = exp− (1)
2σ 2
where Mean is the mean value and σ is the standard deviation of the
Gaussian distribution.
Signal spectrum is flat over the whole simulated bandwidth.
Inputs Outputs
- Electrical
Parameters
Mean [AU]: it is the mean value of the Gaussian distribution shown in
equation (1). Allowed values: min -; max -; default 0.
Standard Deviation [AU]: it is the standard deviation σ of the Gaussian
distribution shown in equation (1). Allowed values: min 0.; max -; default 1.
One-Sided Spectral Density N0 [dB{AU2/GHz}]: it is the one-sided
spectral density of the generated Gaussian distributed noise signal. Allowed
values: min -; max -; default Value related to σ = 1 with the actual VBS
bandwidth, see equations below.
The One-Sided Spectral Density N0 is related to the Standard Deviation
by:
1.6σ 2
N 0 = 10 ⋅ log10 .
BWVBS
x x2
f ( x ) = 2 exp− 2 for x > 0 (1)
S 2S
where S is the parameter of the distribution. Notice that the mean value
is S π 2 while the standard deviation is S 2 −π 2 .
Signal spectrum is flat over the whole simulated bandwidth.
Inputs Outputs
- Electrical
Parameters
S Parameter: it is the S parameter of the Rayleigh distribution shown in
equation (1). Allowed values: min 0.; max -; default 1.
- Electrical
Parameters
Mean [AU]: it is the mean value of the uniform distribution. Allowed values:
min -; max -; default 0.
Range [AU]: it is the interval around the mean value where values are
uniformly distributed. Allowed values: min 0+; max -; default 1.
- Electrical
Parameters
Maximum Absolute Value [AU]: it is the maximum absolute value of the
generated signal. It fixes the interval of values where the output signal is
uniformly distributed. Allowed values: min 0; max -; default 1.
-3dB Bandwidth [GHz]: it is the frequency where signal spectrum is -3 dB
below its DC value. Allowed values: min 0; max 0.25*BWVBS; default 10.
log10 ( N b + 1)
D= (1)
log10 2
where Nb is:
N b = int (Tsim ⋅ Rb )
Tsim is the total simulated time span (it is a global simulation parameter) and
Rb is the actual bit-rate (the value displayed by the Corresponding
Simulated Bit-Rate parameter)
Inputs Outputs
- Logical
- Electrical
5-6 6
7 18
8 16
9 48
Pseudo-Random Sequence Degree Maximum Polynomial Number
10 60
11, 13-27 176
12 144
28 108
Bit Edge with respect to simulation start time [bit fraction] [-]: when
the deterministic option is selected, it is the shift of the bit edge with respect
to the simulation start time. This shift is expressed as a fraction of the bit
time (Edge/T, see Figure 4.1). Allowed values: min 0; max 1; default 0 (if the
Deterministic option is selected).
T=1/BitRate
Edge
Starting Point: it sets how is selected the first bit of the simulation, between
the pseudo-random sequence bits.
Random: the first bit of the simulation is randomly selected. This option
is useful when several data sources are used in the same project, like in
WDM systems, in order to ensure statistical independence of the used
pseudo-random sequences.
Deterministic: the first bit of the simulation is selected inside the chosen
pseudo-random sequence giving the Starting point in the sequence
parameter.
Starting point in the sequence [-]: when the deterministic option is
selected, it sets the first bit of the simulation as the n-th bit of the chosen
pseudo-random sequence. Allowed values: min 0; max -; default 0 (if the
Deterministic option is selected).
Synchronization signal Type: it sets the type of electric synchronization
signal. This signal is synchronized at the baud rate.
Pulse: Pulse signal. Pulses amplitude is fixed to 1.
Square: 50% duty cycle square wave, the amplitudes of the two output
levels are +1 and -1.
Sine: Sine wave with peak amplitude fixed to 1.
Note
If the file does not contain enough bits to fill the whole simulation time
window Ts, the bit sequence will be repeated as many times as necessary.
If the file contains N levels, you can specify in the parameters dialog window
a number of levels M ≤ N. The first M levels in the file will be taken into
account.
- Logical
Parameters
Name: the optional name that appears in the component label.
Bit Rate Nominal [Gbit/s]: it is the bit-rate of the generated clock logical signal.
Simulated [Gbit/s]: it is the actual simulated bit-rate with the current VBS
bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it is the number of samples per bit at the actual
Simulated bit-rate.
Optical
Optical Playback, allows to upload an optical signal from an ASCII data file
containing the time domain samples. The file can be generated by a previous
OptSim simulation or using other software tools.
Optical Recorder, allows to export time domain samples of the optical signal
in an ASCII data file to be used further in a OptSim simulation or for post-
processing analysis to be made with other software tools.
Electrical
Electrical Playback, allows to upload an electrical signal from an ASCII data
file containing the time domain samples. The file can be generated by a
previous OptSim simulation or using other software tools.
Electrical Recorder, allows to export time domain samples of the electrical
signal in an ASCII data file to be used further in a OptSim simulation or for
post-processing analysis to be made with other software tools.
Logical
Logical Playback, allows to upload a logical signal from an ASCII data file
containing the time domain samples. The file can be generated by a previous
OptSim simulation or using other software tools.
Logical Recorder, allows to export time domain samples of the logical signal
in an ASCII data file to be used further in a OptSim simulation or for post-
processing analysis to be made with other software tools.
Record&Playback
Bw Bw
F0 − 4 , F0 + 4 ,
where F0 is the center frequency at which refer the optical signals stored in
the ASCII data file and Bw is the time domain bandwidth of the OptSim
simulation.
SPT Behavior
During the SPT simulation, the spectral information about the signal stored in
the ASCII file is evaluated from time samples: the Optical Playback is
considered like an optical source.
Inputs Outputs
- Optical
Parameters
Name: the optional name that appears in the component label.
File Name: it is the name of the ASCII data file where time samples of the
optical signal is stored in. It is possible to choose from the file system just
clicking on the Browse button (remember that the ASCII data file have to be
stored in the same directory of the project). It is also possible to edit the file
clicking on the Edit button: a window will be spawn with the data file loaded
in an editor. Default editors is VI for Unix and Notepad for Windows NT.
SPT Behavior
The Optical Recorder store the optical signal during the time domain
simulation. Since the SPT simulations propagates only a spectral
representation of the optical signal, the Optical Recorder does not work in the
SPT simulations. SPT simulations can be done anyway when an Optical
Recorder is present.
Inputs Outputs
Optical -
Parameters
Name: the optional name that appears in the component label.
File Name: it is the name of the ASCII data file where the optical signal will
be stored in. In parametric runs, an underscore character and the run
number are appended to the file name (e.g. opt_rec.DAT becomes
opt_rec_1.DAT, opt_rec_2.DAT, etc.).
Recording time-span: it selects the recording signal time window. With the
Whole option the whole Tsim will be considered. With the Limited option only
the time span set with the following parameters will be considered.
Start Time [ns]: it is the starting time of the signal recording. The available
time window is given by the chosen Total Simulated Time Span (Tsim).
Allowed values: min 0; max Tsim; default 0.
End Time [ns]: it is the ending time of the signal recording. The available
time window is given by the chosen Total Simulated Time Span (Tsim).
Allowed values: min 0; max Tsim; default Tsim.
Sampling Factor: it is the ratio between the simulation time domain
sampling rate and the sub-sampling rate; basically if this factor is equal to n,
one sample out of n is recorded. Allowed values: min 1; max -; default 1.
- Electrical
Parameters
Name: the optional name that appears in the component label.
File Name: it is the name of the ASCII data file where time samples of the
electrical signal is stored in. It is possible to choose from the file system just
clicking on the Browse button (remember that the ASCII data file have to be
stored in the same directory of the project). It also possible to edit the file
clicking on the Edit button: a window will be spawn with the data file loaded
in an editor. Default editors is VI for Unix and Notepad for Windows NT.
Electrical -
Parameters
Name: the optional name that appears in the component label.
File Name: it is the name of the ASCII data file where the electrical signal
will be stored in. In parametric runs, an underscore character and the run
number are appended to the file name (e.g. elt_rec.DAT becomes
elt_rec_1.DAT, elt_rec_2.DAT, etc.).
Recording time-span: it selects the recording signal time window. With the
Whole option the whole Tsim will be considered. With the Limited option only
the time span set with the following parameters will be considered.
Start Time [ns]: it is the starting time of the signal recording. The available
time window is given by the chosen Total Simulated Time Span (Tsim).
Allowed values: min 0; max Tsim; default 0.
End Time [ns]: it is the ending time of the signal recording. The available
time window is given by the chosen Total Simulated Time Span (Tsim).
Allowed values: min 0; max Tsim; default Tsim.
Sampling Factor: it is the ratio between the simulation time domain
sampling rate and the sub-sampling rate; basically if this factor is equal to n,
one sample out of n is recorded. Allowed values: min 1; max -; default 1.
- Logical
Parameters
Name: the optional name that appears in the component label.
Bit Rate Bit rate (Nominal) [Gbit/s]: the bit-rate of the logical signal. Allowed
l 0 /2 d f l 0
Logical -
Parameters
Name: the optional name that appears in the component label.
File Name: the name of the ASCII data file where the logical signal will be
stored in. In parametric runs, an underscore character and the run number
are appended to the file name (e.g. log_rec.DAT becomes log_rec_1.DAT,
log_rec_2.DAT, etc.).
Recording time-span: it selects the recording signal time window. With the
Whole option the whole Tsim will be considered. With the Limited option only
the time span set with the following parameters will be considered.
Start Time [ns]: the starting time of the signal recording. The available time
window is given by the chosen Total Simulated Time Span (Tsim). Allowed
values: min 0; max Tsim; default 0.
End Time [ns]: the ending time of the signal recording. The available time
window is given by the chosen Total Simulated Time Span (Tsim). Allowed
values: min 0; max Tsim; default Tsim.
Logical Level: the number of bits per symbol of the input signal. Allowed
values: min 1; max -; default 1.
Sampling Factor: the ratio between the simulation time domain sampling
rate and the sub-sampling rate; basically if this factor is equal to n, one
sample out of n is recorded. Allowed values: min 1; max -; default 1.
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but simply to study the system in terms of average power per
channel and noise level. The only actions of intensity modulation transmitter
that have to be taken into account for SPT purposes are the optical spectral
modifications introduced on the channel levels and on the noise level. An
exactly squared NRZ modulation presents an average power reduction equal
to 3 dB, plus the modulator excess loss: this is the assumption we made for
this component.
Inputs Outputs
- TX_output: optical
Parameters
Bit_rate [Gbit/s]: the bit rate at which the data source operates. Allowed
values: min -; max -; defaut 10.
Laser_frequency [THz]: the CW laser emission frequency. Allowed values:
min flowSPT; max fupSPT; default f0,SPT.
Laser_linewidth [MHz]: the spectrum width at half maximum. Allowed
values: min 0; max -; default: 10.
Laser_power [dBm]: the output power of the laser. Allowed values: min -;
max -; default: 0.
Modulator_chirp_factor: the chirp factor α as it is defined in [1]. Note that
a non-ideal extinction ratio induces an implicit chirping that is not taken into
account by the chirp factor. See section 8.2 for further explanation on the
modulator parameters. Allowed values: min -; max -; default 0.
Modulator_excess_loss [dB]: the value of a constant extra loss introduced
by the modulator, independently by the electrical signal. See section 8.2 for
further explanation on the modulator parameters. Allowed values: min 0.
max -; default 3.
Modulator_extinction_ratio [dB]: the ratio between the output optical
power corresponding to the maximum transmission value and the one
corresponding to the minimum transmission value. See section 8.2 for further
explanation on the modulator parameters.
PRBS_sequence_degree: It is the degree of the PRBS generating
polynomial. The period length of the corresponding pseudo-random sequence
is 2X-1 bits, where X is the degree. Allowed values: min 5; max 28; default 7.
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but simply to study the system in terms of average power per
channel and noise level. The only actions of intensity modulation transmitter
that have to be taken into account for SPT purposes are the optical spectral
modifications introduced on the channel levels and on the noise level. An
exactly squared NRZ modulation presents an average power reduction equal
to 3 dB, plus the modulator excess loss: this is the assumption we made for
this component.
Inputs Outputs
- optical_output: optical
Parameters
Bit_rate [Gbit/s]: the bit rate at which the data sources operate. Allowed
values: min -; max -; defaut 10.
Center_frequency [THz]: the center frequency of the WDM aggregate.
Allowed values: min flowSPT; max fupSPT; default f0,SPT.
Channel_Spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
Laser_linewidth [MHz]: the spectrum width at half maximum of each laser.
Allowed values: min 0; max -; default: 10.
Laser_power_dBm [dBm]: the output power of the lasers. Allowed values:
min -; max -; default: 0.
Modulator_chirp_factor: the chirp factor α as it is defined in [1]. Note that
a non-ideal extinction ratio induces an implicit chirping that is not taken into
account by the chirping factor. See section 8.2 for further explanation on the
modulator parameters. Allowed values: min -; max -; default 0.
Modulator_excess_loss [dB]: The value of a constant extra loss
introduced by the modulators, independent by the electrical signal. See
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but simply to study the system in terms of average power per
channel and noise level. The only actions of intensity modulation transmitter
that have to be taken into account for SPT purposes are the optical spectral
modifications introduced on the channel levels and on the noise level.
The 2DPSK transmitter does not introduce any contribution in term of power
budget so a single tone at the nominal center frequency of the source is
considered.
Inputs Outputs
- dpsk_t_amp_pi_output: optical
Parameters
Bitrate [Gbit/s]: the bit rate at which the data source operates. Allowed
values: min -; max -; defaut Rb.
Fc [THz]: the CW laser emission frequency. Allowed values: min flowSPT;
max fupSPT; default f0,SPT.
PRBS [degree]: the degree of the PRBS generating polynomial for the data
source. Allowed values: min 5; max 28; default 12.
Power [dBm]: the output power of the laser. Allowed values: min -; max -;
default: -12.
Swing [V]: it is the value of the output signal for the driver when a logical
one is transmitted. Allowed values: min -; max -; default +5.
Transmitter_BW [GHz]: it is the –3 dB cutoff frequency for the electrical
filter. Allowed values: min 0; max BWVBS/4; default Rb.
Vpi [V]: the input voltage swing for the optical phase modulator that
generates a π (180°) phase shift on the optical signal. Allowed values: min 0;
max -; default 5.
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but simply to study the system in terms of average power per
channel and noise level. The only actions of intensity modulation transmitter
that have to be taken into account for SPT purposes are the optical spectral
modifications introduced on the channel levels and on the noise level.
The 2DPSK transmitter does not introduce any contribution in term of power
budget so for each channel a single tone at the nominal center frequency of
the source is considered.
Inputs Outputs
- opt_out: optical
Parameters
Bitrate [Gbit/s]: the bit rate at which the data sources operates. Allowed
values: min -; max -; defaut Rb.
Center_frequency [THz]: the center frequency of the WDM aggregate.
Allowed values: min flowSPT; max fupSPT; default f0,SPT.
Ch_power [dBm]: the output power of the lasers. Allowed values: min -;
max -; default: -10.
Channel_spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
PRBS [degree]: the degree of the PRBS generating polynomial for the data
sources. Allowed values: min 5; max 28; default 12.
Transmitter_BW [GHz]: it is the –3 dB cutoff frequency for the electrical
filters. Allowed values: min 0; max BWVBS/4; default Rb.
6.5 References
[1] F. Koyama and K. Iga, Frequency chirping in external modulators, J.
Lightwave Technol., vol. 6, no. 1, pp. 87-33, Feb. 1988.
The Optical fiber link simulates the propagation of the optical signal through a
fiber span. The implemented fiber-model fully takes into account the linear
and non-linear phenomena influencing propagation, as well as polarization
related effects. In particular, PMD and birefringence are simulated
considering their interaction with the non-linearities. PMD and birefringence
are simulated by taking into account of the statistical variation of the
principal axes of the fiber. Non-linear Kerr effect is simulated considering its
instantaneous (SPM, XPM, FWM, PG) and delayed (Raman cross-talk and
amplification) effects on the Optical signal. The frequency variation of the
dispersion is modeled using a 5th order polynomial expansion, and the
statistical variation of the dispersion may also be specified. Fiber loss may
also be defined as a function of frequency (or wavelength) using a second-
order polynomial. In addition to specifying polynomial coefficients, both fiber
dispersion and loss may be defined through an ASCII file (measured values
versus frequency or wavelength). Points of loss ASCII file may also be
interpolated through a spline. The Raman profile may also be specified
through an ASCII file. Raman amplification is simulated together with the
generated in-line ASE noise. Single- and double-scattered ASE noise
components are simulated. The Raman pumps can be specified as either co-
or counter-propagating with respect to the signal propagation. The level of
simulation accuracy obtained can be adjusted by turning on and off
individually many of the above mentioned effects (see “Advanced Control
Parameters” section).
Fibers
SPT Behavior
During SPT simulation fiber links model the propagation of the optical
spectrum taking into account attenuation only. If Raman amplification and
Raman Cross-Talk are turned on, both the spectral-dependent gain and
cross-talk effects are taken into account as well as introduction of the ASE. At
the end of simulation Raman gain and noise spectra plots are available (if
they are enabled).
Note
Choosing the popup menu command View Chart on the fiber link appears
the Raman gain plot only. To display the noise spectra plot you have to start
the OptSim Data Display clicking the button in the OptSim tool bar and
selecting it from the chart tree hierarchy (Chart tab).
Inputs Outputs
Optical Optical
Note
During the analysis of the parameters of a password protected custom fiber,
you may not access the values of the parameters marked as hidden. The
parameters marked as visible are accessible as read-only parameters, but
may not be edited. Whenever a parameter is marked as hidden, its value is
replaced by asterisks.
which are missing in data sheets specification, the default OptSim values
have been used.
Fiber Type Attenuation Chromatic Dispersion MFD Polarization
Note
For some commercial fibers more than one OptSim type can be present in the
list. Each fiber type is defined to represent the same behavior in a different
frequency range.
Moreover, from those data sheets where the measured attenuation curve
versus wavelength is reported, proper values of the attenuation coefficients
have been computed respect to a particular reference wavelength.
As far as Raman profile is concerned some of the OptSim fiber types use
measured data provided by file (see section 7.1.3).
For the Lucent fiber it is possible require a specific maximum individual fiber
loss value during the purchase. Then the attenuation value reported in
OptSim is only a reference and you can change it in order to simulate his own
fiber type.
You can also define your own fiber types by modifying these parameters by
clicking on the Edit/View button. This action opens a secondary dialog
window composed of five sub-sections that allow definition of the parameters
that characterize fiber phenomena. Remember that you cannot modify
directly the predefined realistic fiber, i.e., to create a custom fiber, you have
to select one of the pre-defined realistic fiber-types, and modify it by clicking
on the edit/view button. When you modifies a parameter, OptSim always
creates a copy of the fiber (a custom fiber) and applies your changes. See the
OptSim User Manual for more information on custom fibers.
The Remove button removes a custom fiber from the list. The button is
disabled when you select a predefined fiber (ideal or realistic types).
Length [km]: This parameter specifies the length of the fiber span. Allowed
values: min 0; max -; default 0.
Note
The on-off gain is defined as the ratio of the power level of the signal at the
end of the fiber span with the pump turned on and the power level with the
pump turned off. The noise-figure definition implemented in OptSim is the
one proposed by Hansen in [5]:
1 2 G ASE ( f )
NF ( f ) = 10 log10 1 +
Gon − off ( f ) hf
where Gon-off(f) is the on-off gain at the frequency f, and GASE(f) is the power
spectral density of the ASE noise at the frequency f, at the output of the fiber
span.
Signal and Pumps Powers evolution along fiber Plots. This switch turns
on/off chart generation related to Signals and Pumps Powers evolution along
fiber with Raman amplification, during VBS simulations only. The charts
represent forward signal power evolution, co-propagating and counter-
propagating pumps powers evolution along fiber.
Import from RAMAN Amplifier Design Toolkit, imports the file with
extension ora, coming from the Multi-pump RA Design Toolkit. In this way all
the Raman multi-pump parameters are automatically fitted to obtain the
required Raman flat gain in the SPT bandwidth set in the toolkit.
Pumps Power Filename: name of the data file containing the pump powers
vs. frequency (or wavelength). You may choose the file you wish to use by
clicking on the Browse button. You may also edit the file by clicking on the
Edit button: this opens an editor window and loads the data file. The editor is
VI for Unix or Notepad for Windows. The file syntax must follow the format
specified in section 7.1.3. The frequency must be expressed in [THz], the
wavelength in [nm] and the power in [dBm] or [mW]. The filename extension
must be DAT in capital letters.
No Raman Noise: This choice automatically disable ASE noise generation
due to Raman Amplification
Operating Temperature[K]: It is the absolute temperature of the fiber.
Allowed values: min 0; max -; default 300.
Single This section specifies the single pump condition for Raman amplification. Two
Pumping options are available. Co-propagating assumes the pump is a CW signal
propagating in the same direction as the signal. Counter-propagating
assumes the pump is a CW signal propagating in the opposite direction as the
signal. Note that the pump propagates only through the fiber span where it is
defined, and does not influence other attached fiber spans.
Pump Power [dBm] or [mW]: the pump power injected into the fiber at
the beginning (for co-propagating pumps) or at the end (for counter-
propagating pumps) of the fiber span. Allowed values: min 0; max -; default
100
Pump Frequency/Wavelength [THz] or [nm]: the frequency/wavelength
of the CW Raman amplification pump. Allowed values: min 0; max -; default
1000 (nm).
Pump Attenuation [dB/Km]: the fiber loss at the pump frequency, it
specifies the pump attenuation as it propagates through the fiber span.
Allowed values: min 0; max -; default 1.2.
Multi Pumping This section specifies the multi-pumping condition for Raman amplification.
Pump powers and attenuation must be specified from corresponding files.
Pumps Power Filename: name of the data file containing the pump powers
vs. frequency (or wavelength). You may choose the file you wish to use by
clicking on the Browse button. You may also edit the file by clicking on the
Edit button: this opens an editor window and loads the data file. The editor is
VI for Unix or Notepad for Windows. The file syntax must follow the format
specified in section 7.1.3. The frequency must be expressed in [THz], the
wavelength in [nm] and the power in [dBm] or [mW]. The filename extension
must be DAT in capital letters.
Pumps Attenuation Filename: name of the data file containing the pumps
loss profile vs. frequency (or wavelength). You may choose the file you wish
to use by clicking on the Browse button. You may also edit the file by
clicking on the Edit button: this opens an editor window and loads the data
file. The editor is VI for Unix or Notepad for Windows. The file syntax must
follow the loss format specified in section 7.1.3. The frequency must be
expressed in [THz], the wavelength in [nm] and the loss in [dB/km]. The
filename extension must be DAT in capital letters.
Files format: specifies pump powers measure unit (Power in dBm or
Power in mW), and whether the powers and loss are expressed either as a
function of Frequency or Wavelength.
Where f0,loss is the reference frequency. You may specify the coefficient values
or supply a description file. In the latter case the file must contain the profile
of loss as a function of frequency or wavelength. This feature may be used,
for example, to introduce data from measured results.
OptSim uses the least square method to interpolate the points stored in the
loss ASCII file and to extract the coefficient values to use in (1).
Note
These files are stored in the ProductInstDir\lib\OpLib\Fiber\ directory.
Parameters
From file: specifies that the loss is described through an ASCII data file.
Filename: name of the data file containing the loss profile vs. frequency (or
wavelength). You may choose the file you wish to use by clicking on the
Browse button. You may also edit the file by clicking on the Edit button: this
opens an editor window and loads the data file. The editor is VI for Unix or
Notepad for Windows NT. The file syntax must follow the format specified in
section 7.1.3. The frequency must be expressed in [THz], the wavelength in
[nm] and the loss in [dB/km]. The filename extension must be DAT in capital
letters. OptSim actually derives the above three alpha coefficients through a
least square algorithm and uses the coefficients in the simulation.
Note
The data file must be saved in the same directory where the fiber model is
saved. For a custom fiber saved in the current design, the proper directory is
the one where the design (opf or opm file) is saved. You may avoid this
restriction by using the procedure described for password protection of
custom fibers at the end of section 7.1.3
Parameters
From file: specifies that the dispersion is described through a data file.
Filename: name of the data file that describes the dispersion β2 vs.
frequency or D vs. wavelength. You may choose the file you wish to use by
clicking on the Browse button. You may also edit the file by clicking on the
Edit button: this opens an editor window and loads the data file. The editor is
VI for Unix or Notepad for Windows NT. The file syntax must follow the
format specified in section 7.1.3. The frequency must be expressed in [THz],
the wavelength in [nm], the dispersion coefficients β2 in [ps2/km] and D in
[ps/nm/km]. The filename extension must be DAT in capital letters.
Note
The data file must be saved in the same directory where the fiber model is
saved. For a custom fiber saved in the current design, the proper directory is
the one where the design (opf or opm file) is saved. You may avoid this
restriction by using the procedure described for password protection of
custom fibers at the end of section 7.1.3
File format: specifies that the loss is expressed either as β2 vs. Frequency
or as D vs. Wavelength.
Reference Frequency/Wavelength [THz] or [nm]: the reference
frequency for the following dispersion parameters. Allowed values: min 0;
max -.
β2 [ps2/km] or D [ps/nm/km]: the dispersion value of the fiber at the
reference frequency/wavelength. β2 is related to D: D = − 2πc λ2 β 2 , where λr is
r
the reference wavelength and c is the speed of light. Allowed values: min -;
max -.
β3 [ps3/km] or D' [ps/nm2/km]: the third order fiber dispersion, which is
the variation of the dispersion with respect to the reference
frequency/wavelength. The value may be specified using the third order
derivative of the propagation constant β(ω) with respect to the angular
frequency β3 = ∂3β/∂ω3, or the derivative of the D coefficient with respect to
the wavelength ∂D/∂λ. Allowed values: min -; max -.
β4 [ps4/km]: the fourth order dispersion, that is the fourth order derivative
of the propagation constant β(ω) with respect to the angular frequency β4
= ∂4β/∂ω4. Allowed values: min -; max -.
β5 [ps5/km]: the fifth order dispersion, that is the fifth order derivative of
the propagation constant β(ω) with respect to the angular frequency β5
= ∂5β/∂ω5. Allowed values: min -; max -.
Update: updates the values of the dispersion coefficients for the dispersion
polynomial. As described above, OptSim derives values for these four
coefficients through a least square algorithm.
Note
β2, β3, β4, and β5 are the dispersion coefficients of the series expansion of the
propagation constant β(ω) versus the angular frequency (ω is the angular
frequency, f is the frequency and they are related through ω = 2πf) around
the reference angular frequency ωr, given by the following expression:
β (ω ) = β 0 + β1 (ω − ω r ) +
β2
(ω − ω r )2 + β 3 (ω − ω r )3 +
2 6
β4 β5
+ (ω − ω r )4 + (ω − ω r )5 + 0(ω − ω r )6
24 120
Therefore, the second order dispersion at ω is given by the following
expression
β 2 (ω ) = β 2 + β 3 (ω − ω r ) +
β4
(ω − ω r )2 + β 5 (ω − ω r )3
2 6
where ωr=2πfr, and fr is the reference frequency. The parameters of the
dispersion expression are not constrained. Any value is allowed. It is
important to remember that a higher dispersion value implies longer
simulation times.
OptSimFRaman 1
#
# Fiber Raman gain file
# General comment
#
# - Abscissas in relative frequency [THz]
##
10 0.769
10.5 0.822
11 0.873
11.5 0.921
12 0.959
12.5 0.981
12.75 0.985
13 0.981
13.25 0.975
13.5 0.963
14 0.91
14.5 0.973
14.75 0.882
Example 4
OptSimFRaman 1
#
# Fiber Raman profile (CR ) file
# General comment
#
# - Raman gain profile in [1/Km/W]
# - Abscissas in relative frequency [THz] respect to the reference freq
##
10 0.014
10.5 0.015
11 0.016
11.5 0.017
12 0.0174
12.5 0.0178
12.75 0.0179
13 0.0178
13.25 0.0177
13.5 0.0175
14 0.0165
14.5 0.0177
14.75 0.016
Example 5
#
# Fiber Raman scaling factor (kscale) file
# General comment
#
# - Abscissas in absolute frequency [THz]
##
214 0.947
214.3 0.964
227.3 1
250 1.01
To create a password-protected custom fiber that uses loss, dispersion or
Raman profiles described by an ASCII file, please use the following
procedure.
Note
It is not possible to protect Raman gain user-defined files, which are visible to
all users, but you can protect all other Raman parameters.
Note
All the pump frequencies must be set outside the SPT bandwidth.
Example 1
OptSimFPumps 1
# Abscissas in frequency [THz]
# Pumps power in [mW]
##
Coprop 1
210 500
Counterprop 3
206 300
204 150
212 620
Parameters
Dispersion Statistical Distribution: choice of the statistical variation of the
dispersion along the fiber length. One of two distributions may be chosen for
the dispersion variation: Uniform and the Gaussian, up to 3σ distribution.
Parameters
Reference Frequency/Wavelength for Non-Linearity [THz] or [nm]:
the reference frequency fr for the following fiber non-linearity parameters.
Allowed values: min 0; max -.
Fiber Non-Linearity Coefficient [1/W/km]: the fiber non-linearity
coefficient as defined in [1]. Its value depends on the non-linear refractive
index n2 of the medium and on the effective area Aeff of the fiber:
2πn2 f r
γ= (1)
cAeff
where fr is the reference frequency and c is the speed of light. Allowed
values: min 0; max -.
Note
To take into account the non-linearity effect reduction in single polarization
condition due to the birefringence interaction it is necessary to multiple the
coefficient value by 8/9.
∫ ∫ F (x, y ) dxdy
4
− ∞− ∞
Together with the non-linear refractive index n2, it represents another way to
express fiber non-linearity. Allowed values: min >0; max -.
Fiber Average Beat Length [m]: The beat length LB, as defined in [1], is
the distance after which the polarization of an optical signal propagating in a
fiber span recovers the initial state:
2 2π
LB = = (3)
∆β 0 β x − β y
2π
σβ = σL
0
L2B B
Fiber PMD [ps/km1/2]: the Polarization Mode Dispersion (PMD) effect
coefficient δPMD. This coefficient is the parameter that is usually specified in
data-sheets to quantify the amount of PMD introduced by the fiber.. It is used
to implement the Differential Group Delay (DGD) statistical model as
presented in [5],[12]. Allowed values: min 0; max -.
Note
The PMD effect is simulated through the wave-plate model [12] where the
relationship between the wave-plate length LR , the beat-length LB, and the
correlation length Lc is given by the following expression [13]:
LR
LC = −
L
ln cos π R
LB
The resulting DGD at the and of the fiber span is a random variable with a
Maxwellian distribution and the Probability Density Function takes the
following form [12]:
2 ∆τ 2 ∆τ 2
P(∆τ , z ) = exp − 2 0 < ∆τ < +∞
π q3 2q
π
where q2 = δ PMD
2
Lspan . The average DGD is ∆τ = δ PMD Lspan , and its
8
3π − 8
variance is σ= δ PMD Lspan = 0.422 ∆τ . The effect of PMD on the
8
optical signal may strongly vary, and may also change the correlation length
and beat length parameter.
∂A( z, t )
∂z
{ t
= − jγ (1 − f R )A(z, t ) A(z, t ) + f R A( z, t )∫ hR (t − τ ) A(z,τ ) dτ
2
−∞
2
}
where the impulse response hR(t) assumes the following form [7], obtained
from the inverse Fourier transform of the Raman gain spectral-shape:
τ 12 + τ 22 τ τ
hR (τ ) = exp − sin
τ 1τ 2
2
τ 2 τ1
When a dual-polarization model is considered, the different effect of the
delayed non-linear response on the parallel and orthogonal polarization
components of the optical field [8] is not taken into account. From a physical
point of view it is assumed that the third-order susceptibility constants
χ(3)xxxx, χ(3)xyyx, χ(3)xyxy, χ(3)xxyy are all characterized by the same time
constants of the impulse response.
Channels Pumps
Noise
∂Ps+ (z, f )
= −α( f )Ps+ (z, f ) + rα( f )Ps− (z, f ) +
∂z
f p,max
1 υ
+ ∫ k A f
[ ]
gR (υ − f ) Pp+ (z,υ) + Pp− (z,υ) dυ Ps+ (z, f ) +
f p,min pol eff ref
f p,max
1 υ
+ hf ∫ k A f
1
h(υ− f )
−
[ ]
gR (υ − f ) Pp+ (z,υ) + Pp− (z,υ) dυ
f p,min pol eff ref
1− e KT
∂Ps− (z, f )
= +α( f )Ps− (z, f ) − rα( f )Ps+ (z, f ) +
∂z
f p,max
1 υ
− ∫ k A f
[ ]
gR (υ − f ) Pp+ (z,υ) + Pp− (z,υ) dυ Ps− (z, f ) +
f p,min pol eff ref
f p,max
1 υ
− hf ∫ k A f
1
h(υ− f )
−
[ ]
gR (υ − f ) Pp+ (z,υ) + Pp− (z,υ) dυ
f p,min pol eff ref
1− e KT
where Ps+(z,f)
and Ps-(z,f)
are the forward and backward propagating power
at the frequency f and at the distance z, α(f) is the fiber attenuation at the
frequency f, gR(∆f) is the Raman gain [m/W] at the frequency fref considering
the pump to be at the frequency ν, r is the Rayleigh back-scattering capture
factor [5], kpol is a constant that takes into account the polarization op the
pumps with respect to the signal [14], if the pump is completely depolarized
kpol =2 and this is the hypothesis used in OptSim, Aeff is the effective area of
the fiber, h is Planck's constant, K is the Boltzmann’s constant and T is the
operating temperature.
In the pump bandwidth, we considered all the phenomena taken into account
in the signal bandwidth plus the depletion terms, therefore the equations to
be considered are:
∂Pp+ ( z , f )
= −α ( f )Pp+ ( z , f ) + rα ( f )Pp− ( z , f ) +
∂z
f p , max
υ
+ ∫
1
k pol Aeff f ref
[ ]
g R (υ − f ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ Pp+ ( z , f ) +
f
f p , max
υ
+ hf ∫
1
k pol Aeff f ref
[
g R (υ − f ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ + ]
f
f p , min
− ∫
1 f f
υ k pol Aeff f ref
[ ]
g R ( f − υ ) Ps+ ( z ,υ ) + Ps− ( z ,υ ) dυ Pp+ ( z , f ) +
f s , min
− ∫
1 f f
υ k pol Aeff f ref
[ ]
g R ( f − υ ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ Pp+ ( z , f ) +
f p , min
f p , min
f1 f hυ
−2 ∫ gR ( f −υ ) h ( f −υ )
dυ Pp+ ( z , f ) +
υ k pol Aeff f ref −
f s , min
1− e KT
f
f1 f hυ
−2 ∫ gR ( f −υ) h ( f −υ )
dυ Pp+ ( z , f )
υ k A
pol eff f ref
−
f p , min
1− e KT
∂Pp− ( z , f )
= +α ( f )Pp− ( z , f ) − rα ( f )Pp+ ( z , f ) +
∂z
f p , max
υ
− ∫
1
k pol Aeff f ref
[ ]
g R (υ − f ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ Pp− ( z , f ) +
f
f p , max
υ
− hf ∫
1
k pol Aeff f ref
[
g R (υ − f ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ + ]
f
f p ,min
+ ∫
1 f f
υ k pol Aeff f ref
[ ]
g R ( f − υ ) Ps+ ( z ,υ ) + Ps− ( z ,υ ) dυ Pp+ ( z , f ) +
f s , min
+ ∫ υ k
1
A
f
f
f
[
g R ( f − υ ) Pp+ ( z ,υ ) + Pp− ( z ,υ ) dυ Pp+ ( z , f ) +]
f p ,min pol eff ref
f p , min
f1 f hυ
+2 ∫ gR ( f −υ ) h ( f −υ )
dυ Pp− ( z , f ) +
υ k pol Aeff f ref −
f s , min
1− e KT
f
f1 f hυ
+2 ∫ gR ( f −υ ) h ( f −υ )
dυ Pp− ( z , f )
υ k pol Aeff f ref −
f p , min
1− e KT
where Pp+(z,f) and Pp-(z,f) are the forward and backward propagating pump
power at the frequency f and at the distance z.
If a Lorentzian profile has been chosen for the Raman profile, gR(∆f) is given
by the following expression:
∆ω = 2π ( f p − f )
and
τ 12 + τ 22 1
H R (∆ω ) = F {hR (t )} = 2 2 2 2
τ1 τ 2 1 1
+ j∆ω +
τ2 τ1
otherwise, user-defined or Standard Raman profile [9] are used.
Considering the power spectral density of the optical field at the input of the
fiber and the power spectral density of the pumps – the counter-propagating
pumps launched at the end of the fiber span and the co-propagating pumps
inserted at the beginning of the fiber span – the two sets of differential
equations are solved together using the shooting numerical method [15].
At the end of this preliminary computing phase, the gain profile with respect
to the frequency is available for each section z of the fiber together with the
power spectral density of the ASE noise due to the Raman spontaneous
emission and infinite Rayleigh reflections.
Following this procedure, phenomena taken into account are:
pump-to-signal bi-directional Raman gain
pump-to-pump bi-directional Raman gain (Raman cross-talk in the pump
bandwidth)
pump depletion
bi-directional Amplified Spontaneous Emission Noise in the signal
bandwidth
bi-directional Amplified Spontaneous Emission Noise in the pump
bandwidth
infinite additive noise components due to the interaction of Raman
Spontaneous emission with Rayleigh back-scattering
Note that the signal-to-signal Raman gain (Raman cross-talk in the signal
bandwidth) is not considered in this pre-computing phase because it is fully
taken into account in the time domain simulation.
The effect of Raman gain (a frequency-dependent gain-shape) on the signal
is integrated in the non-linear operator of the split-step method, explicitly
taking into account the effect of increasing signal power during propagation.
In distinction, the ASE noise is added only at the end of the fiber span and is
given by the shape of the power spectrum. No interaction between ASE noise
and the signal is taken into account during the propagation in the amplifying
fiber span.
F ib e r S p a n
S p lit- Ram an
L 1/2 N L L 1/2 S te p Pum p
M e th o d
S ASE (f)
SSM RA SSM RA SSM RA
Raman gain It is possible to set a different gain profile for each Raman pump. Raman
coefficient Gain of a pumped optical fiber (Np pumps) can be simply expressed by using
dependency the following formula:
on pump
frequency 1
Np L
Gon −off ( f ) = 10 log10 (e)∑ C R ( f p ,i − f , f p ,i )∫ Pp ,i ( z ) d z
i =1 k pol
dB
0
Where:
g R (∆f , f p ) 1
C R (∆f , f p ) =
Aeff ( f p ) W⋅ km .
From the theory [17], it is known that:
fp
g R (∆f , f p ) = g R (∆f , f ref ) .
f ref
Hence,
g R (∆f , f ref ) fp
C R (∆f , f p ) = ⋅ =
Aeff ( f p ) f ref
g R (∆f , f ref ) Aeff ( f ref ) f p
⋅ ⋅ =
Aeff ( f ref ) Aeff ( f p ) f ref
1
C R (∆f , f ref ) ⋅ k scale ( f p )
W⋅ km
So this means that the Raman efficiency changes with frequency with a
scaling factor kscale(fp), which takes into account the linear frequency scaling
contribution fp/fref and also the frequency dependence of the effective area Aeff
(fref)/Aeff (fp).
To model the dependency of CR(∆f,fp) on fp, you can select the User Defined
option for the fiber Raman gain profile in the Raman Effects section of the
fiber link.
The CR profile can be then specified through a two-column file containing the
value of CR versus ∆f measured using a pump at the fref frequency:
......
∆fi-1 CR(∆fi-1, fref)
∆fi CR(∆fi, fref)
∆fi+1 CR(∆fi+1, fref)
......
The file syntax must follow the format specified in section 7.1.3.
To specify the kscale dependency on the frequency you can select the Custom
Scaling of Raman Profile button and insert a two-column file containing the
values of kscale versus fref
......
fref, j-1 kscale (fref, j-1)
......
The file syntax must follow the format specified in section 7.1.3.
Figure 7.2 Standard Curve for the Raman Profile for silica fibers [9].
Parameters
Raman Profile SRS is characterized by a profile of the gain vs the frequency difference
between the signal [9] and the pump. This profile depends on the fiber
material. Three profiles are available in the model. Standard is the silica
fiber profile measured at reference wavelength of 1000 nm (see [1] p. 318).
Lorentzian is the profile approximated using the fitto a Lorentzian curve [9],
characterized by three parameters. User Defined is a profile described
through a data file.
The OptSim library includes some files that provide measured Raman profiles
reported in [16].
Reference Frequency/Wavelength [THz] or [nm]: it is the reference
frequency fR for the Raman gain profile. Allowed values: min 0; max -;
default 1000 (nm).
In order to use correctly these data it is necessary to set properly the
Reference Wavelength/Frequency parameter, reported in the file header.
The following table reports the file name and the measured conditions.
File name Reference Wavelength (nm)
Raman_LucentTRUEWAVE.DAT 1455
Raman_LucentALLWAVE.DAT 1455
Raman_CorningLEAF.DAT 1455
Raman_CorningNZDSF.DAT 1455
Raman_CorningDSF.DAT 1455
Raman_NDSF.DAT 1455
Raman_STD.DAT 1000
Note
These files are stored in the ProductInstDir\lib\OpLib\Fiber\ directory.
Raman Profile Filename: name of the data file for the Raman profile vs.
frequency or vs. wavelength. You may choose this file from the file system by
clicking on the Browse button. You may also edit the file by clicking on the
Edit button. This opens an editor window and loads the data file. The editor is
VI for Unix or Notepad for Windows NT. The file syntax must follow the
format specified in section 7.1.3. The frequency must be expressed in [THz],
the wavelength in [nm] and the gain in [m/W·10-13] for gR or [1/Km/W] for
CR. Note that the abscissa represents the difference between the pump
frequency fpump and the signal frequency f. The filename extension must be
DAT in capital letters.
Note
The file must be saved in the same directory where the fiber is saved. For a
custom fiber saved in the current design the directory is where the design
(opf or opm file) is saved. Moreover you can use the procedure described for
the password protected custom fiber at the end of section 7.1.3 to avoid this
restriction.
gr and Cr, toggle options to select the kind of variable specified in the Raman
profile file gR(f) or CR(f).
Custom scaling of Raman profile, specifies the frequency dependence of
the fiber effective area. In particular, for each pump the used Raman
coefficient is evaluated by the following formula:
1
C R (∆f , f p ,i ) = C R (∆f , f ref ) ⋅ k scale ( f p ,i )
W⋅ km
Where kscale is obtained from a data file specified through the following field.
When this option is not selected the scaling factor takes into account only of
the linear scaling.
Custom Scaling Filename: name of the data file for the scaling factor
profile vs. frequency or vs. wavelength. You may choose this file from the file
system by clicking on the Browse button. You may also edit the file by
clicking on the Edit button. This opens an editor window and loads the data
file. The editor is VI for Unix or Notepad for Windows NT. The file syntax must
follow the format specified in section 7.1.3. The frequency must be expressed
in [THz], the wavelength in [nm]. The filename extension must be DAT in
capital letters.
Files format: specifies if the Raman profile and the scaling factor profile are
expressed as a function of Frequency or Wavelength.
Raman Fiber Raman constant [-]: the characteristic constant fR [1] that controls
constants the amount of delayed non-linearity (Raman effect) with respect to the
instantaneous one. Given the non-linear coefficient γ, (1-fR)γ is the coefficient
of the instantaneous non-linear response of the medium, while fRγ is the
coefficient of the delayed (Raman effect) non-linear response of the medium.
If a Standard or User Defined Raman profile is chosen, fR is derived from a
Lorentzian fitting to this profile, and is displayed as a read-only parameter.
Allowed values: min 0; max 1; default: 0.18.
First Raman Time Constant [fs]: the first Raman time constant (τ1) that
controls the sinusoidal part of the impulse response of the Kerr effect. If a
Standard or User Defined Raman profile is chosen, τ1 is derived from a
Lorentzian fitting to this profile, and is displayed as a read-only parameter.
Allowed values: min 0; max -; default: 12.2.
Second Raman Time Constant [fs]: the second Raman time constant (τ2)
that controls the exponential part of the impulse response of the Kerr effect.
If a Standard or User Defined Raman profile is chosen, τ2 is derived from a
Lorentzian fitting to this profile, and is displayed as a read-only parameter.
Allowed values: min 0; max -; default: 32.
Update: updates the above Raman constants By computing the values from
the Lorentzian fitting to the data contained in the gain profile file. It is active
only if gain is introduced from a file (Standard or User Defined profile).
Rayleigh Scattering: enables Rayleigh Scattering in order to evaluate the
amount of ASE noise introduced by the Raman amplification [4]. The Rayleigh
scattering is applied to noise only, hence its effect on the propagating signal
is not taken into account. Two options are available: Ideal and Realistic.
Capture Factor [dB]: the fraction of scattered photons that are back
scattered. It may be specified only if Realistic Rayleigh scattering is chosen.
Allowed values: min -; max 0; default –30.
Reflectivity: enables fiber reflectivity at the input and output of the fiber.
The Ideal case indicates perfect transmission through fiber input and output
interfaces, the Realistic case takes into account input and output fiber
reflectivity.
Input Reflectivity [dB]: the fraction of reflected photons at the input of the
fiber. It may be specified only if Realistic reflectivity is chosen. Allowed
values: min -; max 0; default –40.
Pump Equation
∂I p f p
= −α ⋅ I p − ∫ C B (ξ )I s ( f p − ξ )dξ ⋅ I p (1)
∂z 0
∂I s
= +α ⋅ I s − C B (∆f ) ⋅ I s I p − h ⋅ f s ⋅ C B (∆f ) ⋅ I p (2)
∂z
where
I p ( f p , z) = I p ( z) ⋅ δ ( f − f p ) (3)
is the SBS pump (i.e., the signal propagating in the fiber in the forward
direction)
1 g B (∆f ) 1
C B (∆f ) = ⋅ m ⋅ W (4)
k pol Aeff
where
2
∆f B
2 m
g B (∆f ) = g B ,max ⋅ 2 W (5)
∆f B
+ (∆f − f B )
2
2
and
∆f B m
g B ,max = g B ,0 ⋅ W (6)
∆f B + ∆f p
where
2 ⋅ π ⋅ n 7 ⋅ p122 ⋅ f p2 m
g B,0 = W (7 )
c 3 ⋅ ρ 0 ⋅ V A ⋅ ∆f B
where
• p12 is the longitudinal elasto-optic coefficient (typical value: p12 = 0.286);
• ρ0 is the material density (typical value: ρ0 = 2200 kg/m3);
• ∆fB is the Brillouin linewidth (typical value: ∆fB = 10 MHz);
and ∆fB is the Brillouin gain linewidth and fB of peak of the Brillouin
gain is:
2 ⋅ n ⋅VA ⋅ f P
fB = [Hz] (8)
c
where
• n is the refractive index of the core of the fiber (typical value: n = 1.45);
• VA is the speed of the acoustic wave in the fiber (typical value: VA = 5960
m/s) ;
• fP is the pump frequency;
• c is the velocity of the light in the vacuum.
A typical value for the peak of the SBS gain at λp= c/fp= 1550 nm is: fB ≈ 11
GHz.
• Ip(0) is the amount of power injected in the fiber minus the portion
“reflected” by the SBS, and Is(0) is the amount of power “reflected” by the
SBS. Therefore, the boundary conditions for the problem are:
I p (0) = I p 0
(9)
I s (L ) = 0
where Ip0 is the signal at the input of the fiber
Measurement Units
• Ip, Is: [W/Hz]
• Frequencies: [Hz]
• h = 6.626068 × 10-34 [J·s]
• α : [1/m]
1 gB 1
• CB = ⋅ m ⋅ W
k pol Aeff
A typical value for gB is 4 × 10-10 [m/W]
• kpol in [1;2] : [pure number]. kpol accounts for the random relative
polarization between pump and Stokes wave, in conventional fibers kpol =
1.5.
• Aeff (fiber effective area): [m2]
g B , max m
C B , max = W ,
k pol Aeff
Beq , SBS Beq , SBS
( )
C B ∆f = if ∆f ∈ f B − ; fB +
(10)
2 2
m
0 W , elsewhere
where
2 +∞
∆f 1
Beq ,SBS = B ⋅ ∫ 2
dξ =
2 0 ∆f B
+ (ξ − f B )
2
2
∆f B π f
= ⋅ + arctan 2 B [Hz] (11)
2 2 ∆f B
f p − fB +
Beq , SBS
∂I p 2
= −α ⋅ I p − C B ,max ∫ I s (ξ )dξ ⋅Ip (12)
∂z Beq , SBS
f p − fB −
2
+ α ⋅ I s − C B ,max ⋅ I s I p − h ⋅ f s ⋅ C B ,max ⋅ I p ,
∂I s
= if f s ∈ f p − f B − eq ,SBS ; f p − f B + eq ,SBS
B B
∂z 2 2
0, elsewhere
(13)
We can assume that
I s 0 (0),
B B
I s ( f s , z ) = if f s ∈ f p − f B − eq , SBS ; f p − f B + eq , SBS
2 2
0, elsewhere
(14)
Therefore, the couple of differential equations to be solved is
∂I p
= −α ⋅ I p − C B ,max I s 0 Beq ,SBS ⋅ I p
∂z ∂I s 0
= +α ⋅ I s 0 − C B ,max ⋅ I s 0 I p
∂z
− h ⋅ f s ⋅ C B ,max ⋅ I p
(15)
Cosidering the overall amount of Brillouin power
∂I p
∂z = −α ⋅ I p − CB , max I s ⋅ I p
∂I
s = +α ⋅ I s − C B , max ⋅ I s I p − h ⋅ f s ⋅ Beq , SBS ⋅ CB , max ⋅ I p
∂z
(17)
and the overall amount of power reflected by the SBS is
I p (0) = I p 0
(19)
I s (L ) = 0
In order to solve the differential equations system (17) it could be convenient
to operate the following variable exchange
sL C B ,max
Yielding
∂I pL
= − exp{C B , max ⋅ I sL + α ⋅ z}
∂ z
∂I
sL
[
= − I p 0 ⋅ exp C B , max ⋅ I pL − α ⋅ z ⋅]
∂z
{1 + h ⋅ f ⋅ B
s eq , SBS ⋅ exp[− C B , max ⋅ I sL − α ⋅ z ]}
(21)
The boundary conditions for the differential equations system (21) are
I pL (0) = 0
(22)
I sL (L ) = −∞
The CW Signal Assumption
The steady-state model for the SBS assumes each channel is a CW signal and
does not include the effect of simultaneous presence of more than one
channel at different frequencies. Therefore, studying the effects of SBS using
the steady state model, the users obtain only the effects of power reflection
due to SBS. All the dynamic effects are neglected, i.e., all the effects
influencing the shape of the pulses and all the possible interactions between
the SBS and the other propagation phenomena such as dispersion and
nonlinearities. The interaction between SBS and SRS is also neglected. This
may yield underestimate of SBS effect in case of using the unusual Raman
amplification with forward pumping.
Parameters
Brillouin Gain [m/W]: peak value of Brillouin gain. Allowed values: min 0;
max 1; default 3e-11.
Brillouin Gain Linewidth [MHz]: Brillouin gain linewidth. Allowed values:
min 0; max 1000; default 40.
Brillouin Gain Reference Frequency [THz]: reference frequency for
Brillouin gain. Allowed values: min 0; max 1000; default 193.
Frequency Brillouin Gain Peak [GHz]: frequency of the peak of the
Brillouin gain. Allowed values: min 0; max 1000; default 11.
Signal Linewidth [MHz]: linewidth of the signal. Allowed values: min 0;
max 1e6; default 100.
Signal and Stokes Wave Relative Polarization: relative polarization of
signal and Stokes waves. Allowed values: min 1; max 2; default 1.5.
SBS Spectrum Plot: switch to enable or disable plotting of the SBS
spectrum. Two options are available: On and Off. Default is Off.
7.2 References
[1] G. P. Agrawal, Nonlinear Fiber Optics, Academic Press Inc., Boston, 1995.
[2] OptSim Technical Staff, “TDSS: Time Domain Split-Step Fiber Simulation
Approach,” White paper of erstwhile Artis Software Corporation, Turin
(now RSoft Design Group, New York), 1999.
[3] A. Carena, V. Curri, R. Gaudino, P. Poggiolini and S. Benedetto, “A time-
domain optical transmission system simulation package accounting for
nonlinear and polarization related effects in fiber,” IEEE Journal on
Selected Areas in Communications, Vol. 15, NO. 4, May 1997.
[4] Y. R. Shen, Principles Of Nonlinear Optics, Wiley-Interscience, New York,
1984.
[5] Hansen et al., “Rayleigh scattering limitations in distributed Raman pre-
amplifiers,” IEEE Photonics Technology Letters, Vol. 10, NO. 1, January
1998.
[6] G. J. Foschini and C. D. Poole, “Statistical theory of polarization dispersion
in single mode fibers,” IEEE/OSA Journal of Lightwave Technology, Vol. 9,
NO. 11, November 1991.
[7] K. J. Blow, D. Wood, “Theoretical description of transient Stimulated
Raman Scattering in optical fibers,” IEEE Journal of Quantum Electronics,
Vol. 25, NO. 12, December 1989
[8] E.A.Golovchenko, N. Pilipetkii, “Unified analysis of four-photon mixing,
modulational instability, and stimulated Raman scattering under various
polarization conditions in fibers,” OSA Journal of Optical Society of
America B, Vol. 11, NO. 1, January 1994.
[9] Stolen et al., “Raman response of silica-core fibers,” OSA Journal of
Optical Society of America B, Vol. 6, No. 6, June 1989.
[10] E. Desurvire et al., “Theory and implementation of a Raman active fiber
delay line,” IEEE/OSA Journal of Lightwave Technology, vol. LT-4, No. 4,
April 1986
[11] S. C. Rashleigh, “Origins and control of polarization effect in single-mode
fibers,” IEEE/OSA Journal of Lightwave Technology., Vol. LT-1, pp. 312–
331, 1983.
[12] F. Curti et al., “Statistical treatment of the evolution of the principal
states of polarization in single-mode fibers,” IEEE/OSA Journal of
Lightwave Technology., Vol. 8, N0. 8, pp. 1162–1165, August 1990
[13] F. Corsi et al., “Polarization Mode Dispersion Characterization of Single-
Mode Optical Fiber Using Backscattering Technique,” IEEE/OSA Journal
of Lightwave Technology., vol. 16, N0. 10, pp. 1832–1843, October
1998.
[14] Hansen et al.: et al., “ Capacity Upgrades Of Transmission Systems By
Raman Amplification,” IEEE Photonics Technology Letters, Vol. 9, NO. 2,
February 1997
[15] "Numerical Recipes" Cambridge University Press, (see also www.nr.com
site)
[16] C. fludger, A. Maroney, N. Jolley, R. Mears, An analysis of the
improvements in OSNR from distribuited Raman amplifiers using modern
transmission fibres, OFC 2000, Paper FF2
[17] Miller, Optical Fiber Telecommunications, Academic Press, 1979
πf
sin
H(f ) = B0 (1)
πf
B0
where:
B0 ≅ 2.25 ⋅ BW (2)
You can also choose the level of extinction ratio (corresponding to the ratio
between the maximum and minimum values of the input-output P-V
transmission characteristic) and the chirp factor (see definition in [1]).
The input optical signal is multiplied by factor dependent on:
the input voltages VA and VB applied to the two modulator arms
the excess loss ELdB introduced by the modulator
the extinction ratio ERLIN
the chirp factor α
The output field is:
−
ELdB
1
E out = 10 20
cos φ D − j sin φ D e jφ S e jϑ E in (3)
ERLIN
where:
π Vin − Von
φD = (4)
2 Vπ
π VS
φS = ⋅ (5)
2 Vπ
1
ϑ = α ln Pout (6)
2
dϑ
α = 2 Pout dt (7)
dPout
dt
2
Pout = E out (8)
E out
ERLIN = max
(9)
E out
min
and
V A − VB V A + VB
Vin = , Vs = (10)
2 2
and α is the chirp factor as is defined in [1]. The ERLIN parameter accounts for
the extinction ratio that is defined as below, in the parameter description
section. When the semi-difference between input voltages Vin is equal to Von,
the power of the optical signal is attenuated by the excess loss only, so the
modulator is in the maximum transmission state. To switch to the minimum
transmission state a Vπ voltage must be added or subtracted to Von. A
pictorial explanation of the trans-characteristic is shown in the graph below.
Pout Vπ
Pmax
Pmin
Von Vin
Figure 8.1 The amplitude modulator trans-characteristic
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but to study the systems in terms of average power per channel
and noise level. Therefore, the only actions of amplitude modulators that has
to be taken into account for SPT purposes are the optical spectral
modifications introduced at the channel levels and at the noise level.
In general, an amplitude modulator does not introduce any noise and reduces
the average power of the channel. If the modulation is an exactly squared
NRZ modulation, the average power reduction is well known and it is equal to
3 dB, in addition to the excess loss. Otherwise it has to be evaluated. This
average power reduction is the only parameter used in the SPT simulations
and it is requested in the related section of the parameter dialog window.
If you don’t know which is the value of this parameter, we suggest to perform
a VBS simulation of the subsystem containing the source and the modulator
to measure the optical power before and after the modulator. The average
power reduction is the difference (in dB) between the evaluated powers. Note
that the measured average power reduction contains also the excess loss.
Inputs Outputs
Electrical: the first electrical signal Optical: the modulated optical signal
VA used to modulate the Ein optical Eout at the output of the device
signal
Electrical: the second electrical
signal VB used to modulate the Ein
optical signal
Optical: the optical signal Ein
injected in the device to be
modulated
(2
ERdB = 10 log ERLIN ) (11)
Chirp Factor α [-]: the chirp factor as it is defined in [1]. Note that a non-
ideal extinction ratio (ERdB less than infinite) induces an implicit chirping that
is not taken into account by the chirp factor α. Moreover, also if an
unbalanced Vin voltage, meaning that VA is not equal to -VB, is applied an
extra chirping takes place. Allowed values: min -; max -; default 0.
Note that the pulse generators have been implemented using the chirp factor
C that relates the phase to a function of time. This is the definition used in
[2], and it is different from the chirp parameters α defined for the
modulators. The α parameter refers to the definition of phase that depends
on the power, as it is defined in [1]. Note that the meaning of the two
definitions is completely different, therefore, the impact on system
performance is also different. For a first order Gaussian pulse only α = -C.
Comparing the implemented formulas with the ones presented in [2], you
may note a difference in the sign of C. The reason is that in [2] the
eigenfunction for the analytic signal representation is e-jωT. Therefore, the
instant frequency is
1 ∂Φ (T )
f =− ,
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
Vπ [V]: the voltage swing needed to switch between the minimum and
maximum transmission states. Allowed values: min 0; max -; default 5.
Electrical Filtering with sin(πf/Bw)/(πf) Law: specifies whether or not the
electrical input is filtered by the frequency transfer function (1).
-3dB Bandwidth [GHz]: the -3 dB cutoff frequency of the lowpass filter; it's
available only if the Electrical Filtering with sin(πf/Bw)/(πf) Law is set.
Allowed values: min 0, max 2.25(BWSPT/2), default Rb (reference bit-rate).
First filter notch [GHz]: the resulting first notch in the transfer function.
The transfer function of the filter model (1) is typical for several optical
devices where an optical beam and a traveling wave electrical signal interact
over a given non-negligible length. The filtering function is consequently
determined by the amount of time required by the electrical signal to
propagate along the device. If you need to simulate different filtering
functions than the previous one, simply disable the Electrical Filtering with
sin(πf/Bw)/(πf) Law option, and choose a proper external filter to be
inserted between the electrical signal generator and the electrical input of the
modulator.
1 ∂Φ (T )
f =−
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but to study the systems in terms of average power per channel
and noise level. Therefore, the only actions of amplitude modulators that has
to be taken into account for SPT purposes are the optical spectral
modifications introduced at the channel levels and at the noise level.
In general, an amplitude modulator does not introduce any noise and reduces
the average power of the channel. If the modulation is an exactly squared
NRZ modulation, the average power reduction is well known and it is equal to
3 dB, in addition to the excess loss. Otherwise it has to be evaluated. This
average power reduction is the only parameter used in the SPT simulations
and it is requested in the related section of the parameter dialog window.
If you don’t know which is the value of this parameter, we suggest to perform
a VBS simulation of the subsystem containing the source and the modulator
to measure the optical power before and after the modulator. The average
power reduction is the difference (in dB) between the evaluated powers. Note
that the measured average power reduction contains also the excess loss.
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the two models, Sin2 and
Linear.
Excess loss [dB]: the value of a constant extra loss introduced by the
modulator, which is independent of the electrical input signal. Allowed values:
min 0; max -; default 3.
πf
sin
H(f ) = B0 (1)
πf
B0
where:
B0 ≅ 2.25 ⋅ BW (2)
You can also choose the level of extinction ratio (corresponding to the ratio
between the maximum and minimum values of the input-output transmission
characteristic) and the chirp factor (see definition in [1]).
The input optical signal is multiplied by a factor dependent on:
the input voltage Vin applied to the modulator arm
the excess loss ELdB introduced by the modulator
the extinction ratio ERLIN
the chirp factor α
The output field is:
− dB
EL
r 1 r
E out = 10 20 cos φ D − j sin φ D ⋅ e jϑ Ein (3)
ERLIN
where:
π Vin − Von
φD = (4)
2 Vπ
1
ϑ = α ln Pout (5)
2
dϑ
α = 2 Pout dt (6)
dPout
dt
r 2
Pout = E out (7)
r
Eout
ERLIN = r max
(8)
E out
min
and α is the chirp factor as it is defined in [1]. The ERLIN parameter accounts
for the extinction ratio that is defined as below, in the parameter description
section. When the input voltage is equal to Von, the power of the optical signal
is attenuated by the excess loss only, so the modulator attains the state of
maximum transmission. To switch over to the state of minimum transmission
a Vπ voltage must be applied in addition to the Von one. A pictorial explanation
of the trans-characteristic is shown in Figure 8.1.
Inputs Outputs
Electrical: the electrical signal Vin Optical: the modulated optical signal
used to modulate the Ein optical Eout at the output of the device
signal
Optical: the optical signal injected
in the device to be modulated
Chirp Factor α [-]: the chirp factor as it is defined in [1]. Note that a non-
ideal extinction ratio (ERdB less than infinite) induces an implicit chirping that
is not taken into account by the chirp factor α. Allowed values: min -; max -;
default 0.
Vπ [V]: the voltage swing needed to switch over between the minimum and
maximum transmission states. Allowed values: min 0; max -; default 5.
Electrical Filtering with sin(πf/Bw)/(πf) Law: specifies whether or not the
electrical input is filtered by the frequency transfer function (1).
-3dB Bandwidth [GHz]: the -3 dB cutoff frequency of the lowpass filter; it's
available only if the Electrical Filtering with sin(πf/Bw)/(πf) Law is set.
Allowed values: min 0, max 2.25(BWSPT/2), default Rb (reference bit-rate).
First filter notch [GHz]: the resulting first notch in the transfer function.
The transfer function of the filter model (1) is typical for several optical
devices where an optical beam and a traveling wave electrical signal interact
over a given non-negligible length. The filtering function is consequently
determined by the amount of time required by the electrical signal to
propagate along the device. If you need to simulate different filtering
functions than the previous one, simply disable the Electrical Filtering with
sin(πf/Bw)/(πf) Law option, and choose a proper external filter to be
inserted between the electrical signal generator and the electrical input of the
modulator.
where:
1
ϑ = α ln Pout (2)
2
r 2
Pout = E out (3)
dϑ
α = 2 Pout dt (4)
dPout
dt
0 , Vin < 0
M % 100
f (Vin ) = Vin , 0 < Vin < (5)
100 M%
1 100
, Vin >
M%
Consequently the transmission is turned off for Vin ≤ 0 while it is at maximum
for Vin ≥ (100 / M%). For example, if M% = 50%, the transfer function in
terms of the output power vs. applied voltage is linear in the range between 0
and 2 V, and is hard-limited outside this range, as shown below.
Pout
Pout,max
Inputs Outputs
Electrical: the electrical signal Vin Optical: the modulated optical signal
used to modulate the Ein optical Eout at the output of the device
1 ∂Φ (T )
f =−
2π ∂T
while in OptSim the eigenfunction is e+jωT and the instant frequency is:
1 ∂Φ(T )
f =+ .
2π ∂T
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but to study the systems in terms of average power per channel
and noise level. Therefore, the only actions of amplitude modulators that has
to be taken into account for SPT purposes are the optical spectral
modifications introduced at the channel levels and at the noise level.
In general, an amplitude modulator does not introduce any noise and reduces
the average power of the channel. If the modulation is an exactly squared
NRZ modulation, the average power reduction is well known and it is equal to
3 dB, in addition to the excess loss. Otherwise it has to be evaluated. This
average power reduction is the only parameter used in the SPT simulations
and it is requested in the related section of the parameter dialog window.
If you don’t know which is the value of this parameter, we suggest to perform
a VBS simulation of the subsystem containing the source and the modulator
to measure the optical power before and after the modulator. The average
power reduction is the difference (in dB) between the evaluated powers. Note
that the measured average power reduction contains also the excess loss.
Inputs Outputs
Electrical: the electrical signal used Optical: the modulated optical signal
to modulate the optical signal
Optical: the optical signal injected
in the device to be modulated
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the two models, Linear and
Custom.
Excess loss [dB]: the value of a constant extra loss introduced by the
modulator, which is independent of the electrical input signal. Allowed values:
min 0; max -; default 3.
8.3.1 Linear
This model implements a linear law for the loss introduced by the modulator
and the chirp factor α versus the applied voltage. To model this component
behavior, the minimum and maximum applied voltages and the
corresponding values of loss and chirp factor are required.
For as concerning loss is assumed to be a linear relation of the linear loss (not
dB) of the optical power with respect to the voltage of the applied electric
signal.
If we express the optical field at the output of the device as:
dφ
α = 2 I dt (2)
dI
dt
where
I = Eout (t )
2
(3)
You must provide loss and chirp factor values corresponding to minimum and
maximum applied voltages. If a voltage smaller than the minimum value
indicated is applied to the device, the loss and chirp factor values
corresponding to the minimum voltage are employed for optical field
modulation. If a voltage greater than the maximum value indicated is applied
to the device, the loss and chirp factor values corresponding to the maximum
voltage are employed for optical field modulation. Remember to properly set
the voltage range of the electrical driver to obtain the expected signal
modulation, chirping and extinction ratio.
Electro-optic frequency response can be considered as it is ideal or a filtering
characteristic can be inserted using a simple model or a user-defined one.
Chirping vs. assume. It must be smaller than the maximum value. Allowed values: min: -
Applied ; max: -; default: -5.
Voltage
Applied Voltage (Max) [V]: the maximum value that the electrical input
can assume. It must be greater than the minimum value. Allowed values:
min: -; max: -; default: 0.
Loss (Min) [dB]: the optical loss corresponding to the minimum applied
voltage. Allowed values: min: 0, max: -, default: 30.
Loss (Max) [dB]: the optical loss corresponding to the maximum applied
voltage. Allowed values: min: 0, max: -, default: 0.
Chirping factor (Min) [-]: the chirping factor corresponding to the
minimum applied voltage. Allowed values: min: -, max: -, default: 0.
Chirping factor (Max) [-]: the chirping factor corresponding to the
maximum applied voltage. Allowed values: min: -, max: -, default: 0.
Note
If the chirping factor corresponding to the minimum voltage is set to the
same value of the chirping factor corresponding to the maximum voltage, the
parameter is supposed to be constant versus applied voltage.
Frequency You can define the frequency response of the electro-optic effect of the
Response device, i.e. the spectral characteristic usually measured as S21(f) using the
opto-electronic network analyzer. Choices are the following.
Ideal: the frequency response is assumed flat and non-lossy.
Simple: the frequency response of the device is modeled as a low-pass
single pole filter.
Custom: the frequency response is defined through a data file.
-3 dB Bandwidth [GHz]: the –3 dB bandwidth of the filter. This field is
enabled only if Simple model is selected. Allowed values: min 0; max
BWSPT/2; default 2Rb (reference bit rate).
Filename: name of the data file where the frequency response is stored. This
field is enabled only if Custom model is selected. You can choose the file
from the file system clicking on the Browse button. You can also edit the file
clicking on the Edit button: it opens an editor window loading the data file.
The editor is VI for Unix or Notepad for Windows NT.
The filename extension must be DAT in capital letters and must contain two
columns of ASCII values. The first column represents the frequency values
for which the electrical response is defined. These values, expressed in GHz,
must be positive and in ascending order. The second column represents the
transfer function values, expressed in dB, corresponding to the first column
frequency values. For further details about the file format, please refer to
section 16.1.8
8.3.2 Custom
You can define the law of optical loss and chirp factor as a function of the
input voltage, through an ASCII data file.
Note
The file must be saved in the same directory where the model is saved.
The filename extension must be DAT in capital letters and must contain three
columns of ASCII values. The first column represents the applied voltage
values, which must be in ascending order. If the values are not in ascending
order, the correct behavior of the model is not guaranteed. The second
column represents loss values (expressed in dB) corresponding to the voltage
values of the first column. The loss values must be greater or equal to zero.
The third column represents chirping factors as described in [1]. If you don’t
want chirping factor varying with the applied voltage you must write the
same value in all the rows of this column.
If a voltage less than the minimum value (of the first column values) is
applied to the device, the loss and chirp values corresponding to the
minimum voltage are used for optical field modulation. If a voltage greater
than the maximum value (of the first column values) is applied to the device,
the loss and chirp value corresponding to the maximum voltage are
employed for optical field modulation.
Frequency You can define the frequency response of the electro-optic effect of the
Response device, i.e. the spectral characteristic usually measured as S21(f) using the
opto-electronic network analyzer. Choices are the following.
Ideal: the frequency response is assumed flat and non-lossy.
Simple: the frequency response of the device is modeled as a low-pass
single pole filter.
Custom: the frequency response is defined through a data file.
-3 dB Bandwidth [GHz]: the –3 dB bandwidth of the filter. This field is
enabled only if Simple model is selected. Allowed values: min 0; max
BWSPT/2; default 2Rb (reference bit rate).
Filename: name of the data file where the frequency response is stored. This
field is enabled only if Custom model is selected. You can choose the file
from the file system clicking on the Browse button. You can also edit the file
clicking on the Edit button: it opens an editor window loading the data file.
The editor is VI for Unix or Notepad for Windows NT.
The filename extension must be DAT in capital letters and must contain two
columns of ASCII values. The first column represents the frequency values
for which the electrical response is defined. These values, expressed in GHz,
must be positive and in ascending order. The second column represents the
transfer function values, expressed in dB, corresponding to the first column
frequency values. For further details about the file format, please refer to
section 16.1.8.
where:
V − V0
φ = π in (2)
Vπ
SPT Behavior
The purpose of SPT simulations is not to analyze the modulation of the
channels, but to study the systems in terms of average power per channel
and noise level. The optical phase modulator does not change the average
power of the input optical source, except for the attenuation introduced by
the excess loss. Hence, the optical phase modulator, during SPT simulations,
simply introduces the excess loss attenuation.
Inputs Outputs
Electrical: the electrical signal used Optical: the modulated optical signal
to modulate the optical signal
Optical: the optical signal injected
in the device to be modulated
Parameters
Name: the optional name that appears in the component label.
Excess loss ELdB [dB]: the value of a constant extra loss introduced by the
modulator, which is independent of the electrical input signal. Allowed values:
min 0; max -; default 3.
Zero Phase Voltage V0 [V]: the input voltage that sets the modulator to
the zero phase shift point. Allowed values: min -; max -; default 0.
Voltage Vπ [V]: the input voltage swing that generates a π (180°) phase
shift on the optical signal. Allowed values: min 0; max -; default 5.
Electrical Filtering with sin(πf/Bw)/(πf) Law: specifies whether or not the
electrical input is filtered by the following frequency transfer function.
πf
sin
H(f ) = B0 (3)
πf
B0
where:
B0 ≅ 2.25 ⋅ BW (4)
-3dB Bandwidth [GHz]: the -3 dB cutoff frequency of the lowpass filter; it's
available only if the Electrical Filtering with sin(πf/Bw)/(πf) Law is set.
Allowed values: min 0, max 2.25(BWSPT/2), default Rb (reference bit-rate).
First filter notch [GHz]: the resulting first notch in the transfer function.
The transfer function of the filter model (3) is typical for several optical
devices where an optical beam and a traveling wave electrical signal interact
over a given non-negligible length. The filtering function is consequently
determined by the amount of time required by the electrical signal to
propagate along the device. If you need to simulate different filtering
functions than the previous one, simply disable the Electrical Filtering with
sin(πf/Bw)/(πf) Law option, and choose a proper external filter to be
inserted between the electrical signal generator and the electrical input of the
modulator.
VS3 VS1
^x
Ein Eout
PBS
^y
Amplitude Phase
Section Section
The first electrical signal VS3 is used to split the power of the input optical
signal between horizontal and vertical polarization. If no voltage is applied
and if the device is symmetric, i.e. when Vbal,S3 = 0, the power is equally split
on the two polarization axis, thus obtaining a 45° linear polarization at the
output.
Using the second signal VS1, the relative phase of the two linear polarization
is changed, so that elliptical states of polarization can be obtained.
The component can be used only when the Dual Polarization option is
chosen in the Simulation Parameter dialog window otherwise an error
message is given.
For a brief description of the POLSK modulation features, see section 13.2
and section 12.2 for binary POLSK driver and receiver description. See also
the “POLSK Project Bibliography” at the end of this chapter.
Note
This component works as previously outlined only if the input polarization is
aligned on the x axis. For example, it works properly whenever it is placed
after a laser or an optical pulse generator.
SPT Behavior
Since SPT simulations are based on the power spectrum of the optical signal,
the information about polarization can not be managed with the SPT
simulation technique. Thus SPT simulation does not work with projects
containing this component.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
Excess loss [dB]: the value of a constant extra loss introduced by the
modulator, which is independent of the electrical input signals. Allowed
values: min 0; max -; default 3.
Voltage of Zero Degree Rotation on S3 Axis (Vbal,S3) [V]: the balance
voltage of the power splitter section. Allowed values: min -; max -; default 0.
Voltage of Zero Degree Rotation on S1 Axis (Vbal,S1) [V]: the balance
voltage of the phase modulation section. Allowed values: min -; max -;
default 0.
360° Rotation Voltage, both Axes (V2π) [V]: the input voltage swing that
generates a 2π (360°) rotation of the State of Polarization of the input optical
signal around the S3 or S1 axis in Stokes space. Allowed values: min 0; max -
; default 5.
Electrical Filtering with sin(πf/Bw)/(πf) Law: specifies whether or not the
electrical input is filtered by the following frequency transfer function.
πf
sin
H(f ) = B0 (3)
πf
B0
where:
B0 ≅ 2.25 ⋅ BW (4)
-3dB Bandwidth [GHz]: the -3 dB cutoff frequency of the lowpass filter; it's
available only if the Electrical Filtering with sin(πf/Bw)/(πf) Law is set.
Allowed values: min 0, max 2.25(BWSPT/2), default Rb (reference bit-rate).
The transfer function of the filter model (3) is typical for several optical
devices where an optical beam and a traveling wave electrical signal interact
over a given non-negligible length. The filtering function is consequently
determined by the amount of time required by the electrical signal to
propagate along the device. If you need to simulate different filtering
functions than the previous one, simply disable the Electrical Filtering with
sin(πf/Bw)/(πf) Law option, and choose a proper external filter to be
inserted between the electrical signal generator and the electrical input of the
modulator.
8.6 References
[1] F.Koyama and K. Iga, Frequency chirping in external modulators, J.
Lightwave Technol., vol.6, no.1, pp.87-33, Feb.1988.
[2] Govind P. Agrawal, Nonlinear Fiber Optics, 2nd edition, Academic Press,
San Diego, 1995
SPT Behavior
EDFA devices can be modeled as linear components defined through their
transfer functions as long as the gain transient effects are neglected, so that
a steady-state gain and noise figure can be defined. Thus they can be easily
simulated in SPT simulation that is based on the transfer function model of
each optical component. In formulae, given the SPT input power spectrum
Pin(f), the output spectrum is evaluated as:
Note
Setting Yes this option can slow-down the simulation.
Optical Optical
Parameters
Gain Shape: specifies if the gain is described as a scalar number (Flat) or
through a data file as a function of wavelength (From file).
Gain [dB]: the amplifier gain in dB; it is enable if the Flat option is selected.
Allowed values: min 0; max -; default 25.
Gain Shape File Name: name of the data file where the gain (dB) vs.
wavelength (nm) function is stored. You can choose the file from the file
system clicking on the Browse button. You can also edit the file clicking on
the Edit button: it opens an editor window loading the data file. The editor is
VI for Unix or Notepad for Windows NT. The filename extension must be DAT
in capital letters. See section 9.1.6 for further details on the file format.
No Noise: when selected, noise should is not added to the signal (ideal
case). This feature can help to investigate on effects usually “buried” in the
ASE noise.
Noise figure: specifies if the noise figure is described as a scalar number
(Flat) or through a data file as a function of wavelength (From file).
F [dB] or [lin]: specifies the noise figure, see section 9.1.8 for details on the
EDFA noise figure. This field is enable only if the Flat option is selected.
Allowed values: min 0 (dB); max -; default 4.5 (dB).
Noise figure File Name: name of the data file where the noise figure (dB)
vs. wavelength (nm) function is stored. You can choose the file from the file
system clicking on the Browse button. You can also edit the file clicking on
the Edit button: it opens an editor window loading the data file. The editor is
VI for Unix or Notepad for Windows NT. The filename extension must be DAT
in capital letters. See section 9.1.6 for further details on the file format.
where:
G0 is the small signal gain
P3dB is the output saturation power, i.e. the output power where the gain
is reduced to G0/2 i.e. to one half of the small signal gain (under the
approximation G0>>1)
Pin is the total input power
This model is obtained under following assumptions:
high small signal gain, i.e. G0>>1
output power not much greater than P3dB
uniform pumping of the amplifier
negligible self-saturation due to the internally generated ASE noise
An example of the input-output transfer functions is given in EDFA example,
together with several comments on the model properties.
The same formula can be rewritten, introducing the total input and output
powers and the explicit dependence from wavelength as:
Pin − Pout
log 2
G (λ ) = G0 (λ )e P3dB
(2)
This is the approximated formula used for the wavelength dependent case.
Basically, the actual gain G(λ) is estimated as the small signal gain G0(λ)
multiplied by the scalar value:
Pin − Pout
log 2
P3dB
K =e (3)
Note that this term contains the output power which in turns depend on the
gain G(λ), so that the equation could only be solved iteratively. Consequently,
saturation affects the absolute value of the gain but not its spectral shape.
Inputs Outputs
Optical Optical
Parameters
Output Saturation Power [dBm] or [mW]: the output saturation power
P3dB. Allowed values: min -; max -; default -7 (dBm).
Gain Shape: specifies if the gain is described as a scalar number (Flat) or
through a data file as a function of wavelength (From file).
Small Signal Gain [dB]: the amplifier gain in dB; it is enable if the Flat
option is selected. Allowed values: min 0; max -; default 25.
Gain Shape File Name: name of the data file where the gain (dB) vs.
wavelength (nm) function is stored. You can choose the file from the file
system clicking on the Browse button. You can also edit the file clicking on
the Edit button: it opens an editor window loading the data file. The editor is
VI for Unix or Notepad for Windows NT. The filename extension must be DAT
in capital letters. See section 9.1.6 for further details on the file format.
No Noise: selects if no noise should be added to the signal (ideal case). This
feature can help to investigate on effects usually “buried” in the ASE noise.
Noise figure: specifies if the noise figure is described as a scalar number
(Flat) or through a data file as a function of wavelength (From file).
F [dB] or [lin]: specifies the noise figure, see section 9.1.8 for details on the
EDFA noise figure. This field is enable only if the Flat option is selected.
Allowed values: min 0 (dB); max -; default 4.5 (dB).
Noise figure File Name: name of the data file where the noise figure (dB)
vs. wavelength (nm) function is stored. You can choose the file from the file
system clicking on the Browse button. You can also edit the file clicking on
the Edit button: it opens an editor window loading the data file. The editor is
VI for Unix or Notepad for Windows NT. The filename extension must be DAT
in capital letters. See section 9.1.6 for further details on the file format.
Examples
An example of the input output-transfer functions for a flat case saturating
amplifier with G0=30 dB and P3dB =10 dBm is given in the following pictures.
The last one shows the input-output power transfer function.
The curves can be divided into three regions:
1. for an input power smaller that approx. -20 dBm, the amplifier performs
linearly and has an amplification equal to the small signal amplification
G0=30 dB;
2. for an input power higher than approx. +20 dBm, the amplifier is
completely saturated and the amplification is close to 0 dB;
3. in the intermediate region, i.e. for power close to P3dB =10 dBm, the
amplifier is on the onset of saturation; the amplification gain decreases
from G0 to 0 dB when increasing input power.
Figure 9.2 Gain as a function of input power for G0=30 dB and P3dB =10 dBm.
Figure 9.3 Output power as a function of input power for G0=30 dB and P3dB
=10 dBm.
Optical Optical
Parameters
Output Power [dBm] or [mW]: sets the value of the total power (signal +
noise) at the output of the amplifier. Allowed values: min - (dBm); max -;
default 7 (dBm).
Gain Shape: specifies if the gain is described as a scalar number (Flat) or
through a data file as a function of wavelength (From file).
Maximum Small-Signal Gain [dB]: it is the maximum gain the amplifier
can give to the signal to ensure the requested output power. It is enable only
if the Flat option is selected. Allowed values: min 0; max -; default 35.
Gain Shape File Name: name of the data file where the gain (dB) vs.
wavelength (nm) function is stored. You can choose the file from the file
system clicking on the Browse button. You can also edit the file clicking on
the Edit button: it opens an editor window loading the data file. The editor is
VI for Unix or Notepad for Windows NT. The filename extension must be DAT
in capital letters. See section 9.1.6 for further details on the file format.
No Noise: selects if no noise should be added to the signal (ideal case). This
feature can help to investigate on effects usually “buried” in the ASE noise.
Optical Optical
Parameters
Amplifier Type: several default models are available. They are all based on
the sample description file DOPEDFIB.DAT that will be copied in the working
directory if one of these default types is selected. The defaults are:
Booster 12 dBm, booster amplifier having a maximum output power of
the order of 12 dBm when the input power is in the -6 dBm to 0 dBm
range
Note
The frequency of the co-propagating pump must be inside the bandwidth
defined in the doped fiber data file.
Note
The frequency of the counter-propagating pump must be inside the
bandwidth defined in the doped fiber data file.
Doped Fiber Data: name of the data description file. You can choose the file
from the file system clicking on the Browse button. You can also edit the file
clicking on the Edit button: it opens an editor window loading the data file.
The editor is VI for Unix or Notepad for Windows NT. The filename extension
must be DAT in capital letters.
Gain and Noise Spectra Plots: enables the plotting on screen of the gain
and noise figure of the amplifier. In very long simulation making use of tens
or hundreds of amplifiers, we suggest to disable this option, to avoid the
accumulation of a exceedingly large number of data files.
1542.19946 7.56606483
1542.39941 7.68924379
1542.59937 7.80922413
....
For the physical EDFA model, in the overlap integral approximation, the
data must be written as follows:
− a first line must contain the value of the saturation parameter in
[m-1s-1] followed by a line containing the character '#', see section
9.1.9
− the following lines should contain, for each line, the wavelength in
[nm] followed by the emission coefficient in [dB/m], the absorption
coefficient in [dB/m] and the intrinsic loss in [dB/m] at the specified
wavelength
Note
The bandwidth defined in the file must be greater than the SPT bandwidth.
Here is an example.
#
# EXAMPLE DATA FOR OptSim EDFA MODEL
#
# The first parameter is the saturation power,
# followed by the conventional
# separation sign # and then by the four columns:
# wavelength [nm] emission [dB/m] absorption [dB/m] loss [dB/m]
#
##
8.67e15 saturation parameter [m^-1 s^-1]
#
1450. 0.0189598892 0.104602687 0.0250000004
1450.34485 0.0364919864 0.199739471 0.0250000004
1450.6897 0.0526750833 0.285307109 0.0250000004
1451.03455 0.0677747428 0.366118371 0.0250000004
1451.37927 0.0818995684 0.438738942 0.0250000004
1451.72412 0.0946921483 0.507609963 0.0250000004
1452.06897 0.107067831 0.569540262 0.0250000004
1452.41382 0.11978811 0.629318237 0.0250000004
1452.75854 0.130077243 0.68602562 0.0250000004
1453.10339 0.141509846 0.740024388 0.0250000004
1453.44824 0.152790979 0.79181397 0.0250000004
...
...
PASE = F (G − 1)hv∆v / 2
where F is the noise figure parameter, G the amplifier gain, h the Planck's
constant and ν the optical frequency.
This definition for the output ASE noise evaluation makes sense only when
G>1, otherwise PASE would be negative. You should thus avoid using the
simplified models and specifying a gain smaller than one. Anyway, no error
message is given in this “unphysical” situation, and simply the output ASE
power is set to zero.
nt
χ = πbeff2 (1)
τ
where beff is the effective radius of the doped region (and consequently πbeff2
is the effective transversal area of the fiber doped region), nt is the average
ion density in [m-3] and τ is the so-called “metastable lifetime”.
The gain g(λ) and absorption α(λ) coefficients can be obtained from the
emission σem(λ) and absorption σabs(λ) cross-sections as:
and:
where Γ(λ) is the overlap integral between the optical mode inside the fiber
and the erbium doping profile, while nt is the average ion density in [m-3].
The emission and absorption coefficients in [m-1] should then be converted in
[dB m-1] according to:
SPT Behavior
SOAs are highly non-linear devices. Their simulation as linear components
defining through an input-output transfer function may not be reliable. The
simulation of SOAs is anyway allowed in SPT, but a warning message saying
that results may be unreliable is shown. In SPT, the steady state gain (which
will be defined in the following pages) is used to propagate power spectrum
through the devices by simply assuming that: Pout ( f ) = Gsteady − state ⋅ Pin ( f ) .
In general, we advise to use VBS whenever SOA are involved.
Inputs Outputs
Optical Optical
Model description
The implemented model was initially proposed by Saleh [4], and it is based
on the analysis of a wave equation for the optical field propagation and on a
rate equation for the carrier density evolution. It does not consider device
parameter frequency dependence.
To better understand the above analysis, we resume in the following table the
entire set of symbols and parameters used.
Symbol Meaning
E Electrical field
I SOA bias current (assumed to be time independent)
C Light speed
ε0 Vacuum permittivity
N Active material refractive index
χ Dielectric susceptibility
N Carrier density
α Linewidth enhancement factor
g(z,t) Gain
Γ Confinement factor
A Material gain constant
Symbol Meaning
αP Material loss
N0 Optical transparency carrier density
L Active region length
q Electron charge
V Active region volume
A Active region transverse area
nc
χ (N ) = − (α − j )g (N ) (2)
ω
where g(N) is the carrier density dependent gain. In a small signal analysis,
g(N) is linearly dependent on carriers through the following law:
g ( N ) = Γ[a( N − N 0 )] − α P (3)
∂E 1 + jα
= g ( z , t )E ( z , t ) (4)
∂z 2
Integrating along z, we obtain:
1 + jα L
E (L, t ) = E (0, t )exp ∫ g ( z , t )dz (5)
2 0
On the other hand the rate equation governing the carrier density is given by:
dN I N g (n )
E (t , z )
2
= − − (6)
dt qV τ s hω
This equation describes the carrier density inside the active region and its
time evolution due to the field variation. Please note that the bias current I is
assumed to be time-independent, i.e., just a DC component.
Assuming the gain to be linearly dependent on the carriers, this equation is
modified taking into account Eq.(3) as:
dg ( z , t ) 1 ∂
= −[g ( z , t ) − g 0 ] − E (z, t )
2
τs (7)
dt Ps ∂z
where we have introduced the following parameters:
Ahω
Ps = (saturation power) (8)
Γ aτ s
Iτ s
N0 = (transparency carrier density) (9)
qV
[
g 0 = Γa N 0 − N 0 ] (10)
d E (L, t ) 2 E (0, t ) 2
1 + τ s dt [G (t ) − G0 ] = − P −
Ps
(11)
s
where:
L
G0 = g 0 L and G (t ) = ∫ g ( z , t )dz
0
Numerical solution
The present SOA model, proposed by Saleh [4], is thus characterized by the
following two coupled equations:
1 + jα
E (L, t ) = E (0, t )exp G (t ) (12)
2
d E (L, t ) 2 E (0, t ) 2
1 + τ s dt [G (t ) − G0 ] = − P −
PS
(13)
S
An implicit solution for the evolution of the optical field is thus given by:
1 + jα
E (L, t ) = E (0, t ) ⋅ exp G0
2
(14)
1 + jα E (L, t ) 2 E (0, t ) 2 1 − t
⋅ exp − − * e τs
2 PS PS τ s
(the symbol * means convolution). This equation can be solved using the
component diagram shown in the following figure.
E(0,t) 1 + jα E(L,t)
exp G0
2
exp(⋅ )
2 Filter
⋅ 1 + jα
H (ϖ ) = −
PS 2 (1 + jωτ s )
2
⋅
PS
Parameters
Name: the optional name that appears in the component label.
Bias Current I [mA]: the input bias current. Allowed values: min 0; max -;
default 100.
9.3 Reference
[1] C. R. Giles and E. Desurvire, Modeling erbium-doped fiber amplifiers, J.
Lightwave Technol., vol. 9, no. 2, pp. 271-283, Feb. 1991.
[2] C. R. Giles and E. Desurvire, Propagation of signal and noise in
concatenated Erbium-doped fiber optical amplifiers, J. Lightwave
Technol., vol. 9, no. 2, pp. 271-283, Feb. 1991.
[3] W. Zeiler, F. Di Pasquale, P. Bayvel, J. Midwinter, Modeling of four wave
mixing and gain peaking in amplified WDM optical communication
systems and networks, J. Lightwave Technol., vol. 14, no. 9, pp. 1933-
1942, Sept. 1996.
[4] A. A. M. Saleh, Nonlinear models of travelling-wave optical amplifiers,
Electronics Letters, vol. 24, n. 14, pp. 835-837, Jul. 1988.
[5] S. Benedetto, I. Maio, P. Poggiolini and P. Zanini, Simulation of
multicarrier traveling-wave semiconductor laser amplifiers, IEEE-
Photonics Technology Letters, vol. 3, n. 8., pp.712-714, Aug. 1991.
SPT Behavior
Optical filters are linear components defined through their transfer functions,
thus they can be fully simulated in SPT simulations that is based on the
transfer function model of each optical component. Therefore, the optical
spectrum is filtered using the predefined or user-defined transfer function.
Since SPT simulations are based on the propagation of power spectra, only
the modulus of the transfer functions is relevant. In formulae, given the SPT
input power spectrum Pin(f), the output spectrum is evaluated as:
2
Pout ( f ) = H ( f ) ⋅ Pin ( f )
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the models, Multi-Stage
Lorentzian, User-Defined, Supergaussian, Bessel and Raised-Cosine.
Filter Synthesis: selects the filter type bandpass or notch [1-|H(f)|].
f 3dB = ∆f s 21 N
−1 (2)
The multiple stage Lorentzian filter, after a proper fitting of its three free
parameters f0 , f3dB and N to actual measurements, can be successfully used
to simulate many types of single input – single output optical filters.
Moreover, a bank of parallel multiple stage Lorentzian filters with different f0
but identical f3dB and N, allows to simulate realistic WDM receiver.
For more details, see the related examples.
Inputs Outputs
Optical Optical
Parameters
Number of Stages N [-]: the stages N of the transfer function (1). Allowed
values: min 1; max -; default 1.
Center Frequency [THz] or [nm]: the center frequency f0 of the transfer
function (1). Allowed values: min flowSPT; max fupSPT; default f0,SPT.
-3dB Two-Sided Bandwidth [GHz] or [nm]: the range between the two
BWVBS
cut-off frequencies. Allowed values: min 0 (GHz); max 21 N − 1 ;
2
BWVBS
default min 60, (GHz).
2
Note
If the specified 3-dB filter bandwidth is too large compared to the simulation
bandwidth, an error message is given. In this case, you can either reduce the
filter bandwidth or increase the global bandwidth.
Example
The single stage Lorentzian filter has a rather smooth roll-off, since it is a
single pole bandpass filter. In order to obtain a filter with sharper roll-off, you
can increase the number of stages N while keeping the same f3dB bandwidth.
As an example, we report in the following pictures the transfer function for
Lorentzian filters having the same bandwidth f3dB and N set to 1,5 and 10.
Figure 10.1 Lorentzian filter transfer function for f3dB=1 and N set to 1,5,10
Bw
1 , f − f 0 < (1 − R0 )
2
α
1 π B Bw B
H( f ) = α 1 − sin f − f 0 − w , (1 − R0 ) < f − f 0 < (1 + R0 ) w
2 Ro Bw 2 2 2
B
0 , f − f 0 > (1 + R0 ) w
2
(1)
where f0 is the filter center frequency, Bw is the FWHM (Full Width Half
Maximum) bandwidth and Ro is so-called roll-off parameter (see figure
below). The parameter α is the exponent at which the standard raised cosine
function is raised. For further details on raised cosine filters, see reference
[1].
The filter is numerically implemented using the “Overlap-and-Add” algorithm
(see reference [3]). This is the same algorithm used for the linear operator of
the fiber component. This algorithm allows a very accurate implementation of
any complex transfer functions.
The optical raised cosine filter is useful in many simulations where the exact
shape of the filters are not know in details, and one wants to describe filters
simply with their central frequency and passband.
(1+R0)Bw /2
H(f) , α = 1
(1-R0)Bw /2
1
Bw
0.5
0
f
f0
Inputs Outputs
Optical Optical
Parameters
Raised-Cosine exponent [-]: the exponent α in (1). Allowed values: min 0;
max 1; default 1.
Raised-Cosine Roll-off [-]: the filter roll-off Ro in (1). Allowed values: min
0; max 1; default 0.5.
Center Frequency [THz] or [nm]: the filter center frequency f0 in (1).
Allowed values: min flowSPT; max fupSPT; default f0,SPT.
Bandwidth [GHz] or [nm]: the absolute bandwidth Bw of the filter. Allowed
values: min 0 (GHz); max -; default 60 (GHz).
Amplitude and group delay plots: enables the plotting on screen of the
transfer function and group delay obtained with the above parameters,
together with the transfer function and group delay currently implemented by
the component after the filter synthesis.
10.1.3 Supergaussian
This model implements a Supergaussian filter, i.e. a filter having the following
transfer function:
( f − f 0 )2 N
H ( f ) = exp −
(1)
2∆f 2 N
where f0 is the filter central frequency and N is the filter order. ∆f is
automatically evaluated to have a global 3-dB two-sided bandwidth equal to
the value f3dB specified in the parameter dialog window. The relation between
the two parameters is given by:
f 3dB
∆f = (2)
2 ln (2)
2N
Inputs Outputs
Optical Optical
Parameters
Supergaussian Order []: order N of the supergaussian filter transfer
function. Allowed values: min 1; max -; default 1.
Center Frequency [THz] or [nm]: the center frequency f0 of the transfer
function (1). Allowed values: min flowSPT; max fupSPT; default f0,SPT.
-3dB Two-Sided Bandwidth [GHz] or [nm]: the range between the two
cut-off frequencies. Allowed values: min 0 (GHz); max BWVBS; default
BWVBS
min 60, .
2
Amplitude and group delay plots: enables the plotting on screen of the
transfer function and group delay obtained with the above parameters,
together with the transfer function and group delay currently implemented by
the component after the filter synthesis.
10.1.4 Bessel
This model implements a Bessel filter. The characteristics of this standard
family of filters can be found in any book on analog filters, such as [4].
Inputs Outputs
Optical Optical
Parameters
Number of poles: the number of Bessel filter poles Allowed values: min 1;
max 8; default 1.
Center Frequency [THz] or [nm]: the center frequency f0 of the filter
transfer function. Allowed values: min flowSPT; max fupSPT; default f0,SPT.
-3dB Two-Sided Bandwidth [GHz] or [nm]: the range between the two
cut-off frequencies. Allowed values: min 0 (GHz); max BWVBS; default
BWVBS
min 60, .
2
Amplitude and group delay plots: enables the plotting on screen of the
transfer function and group delay obtained with the above parameters,
together with the transfer function and group delay currently implemented by
the component after the filter synthesis.
gratings
in general, any transfer function of experimentally measured linear device
The numerical algorithm used to implement these components is very
accurate and advanced; see “User-defined Filter Synthesis” in this section for
further details.
The component is very flexible, and the parameters can be inserted in
different formats, namely:
abscissa in [THz] (frequency) or [nm] (wavelength)
absolute frequency (wavelength) or relative to a specified reference
frequency, see the Simulation Parameters description in the OptSim User
Manual
amplitude on a linear scale or in dB
transfer function as a function of group delay or phase
Moreover, the transfer function may be shifted and/or scaled on the
frequency axis, so that you can define a filter shape and then use it in
different situations (see “Frequency Translation” in this section for further
details).
The component may generate an output chart where the following graphs are
shown:
transfer function as specified in the user description file (modulus and
phase or group delay)
currently synthesized transfer function(modulus and group delay)
This component is implemented using the same algorithm (Overlap and add
[3]) used for the linear operator of the fiber component. This algorithm allows
a very accurate implementation of any complex transfer functions. Anyway,
in some peculiar situations, the digital filter synthesis may not be possible.
These situations are usually due to:
1. Some of the input parameters do not match among themselves; for
example, the filter is specified on a frequency range, which is completely
outside the simulation bandwidth.
2. The filter is specified on a frequency range much smaller than the
simulation bandwidth.
3. The filter transfer function comes from an experimental measure where
the measurement uncertainty is high, and thus the measured transfer
function has a lot of noise superimposed to the actual transfer function.
4. The filter has a exceedingly long time-domain impulse response; for
example a bandpass filter whose bandwidth is several orders of
magnitude smaller than the simulation bandwidth.
In all these cases, the simulation is stopped and an error message is given.
Refer to “If Something goes Wrong...” in this section for further details.
Inputs Outputs
Optical Optical
Parameters
Filter Data Filename: name of the data file where the transfer function vs.
frequency is stored. You can choose the file from the file system clicking on
the Browse button. You can also edit the file clicking on the Edit button: it
opens an editor window loading the data file. The editor is VI for Unix or
Notepad for Windows NT. See “User-Defined Filter File Format” in this section
for further details.
File Format Wavelength or Frequency: specifies if the abscissa values of the transfer
function in the description file are expressed in wavelength or frequency.
Ampl. Square and Phase, Ampl. Square and Group Delay or Ampl.
Square only: specifies if the transfer function in the description file is
expressed as a function of amplitude and phase, amplitude and group delay
or amplitude only (the amplitude should always be expressed as the square
of the modulus of the transfer function).
Amplitude Square: specifies whether the transfer function amplitude in the
description file is expressed in dB or in a linear scale.
Absolute abscissa or Relative to Reference Frequency: specifies if the
abscissa values of the transfer function in the description file are expressed in
absolute values or with an offset respect to the reference frequency specified
in the fields below (the allowed values for these field are: min >0; max -;
default f0,SPT).
Smoothing: enables to smoothen the data points supplied in the user data
file. See “Transfer Function Smoothing” in this section for further details.
Smoothing factor [-]: the smoothing integer factor to be applied when
synthesizing the filter. This field is enabled only if Smoothing is set. See
“Transfer Function Smoothing” in this section for further details. Allowed
values: min 1; max -; default 1.
Offset [THz] or [nm]: specifies a shift between the current simulated
transfer function and the transfer function stored on the filter description file.
It should be set to 0 (the default) when offset is not required. See “Frequency
Translation” in this section for further details. Allowed values: min -; max -;
default 0.
Scaling [-]: the real number that specifies the expansion/stretch factor in
the frequency scale of the transfer function stored in the data file. This
parameter should be set to 1 (the default) when scaling is not required. See
“Frequency Translation” in this section for further details. Allowed values: min
>0; max -; default 1.
Actual Filter Amplitude and Group Delay Plots: enables the plotting on
screen of the supplied transfer function and group delay data, together with
the transfer function and group delay currently implemented by the
component. Please note that when this filter is placed in an N-time “iterated”
area and this option is set, N identical charts are generated. In this and other
similar situation, it is advisable to set this option off.
Comment lines are allowed only after the first line and above the '##'
line.
The data must be written in lines, separated by blank, each containing:
− the frequency in [THz] (or wavelength in [nm]) as the first field
− the modulus of the transfer function (linear scale or dB)
− when required, the phase in radians or the group delay in ps
Abscissas (either frequency or wavelength) should be given in ascending
order while abscissas are not required to be strictly equispaced. Two or more
points with the same frequency (or wavelength) are not allowed: if this
situation happens only the first line containing the repeated value is
considered and a warning message is shown. The maximum number of
allowed points is 4095; if the file contains more than 4095 points, only the
first 4095 points are considered and a warning message is shown.
Here is an example of input file:
#
# Filter description file
#
# General comments for the filter:
#
# - Absolute Abscissae in [THz]
# - Amplitude on a linear scale
# - Phase in radians
#
##
192.449997 0.00819663145 6.05006695
192.450546 0.0082128942 6.03798819
192.451096 0.00822920538 6.02592087
192.45166 0.00824601948 6.01353168
192.452209 0.00826242939 6.00148916
192.452759 0.00827888772 5.98945856
192.453308 0.00829539541 5.97744036
192.453873 0.00831241254 5.96510124
192.454422 0.00832901988 5.95310688
...
Please note that, accordingly with the OptSim units for optical signals, the
units to be used in the description file are:
frequency in [THz]
wavelength in [nm]
phases in [radians]
group delay in [ps].
Frequency Translation
The transfer function specified on the filter description file can be shifted
and/or compressed/expanded with respect the frequency (or wavelength)
axis by the use of the two parameters Offset and Scaling. This feature may
prove very useful to define a given filter on a description file, and then
reusing it in different position on the frequency axis.
As an example of how offset and scaling works, the Figure 10.4 shows a
typical sin(x)/x filter centered at 1550 nm, while the following transfer
functions are obtained as follows:
by setting the Offset parameter to -3 nm, the resulting transfer function
is shifted as shown in Figure 10.5
by setting the Scaling parameter to 2, the resulting transfer function is
stretched as shown in Figure 10.6
Figure 10.5 The filter specified in Figure 10.4, after setting Offset parameter
to -3 nm
Figure 10.6 The filter specified in Figure 10.4, after setting Scaling parameter
to 2
Figure 10.7 The filter specified in Figure 10.4, after setting Scaling parameter
to 2 and the Offset parameter to -3 nm
An incorrect use of this two parameters may give errors, for example when
the resulting transfer function is defined on a range which does not overlap at
all with the current VBS and/or SPT simulation bandwidths. See “If
Something goes Wrong...” in this section for further details.
Figure 10.8 An ideal bandpass filter (red line) and a very noisy version of the
same filter (blue line)
Figure 10.9 The filter shown in Figure 10.8, after smoothing with Nfact=4
10.2 References
[1] Leon W. Couch II, Digital and analog communication system, Third
Edition, 497-508, 1989.
[2] C. De Boor, A practical guide to Splines, Springer-Verlag, 1978
[3] M. C. Jeruchim, P. Balaban e K. S. Shanmugan, Simulation of
Communication Systems, Plenum Publishing Corporation, New York,
1992.
[4] A. B. Williams and F. J. Taylor, Electronic Filter Design Handbook,
McGraw-Hill, New York, 1988.
11.1 Photodiode
This component simulates a photodiode. Two different models are
implemented: PIN and APD. The quantum noise generated by photodetection
and avalanche multiplication, the limited bandwidth of the device and the
presence of a dark current are taken into account.
SPT Behavior
This component converts the optical signal into an electrical signal. SPT
simulations manage optical signals only thus this component is not included
in SPT simulation.
Parameters
Name: the optional name that appears in the component label.
Photodiode Type: opens the related dialog windows for the models, APD
and PIN.
where
q
R =η (2)
hv
is the responsivity [A/W] (at the reference frequency ν), η is the quantum
efficiency, In is the shot noise current and Id is the mean value of the dark
current.
Shot noise is approximated as a white process with Gaussian distribution. Its
unilateral Power Spectral Density is flat and is given by Gn(f)=2q(Is+Id),
where Is=RPs. If the Quantum Noise is disabled, no noise is added and
ITot=RPs+Id.
Note that the noise process In is a non-stationary one because its statistical
characterization depends on the value of the instantaneous input power Ps.
Therefore, in principle, the conventional definition of noise Power Spectral
Density, which requires stationarity, could not be applied. However, it can be
shown that in this case, in the model outlined above, a white noise process
whose Power Spectral Density varies in time is valid. To do so, one resorts to
Campbell's theorem for Poisson-distributed non-stationary random processes.
This matter is dealt with in detail in [1].
The photodiode finite bandwidth is modeled using a single-pole transfer
function, of -3 dB bandwidth BW. Filtering can be turned off, in this case the
device has ideally infinite bandwidth.
Inputs Outputs
Optical Electrical
Parameters
Reference Frequency/Wavelength [THz] or [nm]: the reference
frequency for the quantum efficiency and the responsivity parameters.
Allowed values: min 0; max -; default f0,SPT.
Quantum Efficiency η: specifies how many electron-hole pairs are
generated per each incident photon, on average. Allowed values: min 0; max
1; default 0.7.
Responsivity (at reference frequency) R [A/W]: the slope of the current
vs. input optical power. The responsivity is related to quantum efficiency by
(2). Therefore, setting one also sets the other. Allowed values: min 0. A/W;
max 1.25 A/W; default 0.875 A/W.
Single-Pole Electrical Filtering: enables the single-pole filtering. If
disabled, the device bandwidth is unlimited.
-3dB Bandwidth BW [GHz]: the frequency of the single-pole of the
photodiode transfer function. It is enabled only if the previous option is
enabled. Allowed values: min 0; BWSPT/2, default 2Rb (reference bit rate).
Quantum Noise: enables the generation of noise. If disabled, the device is
completely noiseless.
Dark Current Id [nA]: the average value of the current generated by the
reverse bias voltage applied to the photodiode. Allowed values: min 0; max -;
default 0.1.
Ideal: sets the values of parameters as an ideal photodiode, that is
responsivity (at reference frequency) R = 1 A/W and Dark Current Id =0 nA,
and disables electrical filtering and noise generation.
Where
q
R =η (2)
hv
is the responsivity [A/W] (at the reference frequency ν), η is the quantum
efficiency, G is the avalanche gain, Id is the mean value of the dark current
before avalanche multiplication and In is the overall noise current.
Noise is approximated as a white process with Gaussian distribution. Its
unilateral Power Spectral Density is flat and is given by
Gn ( f ) = 2q (I s + I d )G 2 F (3)
where Is=RPs and F is the APD noise figure. Note that if F=1 then there is no
degradation due to the avalanche gain, since both the output signal power
and noise power would increase by G2, leaving the S/N ratio unaltered with
respect to the case of no gain.
If the Photodetection noise is disabled, noise is not simulated and
ITot = GRPs + GI d
Note that the noise process In is non-stationary because its statistical nature
depends on the value of the instantaneous input power Ps. Therefore, in
principle, the conventional definition of noise Power Spectral Density, as a
stationary process, could not be applied. However, it can be shown that in
this case the model outlined above, a white noise process whose Power
Spectral Density varies in time, is valid. To do so, one needs to resort to
Campbell's theorem for Poisson-distributed non-stationary random processes
[1].
The photodiode finite bandwidth is modeled using a single-pole transfer
function, of -3 dB bandwidth BW. Filtering can be turned off In this case the
device has ideally infinite bandwidth.
Inputs Outputs
Optical Electrical
Parameters
Reference Frequency/Wavelength [THz] or [nm]: the reference
frequency for the quantum efficiency and the responsivity parameters.
Allowed values: min 0; max -; default f0,SPT.
Quantum Efficiency η: specifies how many primary (not considering the
avalanche gain) electron-hole pairs are generated per each incident photon,
on average. Allowed values: min 0; max 1; default 0.7.
Responsivity (at simulation center frequency) R [A/W]: the slope of
the current vs. input optical power. The responsivity is related to quantum
efficiency by (2) therefore setting one also the other is set. Allowed values:
min 0; max 1.25; default 0.875.
Single-Pole Electrical Filtering: enables the single-pole filtering. If
disabled, the device bandwidth is unlimited.
-3dB Bandwidth BW [GHz]: the frequency of the single-pole of the
photodiode transfer function; it is enabled only if the previous option is
enabled. Allowed values: min 0; BWSPT/2, default 2Rb (reference bit rate).
Photodetection noise (including Avalanche Gain Noise): enables the
generation of noise. If disabled, the device is completely noiseless.
Total Dark Current (including Avalanche Gain) GId [nA]: the average
value of the current generated by the reverse bias voltage applied to the
photodiode, subjected to avalanche multiplication. Allowed values: min 0;
max -; default 2.5.
Avalanche Gain G [-]: the linear avalanche multiplication gain. Allowed
values: min 1; max -; default 25.
Noise Figure FdB [dB]: determines the amount of noise enhancement due to
the avalanche multiplication process. It is related to the Excess Noise
Factor x by FdB=x10log(G).Therefore, setting one also sets the other .
Allowed values: min 1; max -; default 10.
Excess Noise Factor x [-]: it is an alternative way of expressing noise
enhancement due to the avalanche multiplication process. It is related to the
11.2 References
[1] L. G. Kazovsky, S. Benedetto, A. Willner, Optical Fiber Communication
Systems. Boston : Artech House, copyr. 1996.
Variable Optical
Attenuator BER
TX RX
Tester
Pin BER
BER
Reference
BER
Receiver Pin
Sensitivity
Sensitivity depends on the transmitted pulse shape, the noise power and the
post-detection filter. A set of “typical” options is shown in the parameter
dialog window for both of these aspects. The pulses can be selected to be
NRZ rectangular, RZ rectangular and RZ Gaussian. The post-detection filter
can be either matched to the input pulse (for the definition of a “matched
filter” see [1] p.79, or [2] p.236; a matched filter is optimum with respect to
the receiver additive electrical noise if such noise is Gaussian and its power
spectral density is white), Bessel type up to order 8 or a two real poles type.
Two output electrical signals are available: filtered according to the chosen
filter, or unfiltered. The unfiltered output is useful when one need to simulate
a receiver with a post-detection filter whose type is not the same as the one
at which the sensitivity was measured.
These options encompass the typical test conditions that are specified in the
receiver datasheets. Therefore, at least to a good approximation, you should
Q2 w Q σ
PR = hνG x 20 + hν th (1)
2η r0 Gη qr0
where:
r0 is derived from r(t), sampled at the time-instant t0 that maximizes it: r0
= r(t0) = maxt{r(t)}. In turn, r(t) = p(t)*hL(t), where hL(t) is the
electrical impulse response of the receiver output to the photodiode
photocurrent, and p(t) is the normalized optical power pulse shape. The
p(t ) = P(t ) (2 PR ) , where P(t) is
‘*’ means “convolution”. By definition:
the unnormalized pulse shape in [W] and PR is the average optical power
at the receiver input in [W].
w0 is derived from w(t), sampled at the time-instant t0 that maximizes
r(t). By definition: w(t ) = p(t ) * hL2 (t )
η is the photodiode quantum efficiency (without avalanche gain).
G is the photodiode avalanche gain.
x is photodiode excess noise factor. Note that an equivalent notation is
also in widespread use: Gx = F where F is the avalanche photodetector
noise figure.
σth is the standard deviation of the receiver noise (photodetection shot
noise and avalanche excess noise are not included) due to the electronics,
at the receiver output. It is very common to characterize the receiver
noise by means of an equivalent current noise source added to the
detector photocurrent. If such source has a bilateral Power Spectral
∞
σ th = ∫ Gth ( f ) H L ( f ) df
2
Density (PSD) Gth(f), then where HL(f) is the
−∞
Fourier Transform of hL(t).
The above expression for the Q-factor is exact, under the above assumptions,
and correctly takes into account the contribution of shot noise and avalanche
excess noise. It also accounts exactly for the non-stationary nature of shot
noise, through Campbell's theorem [3], p.123.
The actual error probability is related to the Q-factor through the approximate
relationship:
1 Q
BER ≈ erfc (2)
2 2
This expression is accurate when:
1. No pattern noise is considered.
2. Receiver sensitivity is not very close to the quantum limit where a true
Poisson-quantum detection statistic is valid [9 photons/bit].
3. Receiver electrical noise is Gaussian.
4. Intersymbol interference is negligible.
Typically, even the best receivers with APDs are at least 10-15 dB away from
the quantum limit, so that the expression above can be considered to be very
accurate. In any case, it must be kept in mind that this component is based
on the Q-factor theory, and that the reference BER that you are asked to
specify is simply used to derive a value for the Q-factor, via the expression
above.
Receiver sensibility is measured with respect to a particular BER (say 10-9)
which is achieved by reduced (attenuated) received power at the input of the
receiver.
Based on the receiver parameters, the component computes the PSD of the
input equivalent current noise source necessary to match the specified
(possibly fictitious) BER, or a corresponding Q-factor. At runtime, a white
noise source, followed by a noise shaping linear filter, generates the required
noise PSD. Another white noise source generates shot noise according to the
instantaneous value of the input optical signal.
Note
The maximum Q value evaluated is 30 dB. When greater values are obtained
they are truncated to 30 dB and a warning message is given. The minimum
BER value estimated is 10-40. When smaller values are obtained they are
truncated to 10-40 and a warning message is given.
Shot noise can be turned off. If so, in previous equation of PR the first term
on the right-hand side is neglected and at runtime shot noise is not
generated.
SPT Behavior
SPT simulations manage optical signals only; thus only the optical input and
the optical output of this component are taken into account in simulation
based on the SPT. In particular, the optical input spectrum is filtered using
the appropriate filter transfer function chosen (see the Receiver Filters
section).
Inputs Outputs
Note
A warning is issued if the sensitivity is close to the quantum limit. If a
sensitivity lower than the quantum limit is nevertheless chosen, and the shot
noise is not turned off (see the Advanced section), then the current sensitivity
of the receiver will be set by the shot noise and will result a larger optical
power value (than the supplied value) to reach the target BER.
Second Pole Frequency [fraction of bit rate] [-]: sets the second pole
frequency of the filter, as fraction of bit rate; it is enabled only if Two Real
Poles is chosen. Allowed values: min >0.5; max -; default: 2.
q
R =η
hν
therefore setting one also the other is set. Min and max values are checked
by first computing the corresponding quantum efficiency and then checking
the value against the Quantum efficiency parameter limits. By default it
corresponds to quantum efficiency equal to 0.75.
Photodiode Type: selects the type of photodiode: PIN or APD.
Photodiode avalanche gain [linear]: sets the photodiode avalanche
multiplication gain; it is enabled only if APD is chosen. Allowed values: min
>1; max -; default: 25.
Quantum Photodetection noise: enables or disables the generation of
quantum noise in the photodiode.
Noise Figure FdB [dB]: sets the amount of noise enhancement due to the
avalanche multiplication process; it is enabled only if APD is chosen and the
noise generation is enabled. It is related to the Excess Noise Factor x by
FdB=x10log(G) therefore setting one also the other is set. Allowed values: min
0; max -; default: 10.
Excess Noise Factor x [-]: it is the alternative way of expressing noise
enhancement due to the avalanche multiplication process. It is related to the
Noise Figure FdB by x=FdB/10log(G) therefore setting one also the other is
set. Min and max values are checked by first computing the corresponding
noise figure and then checking the value against the Noise figure parameter
limits. By default it corresponds to a noise figure equal to 10.
Evaluate BER vs. Received Optical Power [-]: enables or disables the
generation of the BER vs. received optical power chart. This option is enabled
only if Electrical Noise Limited is enabled.
Evaluate Required Optical Power at Sensitivity [-]: enables or disables
the evaluation of the required optical power to obtain the reference sensitivity
BER. This option is enabled only if Electrical Noise Limited is enabled.
ASE Noise Limited: enables or disables the ASE Noise Limited Semi-Analytic
BER evaluation technique.
Modulation Type: sets the modulation used IM-DD or POLSK. This option is
enabled only if ASE Noise Limited is enabled.
Include Receiver Noise: includes the receiver noise in BER evaluation. This
option is enabled only if ASE Noise Limited is enabled.
Evaluate Ber vs. Decision Threshold: enables or disables the generation
of the chart of BER vs. Decision Threshold. This option is enabled only if ASE
Noise Limited is enabled.
Measured time-span: it selects the time window where the measurement
will be performed. With the Whole option the whole Tsim will be considered.
With the Limited option only the time span set with the following parameters
will be considered.
Measure Start Time [ns]: the starting time of the measurement window.
Allowed values: min 0; max Tsim (duration of the time-span); default 0.
Measure End Time [ns]: the ending time of the measurement window.
Allowed values: min 0; max Tsim; default Tsim.
Note
A warning is issued if the ASE Noise Limited evaluation technique is
enabled, and no optical filter has been selected.
1 N
1 N
1 Vi − Vth
P ( e) =
N
∑ P(e | V ) = N ∑ 2 erfc
i
2
i =1 i =1
2σ
where:
Vi are the received signal samples at the sampling instant
Vth is the decision threshold
σ2 is the variance of the noise at the sampling instant
For a more general introduction to BER estimation through semi-analytic
techniques in communication systems, see for example [5].
Whenever the above assumption are met, semi-analytic techniques give a
much more precise estimate of the system BER than standard Monte Carlo
simulation of the Q-factor, particularly when system is strongly effected by
inter-symbol interference. This is due to the fact that noise is not actually
generated in the time domain, but it is analytically taken into account through
its PDF at the receiver.
This component implements this technique by:
estimating the noise variance by taking into account all the parameters of
the Basic Attributes and Advanced sections
finding the optimum sampling instant
finding the optimum decision threshold
The generated outputs are:
the evaluated BER at the received input optical power, and the
equivalent Q, obtained inverting the formula:
1 Q
BER = erfc
2 2
a plot of the BER versus the received power; the component evaluates
the BER as though an attenuator is placed in front of the photodiode and
its attenuation value is varied as shown in the following graph
Variable Optical
Attenuator BER
TX RX
Tester
Pin BER
Figure 12.2 Schematic for the RX-Sensitivity component for the semi-
analytical BER evaluation
A schematic diagram of the receiver is shown in Figure 12.3: the input signal
e(t)+n(t) is sent to the cascade of an optical filter with impulse response
ho(t), a photodiode and an electrical filter he(t).
RECEIVER
n(t)
Figure 12.3 Schematic for the RX-Sensitivity component for the ASE noise
limited semi-analytical BER evaluation
1 Q
BER = erfc
2 2
a plot of the BER versus decision threshold
SPT Behavior
This component converts the optical signal into an electrical signal. SPT
simulations manage optical signals only, thus this component is not taken
into account in SPT simulations.
Inputs Outputs
Optical Electrical
Parameters
Name: the optional name that appears in the component label.
File of Parameters: selects the parameters description file: Standard 2.5
Gbit/s, Standard 5 Gbit/s, Standard 10 Gbit/s and Custom (your own
parameters file).
File name contains all parameters used to demodulate the signal.
Custom File: name of the data file where the component parameters are
stored. You can choose the file from the file system clicking on the Browse
button. You can also edit the file clicking on the Edit button: it opens an
editor window loading the data file. The editor is VI for Unix or Notepad for
Windows NT.
The parameters contained in the component description file are:
Bit Rate [Gbit/s]: POLSK signal bit rate
Photodiode Bandwidth: it is the 3 dB cut-off bandwidth of the
photodiode internally used to detect the optical signal
T&H Bandwidth: it is the 3 dB cut-off bandwidth of the Track & Hold
circuit. For more information, please see Ref. 5 of the POLSK Project
Bibliography at the end of this Chapter
Sampling Flag: allows to choose between a random sampling for the
T&H circuit (Sampling Flag = 0) or a deterministic sampling (Sampling
Flag = 1). Here “deterministic” means that the sampling instant has a
fixed offset with respect to the data clock, while “random” means that the
sampling instant offset is randomly chosen.
Note
If the bandwidth of Track & Hold circuit is set to zero, no T&H filter is used.
Here is how the file describing the behavior of this component must look like:
# BITRATE [GHz]
5.
# PHOTODIODE BANDWIDTH [GHz] (>0.)
10.
# TRACK & HOLD CIRCUIT BANDWIDTH [GHz] (>=0.) (IF =0. NO FILTERING)
4.
# SAMPLING FLAG: 0=RANDOM 1=DETERMINISTIC
0
# SAMPLING DELAY 0. <= RDELAY < 1.
0.
# SUBSAMPLING FACTOR USED TO RECOVER REFERENCE VECTOR (>=1)
2
# SHIFT REGISTER LENGHT (>=1)
20
Note
The actual value of the optical delay used in the simulation is an integer
multiple of the simulation time sampling step, so the delay effectively used τ'
is related with the simulation bandwidth through the relationship:
τ ⋅ BWVBS 0.8
τ ′ = Nearest _ Integer
0.8 BWVBS
where BWVBS is the VBS simulation bandwidth and τ is the insert delay value.
The required delay τ can be forced to be equal to the actual used value τ' by
clicking the Force Bit Rate button in the Time Domain section of the
Simulation Parameters dialog window. This uses as reference bit-rate the
bandwidth BW (equal to 1/τ) of the Mach-Zehnder.
Inputs Outputs
Optical Electrical
Parameters
Tuning_Frequency [THz]: the frequency of the optical channel to detect.
Allowed values: min flowSPT; max fupSPT;default 194;
BitRate [Tbit/s]: it is the bit-rate of the optical channel to detect. Allowed
values: min -; max -; default 0.04
This component is useful when one wants to simulate a simple receiver whose
only relevant characteristics are the optical and electrical filter bandwidths,
while the details of the receiver in terms of noise level are irrelevant.
SPT Behavior
This compound component converts the optical signal into an electrical signal.
SPT simulations manage optical signals only, thus this component is not
taken into account in SPT simulations.
Inputs Outputs
Optical Electrical
Parameters
Electrical_filter_3dB_bandwidth [GHz]: the 3-dB bandwidth of the 5
poles electrical Bessel filter. Allowed values: min 0.; max -; default 8.
Optical_filter_bandwidth [GHz]: the bandwidth of the raised cosine
optical filter. Allowed values: min 0.; max -; default 40.
Optical_filter_center_freqency [GHz]: the center frequency of band-pass
in the optical filter. Allowed values: min flowSPT; max fupSPT; default 40.
Optical_filter_roloff [-]: the roll-off of the raised cosine optical filter.
Allowed values: min 0. max 1. default 0.2.
Figure 12.6 The model schematic for the N=4 channel receiver
This component is useful when one wants to simulate a simple WDM receiver
whose only relevant characteristics are the optical and electrical filter
bandwidths, while the details of the receiver in terms of noise level are
irrelevant.
SPT Behavior
This compound component converts the optical signal into an electric signal.
SPT simulations manage optical signals only, thus this component is not
taken into account in SPT simulations.
Inputs Outputs
Optical Electrical
Please note that the number of electrical outputs depends on the chosen
Compound Component, thus assuming the values 4, 8 or 16.
Parameters
Center_freqency [THz]: the center frequency of the WDM frequency comb.
It is the middle frequency between channel 1 and channel N. Allowed values:
min flowSPT; max fupSPT; default 193.1.
Channel_spacing [THz]: the spacing between WDM channels. Allowed
values: min 0.; max BWSPT; default 0.1.
Electrical_filter_3dB_bandwidth [GHz]: the 3dB bandwidth of the 5 poles
electrical Bessel filter. Allowed values: min 0.; max -; default 8.
Optical_filter_bandwidth [GHz]: the bandwidth of the raised cosine
optical filter. Allowed values: min 0.; max -; default 40.
Optical_filter_roloff [-]: the roll-off of the raised cosine optical filter.
Allowed values: min 0.; max 1.; default 0.2.
SPT Behavior
This compound component converts the optical signal into an electrical signal.
SPT simulations manage optical signals only, thus this component is not
taken into account in SPT simulations.
Inputs Outputs
Parameters
Bitrate [Gbit/s]: it is the bit-rate of the optical channel to detect for the
DPSK_REC component. Allowed values: min -; max -; default 0.04.
Channel_frequency [THz]: the frequency of the optical channel to detect.
Allowed values: min flowSPT; max fupSPT;default 194.
Filter_bw [GHz]: the absolute bandwidth Bw of the raised cosine optical
filter. Allowed values: min 0; max BWVBS/4; default Rb.
Roll_off [-]:the raised cosine optical filter roll-off Ro. Allowed values: min 0;
max 1; default 0.5.
SPT Behavior
This compound component converts the optical signal into an electrical signal.
SPT simulations manage optical signals only, thus this component is not
taken into account in SPT simulations.
Inputs Outputs
Parameters
Bitrate [Gbit/s]: the bit rate at which the data sources operates. Allowed
values: min -; max -; defaut Rb.
Center_frequency [THz]: the center frequency of the WDM aggregate.
Allowed values: min flowSPT; max fupSPT; default f0,SPT.
Channel_spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
Filter_bw [GHz]: the absolute bandwidth Bw of the raised cosine optical
filters. Allowed values: min 0; max BWVBS/4; default Rb.
Roll_off [-]:the roll-off Ro for the raised cosine optical filters. Allowed
values: min 0; max 1; default 0.5.
12.8 References
[1] S. Benedetto, E. Biglieri, V. Castellani, Digital Transmission Theory,
Prentice-Hall, Englewood Cliffs, NJ, 1987.
[2] J. G. Proakis, M. Salehi, Communication Systems Engineering, Prentice-
Hall, Englewood Cliffs, NJ, 1993.
[3] L. G. Kazovsky, S. Benedetto, A. Willner, Optical Fiber Communication
Systems, Artech House, Boston, 1996.
[4] P. J. Winzer, A. Kalmar, Sensitivity Enhancement of Optical Receivers by
Impulsive Coding, IEEE-OSA Journal of Lightwave Technology, vol.17,
no.2, pp.171-177, February 1999.
[5] M. C. Jeruchim, Techniques for Estimating the Bit Error Rate in the
Simulation of Digital Communication Systems, IEEE Journal of Selected
Areas in Communications, Vol. SAC-2, no.1, January 1984.
[6] G. Bosco and R. Gaudino, Towards new semi-analytical techniques for
BER estimation in optical system simulation, National Fiber Optics
Engineers Conference NFOEC 2000, Denver (CO), Tuesday Sess. E1, USA,
Aug.~2000.
13.1 Driver
This component simulates an electrical driver. It converts the logical input
signal, a binary sequence of zeros and ones into an electrical signal. Several
mapping laws are available, ranging from simple ones, such as NRZ and RZ
rectangular shaped, to NRZ raised cosine, RZ raised cosine, RZ
supergaussian and RZ soliton shaped pulses.
Parameters
Name: the optional name that appears in the component label.
Signal Type and Shape: it opens the related dialog windows for the
available models: NRZ Rectangular, NRZ Raised Cosine, RZ
Rectangular, RZ Raised Cosine, RZ SuperGaussian and RZ Soliton.
Low Level [-]: it is the value of the output signal when a logical zero is
transmitted. Allowed values: min -; max -; default 0.
High Level [-]:it is the value of the output signal when a logical one is
transmitted. Allowed values: min -; max -; default +5.
Logical Electrical
Parameters
Time slope [-]: it is the value of the signal time slope, as a fraction of bit
duration, defined as time between 10% to 90% of the rise signal amplitude,
when a transition between “0” to “1” occurs. Allowed values: min 0; max
0.8; default 0.
Figure 13.1 NRZ raised cosine driver: the output signal waveform.
Inputs Outputs
Logical Electrical
Parameters
Resulting Raised Cosine Shape through a linear modulator is in: it
allows to selects what the driver will shape if connected to an amplitude
optical modulator with a linear transfer function. Choice is between
Amplitude or Power, so that the raised cosine waveform will be applied to
the amplitude or the power of the optical field.
Time slope [-]: it is the value of the signal time slope, as a fraction of bit
duration, defined as time between 10% to 90% of the rise (fall) of signal
amplitude, when a transition between “0” to “1”occurs. It is related to roll-off
α through the formula:
α
Tslope = (arcsin(0.8) − arcsin(−0,8) )
π
Allowed values: min 0; max corresponding to α = 1; default: corresponding
to α =0.8
Roll-off [-]: it is the roll-off parameter of the raised cosine pulse and fixes
the raise (and fall) time during switching. It is defined as the ratio between
the complete 0%-100% raise (fall) time and the bit time. Allowed values: min
0.; max 1; default 0.8.
13.1.3 RZ Rectangular
This component simulates the RZ rectangular driver. It has an output signal
that can assume two electrical levels. When a "1" is transmitted, the output
signal is at the High Level for a time equal to the product of the duty cycle
by the bit time. Then it goes down to the Low Level for the remaining time.
When a "0" is transmitted, the output is constant at the low level for the
entire bit time. Switching between the two levels is instantaneous with
resulting square edges.
Inputs Outputs
Logical Electrical
Parameters
Duty Cycle [-]: it is the ratio between the time at High Level for the first
part of the bit, when a "1" is transmitted, and the bit time. Allowed values:
min 0; max 1; default 0.5.
V (T ) = ( Ah − Al ) f (T ) + Al
if a raised cosine pulse in power is chosen and
V (T ) = ( Ah − Al ) f 2 (T ) + Al
if a raised cosine pulse in amplitude is chosen.
Ah is the high level, Al is the low level and f(t) is the standard raised cosine
function
TL (1 + α )
0 , T >
2
1 T − TL TL (1 − α ) TL (1 + α )
f (T ) = 1 − sin π ,
≤ T ≤
2 αTL 2 2
T (1 − α )
1 , T < L
2
where
Tb
TL =
1+α
α is the roll-off parameter and Tb is the bit time.
Inputs Outputs
Logical Electrical
Parameters
Resulting Raised Cosine Shape through a linear modulator is in: it
allows to selects what the driver will shape if connected to an amplitude
optical modulator with a linear transfer function. Choice is between
Amplitude or Power, so that the raised cosine waveform will be applied to
the amplitude or the power of the optical field.
Duty Cycle [-]: it is the ratio between the time at High Level for the first
part of the bit, when a "1" is transmitted, and the bit time. Allowed values:
min 0; max 1; default 0.5.
Roll-off [-]: it is the roll-off parameter of the pulse (see expression above).
Allowed values: min 0.; max 1; default 0.5.
13.1.5 RZ Supergaussian
This component simulates the RZ supergaussian driver. When “1” is
transmitted it generates a pulse with supergaussian shape. When a "0" is
transmitted, the output signal is set to the Low Level for the entire bit time.
Considering a relative time origin in the middle of the bit time (where the
pulse peak is centered), when a "1" is transmitted, the electrical output has
the following expression:
2m
1 t
−
V (t ) = ( Ah − Al )
2 T
e 0 + Al
where Ah is the High Level amplitude, Al is the Low Level amplitude and m
is the supergaussian order. T0 is related to the TFWHM (full width half
maximum time) through this function:
TFWHM
T0 = 2m
ln 2
Inputs Outputs
Logical Electrical
Parameters
FWHM resulting through linear modulator [ps]: it is the full width half
maximum time of the optical pulse obtained driving a linear optical amplitude
modulator. Points at half of the High Level power are separated by TFWHM.
Allowed values: min 0; max -; default 30.
Order [-]: it is the order of the supergaussian function. Allowed values:
min 1; max -; default 1.
13.1.6 RZ Soliton
This component simulates the RZ soliton driver. When a "1" is transmitted it
generates a pulse with soliton shape. When a "0" is transmitted the output
signal is set to the Low Level for the entire bit time. Considering a time
origin in the middle of the bit time (where the pulse peak is centered), when
a "1" is transmitted, the electrical output signal has the following expression:
t
V (t ) = ( Ah − Al )sech 2 + Al
T0
where Ah is the high level amplitude, Al is the low level amplitude. T0 is
related to the TFWHM (full width half maximum time) through this function:
TFWHM
T0 =
ln 3 + 2 2
Inputs Outputs
Logical Electrical
Parameters
FWHM resulting through linear modulator [ps]: it is the full width half
maximum time of the optical pulse obtained driving a linear optical amplitude
modulator. Points at half of the peak power are separated by TFWHM. Allowed
values: min 0; max -; default 30.
V2π
±
4
while the second is null; see Polarization Modulator (section 8.5) for further
details. If default values for the polarization modulator are used, the resulting
output optical signals are switched at the bit rate speed between x and y
linear polarization.
PolSK modulation offers several potential advantages with respect to
standard intensity modulation (IM-DD), as shown in details in the papers
reported in “POLSK Project Bibliography” at the end of the chapter.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
13.3 References
PolSK Project Bibliography
S.Benedetto and P.Poggiolini, Theory of polarization shift keying modulation,
IEEE Trans. on Comm., vol. 40, pp. 708-721, Apr. 1992.
S. Benedetto and P. Poggiolini, Multilevel polarization shift keying: optimum
receiver structure and performance evaluation, IEEE Trans. on Comm.,
vol. 42, pp. 1174-1186, Feb./Mar./Apr. 1994.
S. Benedetto, R. Gaudino, P. Poggiolini, Performance of coherent optical
polarization shift keying modulation in the presence of phase noise, IEEE
Trans. on Comm., vol. 43, no. 2/3/4, pp. 1603-1612, Feb./Mar./Apr. 1995.
S. Benedetto, R. Gaudino, P. Poggiolini, Direct detection of optical digital
transmission based on polarization shift keying modulation, IEEE J. of Select.
Areas in Comm., special issue on “High speed point-to-point optical
communications systems”, vol. 13, no. 3, Apr.1995. This paper has received
a nomination for the “Leonard G. Abraham Prize”, 1995 edition, that is
awarded to the best paper published in the IEEE Journal on Selected Areas in
Communications in that year.
S. Benedetto, R. Gaudino, P. Poggiolini, Polarisation recovery in optical
polarisation shift keying systems, IEEE Trans. on Comm., vol. 45, n. 9.,
pp. 1269-1279; Oct. 97.
P. Poggiolini, S. Benedetto, Theory of polarization spreading techniques: part
I, IEEE Transactions on Communications, vol. 42, n. 5, pp. 2105-2118, May
1994. This paper has received a nomination for the “S. O. Rice Prize Paper
Award”, 1994 edition, that is awarded to the best paper published in the IEEE
Transactions on Communications in that year.
S. Benedetto, P. Poggiolini Theory of polarization spreading techniques: part
II, IEEE Transactions on Communications, vol. 42, n. 6, pp. 2291-2304, June
1994.
S. Benedetto, G. Olmo, P. Poggiolini, Trellis coded polarization shift keying
modulation for coherent optical communications, IEEE Transactions on
Communications, vol. 43, n. 2/3/4, pp. 1591-1602, Feb. /Mar./Apr. 1995.
RF Mo-Demodulators
Analog
This category contains several components implementing analog mo-
demodulator. An analog electrical signal is modulated over an electrical
carrier using amplitude, phase or frequency modulation.
Amplitude Modulator (AMPMO2), simulates an analog amplitude modulator.
Digital
This category contains several components implementing digital mo-
demodulators for ASK, PSK QAM and FSK digital modulation formats. The
typical structure of all digital modulators is the following.
RF Mo-Demodulators
Input Output
Logical Signal DIGITAL Modulated Signal
MODULATOR
Parameters
• Central frequency
• Bit rate
• Modulation format
• …
The input logical signals is mapped by the modulator into a set of signals
giving rise to the output modulated electrical signals. Typical component
parameters are the output signal central frequency, the bit or baud rate and
other details required to specify the chosen modulation format. The typical
structure of all digital demodulators is the following.
Input
Carrier-recovery Signal
Input Output
Modulated Signal DIGITAL Demodulated Logical Signal
DEMODULATOR
Input
Clock Signal Parameters
• Central frequency
• Bit rate
• Modulation format
• …
Parameters
Carrier_Frequency [THz]: carrier frequency. Allowed values: see section
1.4 for further information; default 0.002 THz.
Carrier_Amplitude: fixed amplitude of the carrier (zero-to-peak amplitude).
Allowed values: any positive real number; default 1.
Carrier_Phase [rad]: fixed initial phase of the carrier. Allowed values: any
real number; default 0 rad.
Modulator_Sensibility: sensibility of the modulator. It is a scale factor that
multiplies the input modulating signal. Allowed values: any positive real
number; default 1.
Fixed_Mean: fixed mean value to be added to the input signal. Allowed
values: any positive real number; default 1.
Parameters
Carrier_frequency [THz]: frequency of the input signal. Allowed values:
see section 1.4 for further information; default 0.002 THz.
Gain: demodulator signal gain. Allowed values: any positive real number;
default 1.
Parameters
Carrier_Frequency [THz]: carrier frequency. Allowed values: see section
1.4; default 0.002 [THz].
Carrier_Amplitude: fixed amplitude of the carrier (zero-to-peak amplitude).
Allowed values: any positive real number; default 1.
Carrier_Phase [rad]: fixed initial phase of the carrier. Allowed values: any
real number; default 0 rad.
Frequency_Offset [rad]: initial frequency offset of the carrier with respect
to the center frequency. Allowed values: any real number that allows the
resulting starting frequency, given by the sum of the signal frequency plus
the frequency offset, to be inside baseband, see section 1.4 for further
information; default 0 rad.
Mod_sensitivity [rad/(ps*amplitude units)]: sensitivity of the
modulator. It is a scale factor that multiplies the input modulating signal.
Allowed values: any positive real number; default 1.
y (t ) = f ist (t )
Inputs Outputs
Parameters
Signal_Frequency [THz]: carrier frequency. Allowed values: see section 1.4
for further information; default 0.002 THz.
y (t ) = Carrier _ amplitude
⋅ cos(2π ⋅ carrier _ frequency ⋅ t + mod ulator _ sensitivity ⋅ x(t ) + carrier _ phase )
Inputs Outputs
Parameters
Carrier_Frequency [THz]: carrier frequency. Allowed values: see section
1.4 for further information; default 0.002 THz.
Carrier_Amplitude: fixed amplitude of the carrier (zero-to-peak amplitude).
Allowed values: any positive real number; default 1.
Carrier_Phase [rad]: fixed initial phase of the carrier. Allowed values: any
real number; default 0 rad.
Mod_sensitivity: it is a scale factor that multiplies the input modulating
signal. Allowed values: any positive real number; default 1.
Parameters
Signal_Frequency [THz]: carrier frequency. Allowed values: see section 1.4
for further information; default 0.002 THz.
Parameters
Fcentr [THz]: center frequency (section 1.4) of the signals.
where hLOW(t) is an ideal low pass filter that selects the baseband component
of the signal |x(t)|2 and the symbol "*" means convolution.
Inputs Outputs
Parameters
Carrier_Frequency [THz]: carrier frequency. Allowed values: see section
1.4 for further information; default 0.002 THz.
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Output_Amplitude: amplitude of the carrier.
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
Note
The Level of the input logical signal determines the number of levels of the
modulation. It must be not greater than 6 and this constraint is checked.
Lin: input logical signal containing Eoutl: output binary PSK electrical
the information bits signal
Parameters
Output_amplitude: carrier amplitude.
Frequency [THz]: carrier center frequency. See section 1.4 for further
information.
Modulation_index [rad]: modulation index - The normal value that
eliminates the carrier is 1.5707963268.
Baud_rate [Tb/s]: input signal baud rate.
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
Note
The PSK modulation is binary.
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Note
The number of phases of the PSK modulation of the input signal is implicitly
chosen by the user when he chooses the Level of the output logical signal.
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Output_Amplitude: amplitude of the carrier.
Baud_rate [Tb/s]: input signal baud rate.
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
E_clock: clock signal; the period is L_out: logical signal that contains
equal to the symbol period; the the output bits (Level = 2)
sampling instant is defined by XCLK
that is the sampling instant for the
in-phase component. See section
1.6 for further details
E_carrier: reference electrical
carrier used in the demodulation
E_signal: electrical signal to be
demodulated
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Baud_rate [Tb/s]: input signal baud rate.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Output_Amplitude: amplitude of the carrier.
Symbol_period [ps]: duration of the symbol (inverse of the bit rate).
CCITT_flag: integer flag:
if set to 0 the CCITT A scheme is used
if set to 1 the CCITT B scheme is used (an additional phase step is
introduced for each symbol)
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
Note
The Level of the input logical signal determines the number of levels of the
modulation. It must be not greater than 6.
Parameters
Frequency [THz]: carrier center frequency. See section 1.4 for further
information.
Baud_rate [Tb/s]: input signal baud rate.
CCITT_type: flag for selecting the modulation scheme:
1: the CCITT scheme B is used
0: the CCITT scheme A is used
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Parameters
Frequency [THz]: carrier frequency of the generated signal. See section 1.4
for further information.
Output_Amplitude: r.m.s. value of the envelope of the output signal.
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
Note
The dimension of the input logical signal determines the number of levels of
the modulation. It must be even and this constraint is checked.
Parameters
Frequency [THz]: carrier frequency. See section 1.4 for further information.
Decision_threshold: is the demodulator decision threshold Th. It must be
equal to the distance between two adjacent signal of the constellation. In the
case of AWGN linear channel the Th value is given by the following formula:
2A
Th = (N )
2 2 −1
where A is the amplitude of the modulation and N is the dimension of the
output logical signal.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.sono valide solo le potenze di 4
Signal L_out logical level: it is the dimension of the output logical signal,
i.e. the number of bits per symbol. It MUST be even and greater than 2 (4, 6,
8, ...), it fixes the number of points of the M-QAM constellation (M = 2L_out).
Parameters
Baud_rate [Tb/s]: input signal baud rate.
Carrier_Amplitude: amplitude of the generated carrier.
Frequency [THz]: carrier center frequency. See section 1.4 for further
information).
Precoding_option: option to choose precoding of binary symbols:
0 - no precoding
1 - with precoding
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
E_clock: clock signal for decision. E_out: electrical signal at the output
See section 1.6 for further details. of the product detector
E_signal: electrical signal to be L_out: output logical signal
demodulated containing the detected bit (Level =
1)
Parameters
Frequency [THz]: center frequency of the signal to be demodulated. See
section 1.4 for further information.
Baud_rate [Tb/s]: input signal baud rate.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Parameters
Baud_rate [Tb/s]: input signal baud rate.
Output_amplitude: carrier amplitude.
Frequency [THz]: carrier frequency. See section 1.4 for further information.
Tone_spacing [THz]: frequency spacing between adjacent tones.
Signal Lin logical level: dimension of the input logical signal i.e. the
number of bits per symbol.
Note
The dimension of the input logical signal determines the modulation levels
and cannot exceed 4. A check is performed to ensure this constraint.
Parameters
Tone_spacing [THz]: tone spacing between adjacent levels.
Frequency [THz]: carrier frequency. See section 1.4 for further information.
Filter_bandwidth [THz]: 3-dB bandwidth of the bandpass filters.
Filter_number_of_poles: number of poles of the bandpass filters.
Baud_rate [Tb/s]: input signal baud rate.
Delay_sync [ps]: time instant at which the integrate and dump filter is
reset.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Note
The number of levels of the FSK is derived from the Level of the output
logical signal.
Parameters
Threshold_level: threshold level.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
Parameters
Frequency [THz]: carrier center frequency. See section 1.4 for further
information.
Baud_rate [Tb/s]: input signal baud rate.
Correl_option: option to choose error correction - 0 means no correction, 1
means correction.
Signal L_out logical level: dimension of the output logical signal i.e. the
number of bits per symbol.
E_phase
x
FILBUT BBHLIM x
pi/2
FILBUT BBHLIM x
x
E_quad
The Loop filter FILOOP can have one of the following transfer functions (in
terms of Laplace Transforms) :
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D, and G are real constants.
The filters labeled FILBUT in the previous graph are lowpass Butterworth
filters with user specified order and bandwidth.
Inputs Outputs
Parameters
Integration_Order: order of integration (0, 1 or 2) for the loop filter,
integer variable. Allowed values: min 0; max 2; default 1.
Loop_Filter_Par_A: loop filter constant (A in the formulae). Allowed values:
any real number; default 1.
Loop_Filter_Par_B: loop filter constant (B in the formula (3), ignored in
other cases). Allowed values: any real number; default 1.
Loop_Filter_Par_D: loop filter constant (D in form. (1) and (2), ignored in
(3)). Allowed values: any positive real number; default 1.
Loop_Filter_Gain: loop filter gain (G in the formulae). Allowed values: any
positive real number; default 1.
VCO_Initial_frequency [THz]: center frequency of the VCO; it should be
equal (or as close as possible) to the input signal central frequency. Allowed
values: see section 1.4 for further information; default 0.002 THz.
Amp: amplitude of the VCO output signal. Allowed values: any real number;
default 1.
VCO_sens: VCO gain, real variable. Allowed values: any real number;
default 1.
Butterworth_Filter_Npoles: number of poles of the lowpass Butterworth
Arm Filters. Integer variable. Allowed values: integer number greater or equal
to 1; default 1.
Butterworth_Filter_Band: 3 dB bandwidth of the lowpass Butterworth Arm
Filters. Allowed values: any real number greater than zero, and giving a filter
bandwidth inside the simulation bandwidth, see section 1.4 for further
information; default -.
Electrical Electrical
Parameters
Name: the optional name that appears in the component label.
Gain [dB] or [lin]: it is the ideal electrical gain. The linear gain has to be
interpreted as an amplitude gain. It means: GaindB = 20 log10 (Gainlin )
Allowed values: min 0 (lin); max -; default 1 (lin).
Note
If a loss has to be inserted, please, specify a negative value in dB or in the
range (0,+1) in the linear scale.
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog window for the two available models, 50-ohm
and User-defined.
15.2.1 50 Ohm
This component has been included in the OptSim library to allow the
simulation of a 50 Ω amplifier. Specifically, it has been thought as an optical
receiver front-end, directly connected to a photodetector.
The implementation of this component assumes that the input signal is a
current expressed in Ampere [A]. If the input signal has a different nature,
the noise is incorrectly generated. Therefore, in this case, the user must
apply the proper conversion by multiplying the input signal by a suitable unit
conversion constant.
The output of the OptSim photodiode component can be directly connected to
this component input without any normalization, since both the photodiode
output and amplifier input has to be intended as currents expressed in
Ampere [A].
The considered model of the 50 Ω amplifier does not consider frequency
dependence of the gain and of noise. Hence, gain and noise power spectral
density are considered flat over the whole simulation bandwidth.
The implemented model of this component is based on the assumption that
the 50 Ω matching is satisfied. It means that it must be possible to represent
the input signal through an equivalent Norton circuit (ideal current generator
plus a parallel resistor) with a 50 Ω resistor. Since the modeled device has a
50 Ω input resistance, the input signal propagates through two 50 Ω
resistances placed in parallel. Also the output of the device is supposed to be
50 Ω matched, therefore a final terminating 50 Ω resistance is assumed to be
present for the sake of calculating the output current Iout The described
schematic of the implemented model can be observed below. Iin is the input
current to the component. Iout is the output current.
G DC
I out = I in
50Ω 50Ω
2
Iin IN
50Ω 50Ω
Note
50 Ω amplifiers generally have worse noise performance than the specifically
designed transimpedance amplifiers. However, as the bit-rate grows and if an
optical preamplifier is also used, their poorer noise performance becomes less
important, and their use may be viable even in high-performance systems. In
other situations, the 50 Ω solution can be the one of choice due to the
extremely low cost of some of these devices.
GDC
I out = I in (1)
4
Noise is modeled as an equivalent current source IN . According to the
definition of noise figure F [2], all terminals are assumed 50 ohm matched, as
in the above schematic. Then the double-sided power spectral density (PSD)
of IN is:
2 K BTGDC
F
50
where T is the absolute temperature expressed in Kelvin and KB the
Boltzmann's constant.
Even if the component has been modeled to simulate a 50 Ω matched
amplifier, i.e., a layout including a 50 Ω resistance placed between the
current generator component and the amplifier input, it is nevertheless
possible to simulate a layout where the matching input resistance is missing.
To get this result the user has to consider the nominal DC gain and noise
figure of the 50 Ω device, and then she/he has to set the component
parameters in the following way.
6 dB have to be added to the nominal DC gain, therefore the gain value
to be introduced in the dialog window is Gblock=GDC+6, where GDC is the
50 Ω amplifier nominal gain and Gblock is the actual value, expressed in
dB, to be introduced in the window.
The actual noise figure to be introduced must to be derived from the
nominal value employing the following formula: Fblock=10log10[1+(F-
2)/4]. Where F is the nominal noise figure of the amplifier, expressed in
linear unit, and Fblock is the actual value, in dB, to be introduced in the
dialog window.
Inputs Outputs
Electrical: the input current [A] Electrical: the output current [A]
Parameters
DC Gain [dB]: the gain of the amplifier. Allowed values: min 0; max -;
default 0.
Noise Figure F [dB]: the noise figure of the amplifier [2]. Allowed values:
min 0; max -; default 3.
Reference Temperature [K]: the reference temperature of the amplifier
[2]. Allowed values: min 0; max -; default 300.
15.2.2 User-Defined
This model implements a noisy amplifier whose transfer function and noise
characteristics are described by the user through ASCII files. This component
has been introduced to be flexible and to allow modeling of any linear
receiver structure, independently of the underlying hardware implementation.
The noise added by the receiver is assumed to be a Gaussian random
process.
The user needs to supply an ASCII file containing the samples of the transfer
function vs. the frequency, and an ASCII file containing the samples of either
the output noise spectral density, or the amplifier noise figure as a function of
the frequency.
Several amplifiers can be cascaded in an OptSim project, but no reflections
are taken into account. It means that a perfect matching between the
cascaded components is always supposed.
The implementation of this component is largely based on the “User-defined
Electrical Filter” (see section 16.1.8).
Inputs Outputs
Electrical Electrical
Modeling Principles
The user provides the amplifier transfer function HL(f). It is implemented
through a filter whose impulse response is hL(t)=IFT{HL(f)}, where IFT is the
Inverse Fourier Transform. At runtime, the input electrical signal sin(t) is
therefore filtered accordingly to:
where s'out(t) is the output signal before the noise addition, and ‘✳’ means
convolution.
The overall output electrical signal sout(t) is the sum of the filtered input
s'out(t) and the noise at the output nout(t):
The output noise nout(t) is a Gaussian and zero-mean random process. Its
power spectral density matches with the output noise bilateral power
spectral density Gn(f) defined through its shape or through the noise figure
dependency on frequency. This is done by synthesizing a linear filter that
shapes a white noise.
Note
The output noise power spectral density Gn(f) is always considered as a
bilateral distribution, i.e., positive and negative frequencies are considered.
It means that the output noise variance is equal to
+∞
−∞
∫ G ( f )df
n
Since electrical signals in OptSim have no declared measure units, the user
has to take care of the consistency of the supplied data. In particular, the
measure unit of noise n(t) and signal s'out(t) must be the same. It means,
that if the output of the amplifier is assumed to be a current expressed in
Ampere [A], n(t) must be a current. Therefore, its power spectral density
Gn(f) must be defined as current squared per Hz [A2/Hz]. If the output of the
amplifier is assumed to be a voltage expressed in Volts [V], n(t) has to be a
voltage, and its power spectral density Gn(f) must be measured in Volts
squared per Hz [V2/Hz].
Alternative to providing the power spectral density Gn(f), the user can supply
samples of the noise figure [2] as a function of the frequency. If this option is
selected, the user is requested to furnish the value of the reference
resistance and of the reference temperature of the device. Moreover, the user
has to define if the noise generated by the reference resistance is a current or
a voltage. Such an explicit noise dimension declaration, is needed because of
the formula giving the noise power spectral density from the noise figure F(f):
Gn ( f ) = H L ( f ) ⋅ F ( f ) ⋅ 2 K t T ⋅ W
2
. (3)
Cascading Amplifiers
Cascading two or more amplifier components poses no special problems. The
overall transfer function is the product of the individual transfer functions and
each amplifier introduces the defined amount of noise.
Note that the model does not take into account either impedance
mismatch or reflections. Therefore, if the user needs to model an amplifier
chain where inter-stage reflections occur, the user has to employ one
amplifier only whose transfer function is the overall chain transfer function
and the total output noise power spectral density is the equivalent power
spectral density of the chain. Note that when cascading components, with no
other components in between, the dimensions of the input and output must
obviously the same. Then noise power spectral density have to be supplied in
[A2/Hz] if the input signals are currents, or in [V2/Hz] if the input signals are
voltages.
If the noise figure is used, then you need to indicate explicitly whether the
noise generated by the reference resistance is a voltage or a current.
Dimensions must match those of the signals.
Note
A word of caution is necessary regarding the reference resistance supplied
with noise figure samples. Such a resistance is used only for the
generation of the correct amount of noise, as indicated in (3). It is not
used to compute impedance mismatch or reflections between cascaded.
Special Features
Transfer function and noise data can be smoothed, and the user can define a
smoothing factor to control the amount of smoothing to be applied. This
feature is useful in case data come from measurements affected by
uncertainty and instrument noise.
In addition, the frequency scale of the transfer function and noise data can be
stretched or shrank. This option can be used to reuse data. For instance, the
data of a 2.5 Gbit/s receiver can be stretched to support a 5 Gbit/s
simulation. Any amount of stretching/shrinking can be applied.
For more details please refer to “User-defined Electrical Filter” (section
16.1.8).
Parameters
Receiver Type: selects the model to be simulated between the predefined
Transimpedance 2.5 GHz or Transimpedance 10 GHz, or the completely
user-defined Custom.
Note
Two predefined amplifiers (Transimpedance 2.5 GHz, Transimpedance 10
GHz) are available. In both cases they have been generated from theoretical
models of transimpedance receivers. Both have a single pole, whose
frequency is 2.5 GHz and 10 GHz. The output noise power spectrum density
has been defined to experience the “f2” behavior at high frequency that is
typical of this kind of receivers. These predefined amplifiers are only meant to
serve as examples. They do not refer to specific components. The models for
the generation of these transfer functions have been taken from [1]. Note
that the input signal for these defaults must be a current expressed in
Ampere [A]. The output electrical signal is a voltage, expressed in Volts [V].
The transfer functions and output noise power spectral densities can be
viewed running a simulation with the option Transfer Function and Noise
Spectrum Plot set.
Transfer Function Data Filename: name of the ASCII file where the
transfer function vs. frequency is stored. The file is a three-column file. First
column defines the frequency to which the amplitude (second column) and
the phase/group-delay (third column) refer. The user can choose the file
from the file system clicking on the Browse button. The user can also edit
the file clicking on the Edit button: it opens an editor window loading the
data file. The editor tool is VI for Unix systems or Notepad for Windows NT
systems. The filename extension must be DAT in capital letters.
Transfer Function Data Format: specifies whether the transfer function is
given as a set of samples of the Amplitude Square and Phase or of the
Amplitude Square and Group Delay. See “User-Defined Filter File Format”
in section 16.1.8 for further details.
Amplitude Square: specifies whether the amplitude in the description file is
expressed in dB or in a linear scale.
Amplifier Noise Data Filename: name of the data file where samples of
the output noise bilateral power spectral density, or of the noise figure, vs.
frequency are stored. The user can choose the file from the file system
clicking on the Browse button. You can also edit the file clicking on the Edit
button: it opens an editor window loading the data file. The editor is VI for
Unix or Notepad for Windows NT. The filename extension must be DAT in
capital letters.
The format of this file is the same as for the file describing the transfer
function, except that there is no phase or group delay data to be specified.
Troubleshooting
Error messages can be issued by this component whenever the data supplied,
regarding either the transfer function, or the amplifier noise characteristics, is
ill conditioned and the linear filter synthesis on which the component
implementation is based fails.
For all troubleshooting and the explanation of all error messages, please refer
to “If something goes wrong” in section 16.1.8.
15.3 References
[1] G. Kazovsky, S. Benedetto, A. Willner, Optical Fiber Communication
Systems. Boston Artech House, 1996.
[2] L. W. Couch II, Digital and Analog Communication Systems, Fifth Edition,
Prentice Hall, 1997.
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the filter models, which can be
chosen among Single Pole Lowpass, Two Pole Lowpass, Butterworth,
Bessel, Chebychev, User Defined, Raised-Cosine and Matched Filter.
1 1
H(f )= H( f ) =
2
⇒ (1)
f f
2
1+ j 1 +
f 3dB
f 3dB
where f3dB is the single pole frequency.
The filter is numerically implemented using an IIR (Infinite Impulse
Response) algorithm (see reference [3]) together with the bilinear
transformation method.
Inputs Outputs
Electrical Electrical
Parameters
Pole Frequency [GHz]: the frequency f3dB of the pole. Allowed values: min
0; max BWVBS/2; default 10.
Note
The plot of this filter transfer function is not available in the current version.
Examples
An example of use of a single-pole lowpass filter is shown in the following
figures. The spectrum of the input signal (a white Gaussian noise with
spectral density 1 mW/GHz) and the corresponding filtered output signal (in a
large bandwidth and in a reduced bandwidth) were obtained by setting the
pole frequency to 10 GHz.
Figure 16.1 White noise and the filtered signal (large bandwidth)
Figure 16.2 White noise and the filtered signal (reduced bandwidth)
s1 ⋅ s2
H(f )= (1)
( j 2πf + s1 ) ⋅ ( j 2πf + s2 )
The two poles s1 and s2 are determined by the following two parameters:
The poles “natural frequency” f0
(
s1 = 2πf 0 ξ + ξ 2 − 1 ) s2 =
(ξ +
2πf 0
ξ 2 −1 )
Moreover when ξ < 2 2 , the filter transfer function has an overshoot, i.e. a
resonance, which is usually characterized through the two following
parameters:
f 04
the resonance overshoot, given by 10 log10 dB
0 ( p )
f 2 − f 2 2 + (2 f f ξ )2
0 p
These two values are evaluated after each entry of new parameters and
automatically shown in the lower part of the parameter dialog window in a
read-only box.
The filter is numerically implemented using an IIR (Infinite Impulse
Response) algorithm (see reference [3]) together with the bilinear
transformation method.
Some examples of the transfer functions of two-pole lowpass filters are
shown in the following figure. All graphs were obtained by setting the same
double-pole natural frequency (f0=1) and changing the damping factor ξ.
Figure 16.3 The two-pole lowpass transfer function for different damping
factor ξ and identical natural frequency f0
Inputs Outputs
Electrical Electrical
Parameters
Double-Pole Natural Frequency [GHz]: the natural frequency f0 of the
double-pole. See previous page formulas for further information. Allowed
values: min 0; max BWVBS/2;default 10.
Damping factor [-]: Allowed values: min 0; max -; default 0.5.
Frequency of Resonant Peak [GHz]: (read-only field) the frequency of the
resonant peak fp.
Overshoot [dB]: (read-only field) the filter overshoot, whenever defined.
Note
The plot of the transfer function is not available in the current version.
Examples
The following figures show two examples of two-pole lowpass filters. The
spectrum of the input signal (a white Gaussian noise with spectral density
1 mW/GHz) and the corresponding filtered signal are shown. The double-pole
natural frequency is f0=10 GHz and the damping factor is respectively
ξ=0.04, ξ=0.4 and ξ=4.
A white Gaussian noise filtered by a two-pole lowpass filter, with a damping
factor of ξ=0.04 is shown in the following figure (in a large bandwidth and in
a reduced bandwidth).
Figure 16.4 White noise and the filtered signal (large bandwidth, ξ=0.04)
Figure 16.5 White noise and the filtered signal (reduced bandwidth, ξ=0.04)
Figure 16.6 White noise and the filtered signal (reduced bandwidth, ξ=0.4)
Figure 16.7 White noise and the filtered signal (reduced bandwidth, ξ=4)
16.1.3 Butterworth
This model implements lowpass, highpass and band-pass Butterworth filters.
This is a standard family of filters that are also sometimes called “Maximally
Flat Amplitude” filters. Their characteristics can be found in any book on
analog filters, such as [4].
The filter is numerically implemented using an IIR (Infinite Impulse
Response) algorithm (see reference [3]) together with the bilinear
transformation method.
Inputs Outputs
Electrical Electrical
Parameters
Lowpass, Band-pass, Highpass: selects the type of Butterworth filter.
Center Frequency [GHz]: the center frequency of the band-pass filter. This
field is enabled only if the Band-pass option is selected. Allowed values: min
0; max BWVBS/4; default -.
Number of Poles [-]: the number of current poles for lowpass and highpass
filters. For the band-pass filter, this parameter corresponds to the number of
poles of the lowpass equivalent filter, so that in this case the number of
current poles is twice this parameter. Note that the maximum number of
poles has been set to 10. Allowed values: min 1; max 10; default 1.
-3dB Bandwidth [GHz]: it is the –3dB cutoff frequency for lowpass and
highpass filters, or it is the range between the two cutoff frequencies for
band-pass filter. Allowed values: if lowpass or highpass: min 0;
max BWVBS/4;
Amplitude Plots: enables the plotting of the filter transfer function.
Examples
Two examples of use of Butterworth filters are shown in the following figures.
A lowpass filter and a band-pass filter with 3 poles and a –3 dB bandwidth of
10 GHz are considered. The spectrum of the input signal (a white Gaussian
noise with spectral density 1 mW/GHz) and the corresponding filtered signal
are shown.
The ideal and the currently implemented transfer functions of the lowpass
Butterworth filter are shown in the next figure. This kind of graph is
generated when the request of an amplitude plot is set in the parameter
dialog window.
Ideal
Currently
implemented
Figure 16.8 The ideal and the currently implemented lowpass Butterworth
transfer functions.
Figure 16.9 White noise and the filtered signal (large bandwidth)
Figure 16.10 White noise and the filtered signal (reduced bandwidth)
The ideal and the currently implemented transfer functions of the Butterworth
band-pass filter are shown in the following figure. The center frequency is
200 GHz.
Ideal
Currently
implemented
Figure 16.11 The ideal and the currently implemented band-pass Butterworth
transfer functions.
Figure 16.12 White noise and the filtered signal (large bandwidth)
Figure 16.13 White noise and the filtered signal (reduced bandwidth)
16.1.4 Bessel
This model implements lowpass, highpass and band-pass Bessel filters. This
is a standard family of filters that are also sometimes called “Maximally Flat
Delay” filters. Their characteristics can be found in any book on analog filters,
such as [4].
The filter is numerically implemented using an IIR (Infinite Impulse
Response) algorithm (see reference [3]) together with the bilinear
transformation method.
Inputs Outputs
Electrical Electrical
Parameters
Lowpass, Band-pass, Highpass: selects the type of Bessel filter.
Center Frequency [GHz]: the center frequency of the band-pass filter. This
field is enabled only if the Band-pass option is selected. Allowed values: min
0; max BWVBS/4; default -.
Number of Poles [-]: the number of current poles for lowpass and highpass
filters. For the band-pass filter, this parameter corresponds to the number of
poles of the lowpass equivalent filter, so that in this case the number of
current poles is twice this parameter. Note that the maximum number of
poles has been set to 10. Allowed values: min 1; max 10; default 1.
-3dB Bandwidth [GHz]: it is the –3 dB cutoff frequency for lowpass and
highpass filters, while it is the range between the two cutoff frequencies for
band-pass filter. Allowed values: min 0; max BWVBS/4; default Rb (if lowpass
or highpass) or 2·Rb (if band-pass).
Amplitude Plots: enables the plotting of the filter transfer function.
Examples
Two examples of use of Bessel filters are shown in the following figures. A
lowpass filter and a band-pass filter with 3 poles and a –3 dB bandwidth of
10 GHz are considered. The spectrum of the input signal (a white Gaussian
noise with spectral density 1 mW/GHz) and the corresponding filtered signal
are shown.
The ideal and the currently implemented transfer functions of the lowpass
Bessel filter are shown in the next figure. This kind of graph is generated
when the request of an amplitude plot is set in the parameter dialog window.
Ideal
Currently
implemented
Figure 16.14 The ideal and the currently implemented lowpass Bessel transfer
functions.
Figure 16.15 White noise and the filtered signal (large bandwidth)
Figure 16.16 White noise and the filtered signal (reduced bandwidth)
The ideal and the currently implemented transfer functions of the Bessel
band-pass filter are shown in the following figure. The center frequency is
200 GHz.
Ideal
Currently
implemented
Figure 16.17 The ideal and the currently implemented band-pass Bessel
transfer functions.
Figure 16.18 White noise and the filtered signal (large bandwidth)
Figure 16.19 White noise and the filtered signal (reduced bandwidth)
16.1.5 Chebychev
This model implements lowpass, highpass and band-pass Chebychev filters.
This is a standard family of filters whose characteristics can be found in any
book on analog filters, such as [4].
The filter is numerically implemented using an IIR (Infinite Impulse
Response) algorithm (see reference [3]) together with the bilinear
transformation method.
Inputs Outputs
Electrical Electrical
Parameters
Lowpass, Band-pass, Highpass: selects the type of Chebychev filter.
Center Frequency [GHz]: the center frequency of the band-pass filter. This
field is enabled only if the Band-pass option is selected. Allowed values: min
0; max BWVBS/4; default -.
Number of Poles [-]: the number of current poles for lowpass and highpass
filters. For the band-pass filter, this parameter corresponds to the number of
poles of the lowpass equivalent filter, so that in this case the number of
current poles is twice this parameter. Note that the maximum number of
poles has been set to 10. Allowed values: min 1; max 10; default 1.
Bandwidth at Cutoff [GHz]: it is the –3 dB cutoff frequency for lowpass
and highpass filters, while it is the range between the two cutoff frequencies
for band-pass filter. Allowed values: min 0; max BWVBS/4; default Rb (if
lowpass or highpass) or 2·Rb (if band-pass).
Ripple [dB]: band-pass ripple. Allowed values: min 0; max -; default 0.3.
Amplitude Plots: enables the plotting of the filter transfer function.
Examples
Two examples of use of Chebychev filters are shown in the following figures.
A lowpass filter and a band-pass filter with 3 poles, a –3 dB bandwidth of
10 GHz and a ripple of 0.3 dB are considered. The spectrum of the input
signal (a white Gaussian noise with spectral density 1 mW/GHz) and the
corresponding filtered signal are shown.
The ideal and the currently implemented transfer functions of the lowpass
Chebychev filter are shown in the next figure. This kind of graph is generated
when the request of an amplitude plot is set in the parameter dialog window.
Ideal
Currently
implemented
Figure 16.20 The ideal and the currently implemented lowpass Chebychev
transfer functions.
Figure 16.21 White noise and the filtered signal (large bandwidth)
Figure 16.22 White noise and the filtered signal (reduced bandwidth)
The ideal and the currently implemented transfer functions of the band-pass
Chebychev filter are shown in the next figure. The center frequency is
200 GHz.
Ideal
Currently
implemented
Figure 16.23 The ideal and the currently implemented band-pass Chebychev
transfer functions.
Figure 16.24 White noise and the filtered signal (large bandwidth)
Figure 16.25 White noise and the filtered signal (reduced bandwidth)
16.1.6 Matched
This model implements a baseband matched electrical filter. Given a signal
s(t), its matched filter has an impulse response function given by h(t)=s*(t0-
t). It is known from the general theory of digital communications [1] that,
when the signal s(t) is received together with white Gaussian noise, the
matched filter gives at its output the maximum signal-to-noise ratio among
all possible filters. This corresponds to the minimum bit error rate at its
output.
OptSim implements a matched filter for rectangular and half-sinusoidal
signals. A half-sinusoidal signal in the time interval [0,T] is defined as:
π
s (t ) = sin t for t ∈ [0, T ]
T
s (t ) = 0 for t ∉ [0, T ]
Inputs Outputs
Electrical Electrical
Parameters
Filter type selector: allows selecting the waveform to be matched between
rectangular or half-sinusoidal.
Impulse response duration [ps]: the impulse response duration of the
matched filter. In a digital transmission environment it is usually given by the
inverse of the system bit rate. Allowed values: min >0; max -; default: 1/Rb.
Simulated impulse response duration [ps]: (read-only field) the actually
simulated impulse response duration of the matched filter. It may be slightly
different from the value inserted in the previous field due to the fact that an
integer number of time samples must fit into the impulse duration. Please see
the OptSim User manual for further information on this issue.
Examples
Two examples of matched filters are shown in the following figures. They
correspond to rectangular and half-sinusoidal waveforms with identical
impulse response duration of 100 ps. The spectrum of the input signal (a
white Gaussian noise with spectral density 1 mW/GHz) and the resulting
filtered signal are shown.
Figure 16.26 White noise and the signal filtered by a Matched rectangular
filter
Figure 16.27 White noise and the signal filtered by a Matched half-sin filter
H(f )= A
[C ( f , Ro )]
α
(1)
R( f )
where:
C(f,Ro) is a raised cosine function, characterized by its roll-off Ro,
described by the following expression (2)
A is an amplitude factor
α is the (real) exponent at which C(f,Ro) is raised
R(f) is a shaping function, that can be either the spectrum of a symmetric
rectangular pulse of duration equal to the inverse symbol rate, or a
constant equal to 1
Bw
1 , f ≤ (1 − Ro )
2
1 π B
C ( f ) = 1 − sin f − w , (1 − Ro ) Bw ≤ f ≤ (1 + Ro )
Bw
2 Ro B w 2 2 2
Bw
0 , f ≥ (1 + Ro )
2
(2)
The parameter Bw is the filter bandwidth.
Raised cosine filters are often used in RF digital modulation (for format like
PSK, FSK and QAM) in order to satisfy the Nyquist criterion that allows
minimizing intersymbol interference. For further information, you can see any
book about digital modulation, such as Ref. [5].
Inputs Outputs
Electrical Electrical
Parameters
Shaping function: the shaping function that determines R(f). It could be the
inverse Fourier transform of a symmetric rectangular pulse of duration equal
to the inverse of the bandwidth Bw specified in (2) (Sinc option) or a constant
equal to 1 (Constant option).
Raised-cosine exponent α [-]: Allowed values: min 0.; max -; default 1.
Raised-cosine roll-off [-]: roll-off parameter Ro. Allowed values: min 0;
max 1; default 0.5.
Bandwidth [GHz]: filter bandwidth Bw. Allowed values: min 0.; max
BWVBS/2; default 10.
Amplitude Plots: enables the plotting of the filter transfer function.
Examples
Examples of Raised-Cosine filters are shown in the following figures. In
particular, Raised Cosine filter with Bw=10 GHz (–3 dB bandwidth) are shown
for different values of the roll-off Ro and with or without the Sinc shaping
factor. The exponent α is set to 1 in all cases, while the input signal is a white
Gaussian noise with power spectral density equal to 1 mW/GHz.
The output of a Raised-Cosine filter with a Sinc shaping factor is shown in the
next figure. The roll-off is set to 0, 0.5 and 1.
Roll-off factor
1.
0.5
0.
Figure 16.28 White noise and the signal filtered by a Raised-Cosine with the
Sinc shape option.
Roll-off factor
1.
0.5
0.
Figure 16.29 White noise and the signal filtered by a Raised-Cosine Constant
without the sinc shaping option
Electrical Electrical
Parameters
Filter Data Filename: name of the data file where the transfer function vs.
frequency is stored. You can choose the file from the file system clicking on
the Browse button. You can also edit the file clicking on the Edit button: it
opens an editor window loading the data file. The editor is VI for Unix or
Notepad for Windows NT. See “User-Defined Filter File Format” in this section
for further details.
File Format Ampl. Square and Phase, Ampl. Square and Group Delay or Ampl.
Square only: specifies if the transfer function in the description file is
expressed as a function of amplitude and phase, amplitude and group delay
or amplitude only (the amplitude should always be expressed as the square
of the modulus of the transfer function).
Amplitude Square: specifies whether the amplitude in the description file is
expressed in dB or in a linear scale.
Smoothing: enables to smoothen the data points supplied in the user data
file. See “Transfer Function Smoothing” in this section for further details.
Smoothing factor [-]: the smoothing factor to be applied when
synthesizing the filter. This field is enabled only if Smoothing is set. Allowed
values: min 1; max -; default 1.
Offset [GHz]: specifies a shift between the current simulated transfer
function and the transfer function stored on the filter description file. It
should be set to 0 (the default) when offset is not required. See “Frequency
Translation” in this section for further details. Allowed values: min -; max -;
default 0.
Scaling [-]: the real number that specifies the expansion/stretch factor in
the frequency scale of the transfer function stored in the data file. This
parameter should be set to 1 (the default) when scaling is not required. See
“Frequency Translation” in this section for further details. Allowed values: min
>0; max -; default 1.
Actual Filter Amplitude and Group Delay Plots: enables the plotting on
screen of the supplied transfer function and group delay data, together with
the transfer function and group delay currently implemented by the
component.
the first line containing the repeated value is considered and a warning
message is shown. The maximum number of allowed points is 2043; if in the
file there are more than this number only the first 2043 points are considered
and a warning message is shown.
Here is an example of input file:
#
# Electrical filter description file
# General comment on the filter:
# - abscissas in GHz
# - modulus in dB
# - phase in radians
##
0.0 0.0 0.0
1.01502502 -0.0445154086 0.10150251
2.02005029 -0.173698097 0.202005035
3.02507544 -0.380280584 0.302507538
4.03010035 -0.653619885 0.40301004
5.03512573 -0.98132962 0.503512573
6.04015064 -1.35079885 0.604015076
7.04517603 -1.75031734 0.704517578
8.05020142 -2.16972327 0.805020142
9.05522728 -2.60064578 0.905522705
10.0602522 -3.03646755 1.00602527
11.0652781 -3.4721365 1.10652783
12.0703039 -3.90391445 1.2070304
13.0753298 -4.32911968 1.30753296
14.0803547 -4.74589777 1.40803552
15.0853806 -5.1530242 1.50853809
...
Please note that, accordingly with the OptSim units for electrical signals, the
units to be used in the description file are:
frequency in [GHz]
phases in [radians]
group delay in [ps]
Frequency Translation
The transfer function specified on the filter description file can be shifted
and/or compressed/expanded on the frequency axis by the use of the two
parameters Offset and Scaling. This feature may prove very useful to define
a given filter on a description file, and then reuse it in a different position on
the frequency axis.
Please refer to the examples of “Frequency Translation” in section 10.1.5 for
further explanation.
An incorrect use of this two parameters may give errors, for example when
the resulting transfer function is defined on a range which does not overlap
with the current simulation bandwidth. See “If something goes wrong” in this
section for further details.
In such situations, the Smoothing parameter may prove useful. When the
feature is set with a Smoothing factor Nfact, the currently simulated transfer
function is obtained by averaging Nfact adjacent points of the original transfer
function.
Please refer to the examples of “Transfer Function Smoothing” in section
10.1.5 for further explanation.
16.2 References
[1] Leon W. Couch II, Digital and analog communication system, Third
Edition, 497-508, 1989.
[2] C. De Boor, A practical guide to Splines, Springer-Verlag, 1978
[3] M. C. Jeruchim, P. Balaban e K. S. Shanmugan, Simulation of
Communication Systems, Plenum Publishing Corporation, New York,
1992.
[4] A. B. Williams and F. J. Taylor, Electronic Filter Design Handbook,
McGraw-Hill, New York, 1988.
[5] J. G. Proakis, Digital Communications, 2nd edition, McGraw-Hill, New York,
1989.
Optical
See the “Optical Network Components” (chapter 18).
Electrical
Two Inputs Multiplexer (MPX).
Logical
Merge Of Two Logical Signals (RMUX1).
Parameters
KSEL: the selection parameter. Allowed values: min 0; max 1; default -.
Parameters
KSEL: the selection parameter. Allowed values: min 0; max 1; default -.
Parameters
Baud rate [Tbaud/s]: it is the baud rate for the output logical signal. Input
baud rate must be half of this.
Delay [ps]: the delay waited before the operation starts.
Note
An error condition will be issued if the levels of the three logical signals are
not equal. No test is performed whether the input baud rate is half of the
output one.
Note
An error condition will be issued if the levels of the three logical signals are
not equal.
Optical Circulator
Optical Demultiplexer
Optical Multiplexer
SPT Behavior
The ideal optical switch is a linear component, therefore, its effects on the
optical spectra are fully simulated during SPT simulations. The used model is
the same employed for VBS simulations.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Connection Vector: the connections between the input and output ports are
specified by a vector of integer values. The position of each vector element
corresponds to a specific connected input port. For instance the 4th element of
the vector corresponds to the 4th input port. The integer value of each vector
element identifies the connection with the output port. For instance if the 4th
element is equal to 6, then the 4th input port is connected to the 6th output
port. A particular case is represented by the 0 value. When the entry in the
Connection Vector is equal to zero, the correspondent input port is not
connected to any output port. Allowed values: min 0; max 64; default -.
INPUT PORT OUTPUT PORT
1 k 1 2
2 0
3 3
4 6
m n m
Examples
4|3|2|1|8|7|6|5
12|0|10|9|0|0|8|5
2|0|0|1
Error cases The configurations of the switch component that can produce an error
message during the simulation are summarized in the following list:
Switch m×k with k>m (the number of connected outputs is higher than
the number of the connected inputs).
Not all the connected output ports are specified in the Connection Vector.
The Connection Vector length is not equal to the maximum input port
number among the connected ones.
SPT Behavior
The optical switch is a linear component, therefore, its effects on the optical
spectra are fully simulated during SPT simulation. The used model is the
same employed for VBS simulation.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Insertion_loss [dB]: the insertion loss of the component. Allowed values:
min 0.; max -; default 0.5.
Crosstalk [dB]: the cross-talk on all signals. Allowed values: min 0.; max -;
default 90.
Out_in1, Out_in2, Out_in3, Out_in4: these parameters express the
connection between the input and the output ports. They are the elements of
the Connection Vector of the Ideal Optical Switch used within the
configuration. Allowed values: min 0.; max 4.
OXC_K×K_Nch
λ1 λ1
K×K
λ1, … λN Optical
λ1, … λN
1
Switch 1
λN λN
λ1 λ1
K×K
K Optical K
λN Switch λN
SPT Behavior
The OXC is a linear component, therefore, its effects on the optical spectra
are fully simulated during SPT simulations. Filtering function at demultiplexer
and multiplexer sections are applied to signal spectra.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Channel_Spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
Center_Frequency [THz]: the center frequency of the WDM signal. It is the
frequency in the middle of the two central channels. Allowed values:
min flowSPT; max fupSPT; default 193.
Filter_Bandwidth [GHz]: the 3dB bandwidth of the demultiplexing raised
cosine filter, refer to section 10.1.2 for better filter parameter definition.
Allowed values: min 0.; max -; default 50.
Filter_Roloff: the raised cosine filter roll-off, refer to section 10.1.2 for
better filter parameter definition. Allowed values: min 0.; max 1.; default 0.2.
Insertion_loss [dB]: the insertion loss of the component. Allowed values:
min 0.; max -; default 3.
Insertion_loss_switch [dB]: the insertion loss of the internal switch
component. Allowed values: min 0.; max -; default 0.5.
Crosstalk [dB]: the cross-talk on all signals. Allowed values: min 0.; max -;
default 90.
Out_chij : this parameter expresses the path of a specific channel at an
input port towards an output port. The Out_chij value is the output port
number of the ith channel at the jth input port. The generic OXC_K×K_Nch
module has N×K Out_chij parameters. They are the elements of the
Connection Vectors of the ideal optical switches used within the configuration.
For instance the OXC_4×4_4ch has 16 Out_chij parameters. Allowed values:
min 0; max 16.
SPT Behavior
The optical circulator is a linear component, therefore, its effects on the
optical spectra are fully simulated during SPT simulations. The used model is
the same employed for VBS simulations.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Input_Insertion_loss [dB]: the insertion loss at each input port. Allowed
values: min 0.; max -; default 0.5.
Output_Insertion_loss [dB]: the insertion loss at each output port.
Allowed values: min 0.; max -; default 0.5.
Crosstalk [dB]: the cross-talk contribution along the counterclockwise
direction. Allowed values: min 0.; max -; default 90.
SPT Behavior
Filtering function are applied to signal spectra. When adding channels, we
assume them as in-phase: resulting power spectrum is the sum of input
channels.
Inputs Outputs
Optical Optical
Parameters
Add_Drop_Frequency [THz]: frequency of the channel to be dropped out
of the incoming WDM aggregate. A new signal at this frequency can be added
at the Add port. Allowed values: min flowSPT; max fupSPT; default 193.
Filter_Bandwidth [GHz]: the 3 dB two-sides bandwidth of the add and
drop optical filters. Allowed values: min 0; max -; default 20 GHz.
Insertion_Loss [dB]: the insertion loss of the device at each input port.
Allowed values: min 0; max -; default 3.
Crosstalk [dB]: the cross-talk contribution between the added and dropped
signals. Allowed values: min 0.; max -; default 90.
SPT Behavior
This optical demultiplexer is a linear component defined through transfer
functions of the filters that contain. It can be fully simulated in SPT
simulations that are based only on transfer function models of each
component. Therefore, the optical spectrum at each output is the one at the
input filtered using the raised cosine shaping. Since SPT simulations are
based on power spectra propagation, only the square modulus of the transfer
functions is considered.
Inputs Outputs
Optical Optical
Parameters
Channel_Spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
SPT Behavior
The optical spectrum at the output is the sum of the input power spectra
filtered by the chosen raised cosine function.
Inputs Outputs
Optical Optical
Parameters
Channel_Spacing [THz]: the frequency offset between each channel.
Allowed values: min -; max -; default 0.1.
Excess_Loss [dB]: the excess loss of the component. Allowed values: min
0.; max -; default 3.
Filter_Bandwidth [GHz]: the 3dB bandwidth of the multiplexing raised
cosine filter, refer to section 10.1.2 for better filter parameter definition.
Allowed values: min 0.; max -; default 50.
Filter_Roloff: the raised cosine filter roll-off, refer to section 10.1.2 for
better filter parameter definition. Allowed values: min 0.; max 1.; default 0.2.
SPT Behavior
Due to the non-linear behavior of SOA, simulation as linear component
defined through a transfer function is not reliable. Anyway, SPT simulations
are allowed, after displaying a warning message saying that results may be
unreliable. Refer to SPT behavior of each block to "understand" the results.
Inputs Outputs
Optical Optical
Parameters
Refer to Semiconductor Optical Amplifier (section 9.2) for a detailed
description and parameter range.
CW_linewidth [MHz]: the linewidth of the local CW laser, working at the
new wavelength. Allowed values: min 0.; max -; default 10.
CW_Power [dBm]: the power of the local CW laser, working at the new
wavelength.
Excess_Loss [dB]: the excess loss of the device. Allowed values: min 0.;
max -; default 10.
New_Wavelength [nm]: the new wavelength of the modulated optical
signal at the output of the WC. Allowed values: min c/fupSPT; max c/flowSPT;
default 1550.
SOA_Confinement_factor: the optical field confinement factor of SOA.
Allowed values: default 0.35.
SOA_Ibias [mA]: the bias current of SOA. Allowed values: default 100.
SOA_Input_loss [dB]: the input insertion loss of SOA. Allowed values:
default 3.
SOA_LW_enhancement_factor: the linewidth enhancement factor of SOA.
Allowed values: default 3.
SOA_Length [µ
µm]: the amplifier length of SOA. Allowed values: default
500.
SOA_Material_gain_const [cm2]: the material gain constant of SOA.
Allowed values: default 3*10-16.
SOA_Material_loss [cm-1]: the material loss of the SOA. Allowed values:
default 10.5.
SOA_Output_loss [dB]: the output insertion loss of SOA. Allowed values:
default 3.
SOA_Spont_carrier_lifetime [ns]: the spontaneous carrier lifetime of
SOA. Allowed values: default 0.2.
SOA_Thickness [µ
µm]: the active layer width of SOA. Allowed values:
default 0.15.
SOA_Transp_carrier_density [cm-3]: the transparency carrier density (N0)
of SOA. Allowed values: default 1*10-18.
SOA_Width [µ
µm]: the active layer width of SOA. Allowed values: default
1.5.
SPT Behavior
Due to the non-linear behavior of SOAs and interferometric structure,
simulation as linear component defined through a transfer function is not
reliable. Anyway, SPT simulations are allowed, after displaying a warning
message saying that results may be unreliable. Refer to SPT behavior of each
block to "understand" the results.
Inputs Outputs
Optical Optical
Parameters
Refer to Semiconductor Optical Amplifier (section 9.2) for a detailed
description and parameter range.
CW_linewidth [MHz]: the linewidth of the local CW laser, working at the
new wavelength, contained inside the WC. Allowed values: min 0.; max -;
default 10.
CW_Power [dBm]: the power of the local CW laser, working at the new
wavelength, contained inside the WC.
Excess_Loss [dB]: the excess loss of the device. Allowed values: min 0.;
max -; default 10.
New_Wavelength [nm]: the new wavelength of the modulated optical
signal at the output of the WC. Allowed values: min c/fupSPT; max c/flowSPT;
default 1550.
Passive
Attenuator, simulates an ideal optical attenuator.
Non-Linear
Frequency Shifter, shifts the optical signal in the spectral domain.
SPT Behavior
The attenuator introduces the requested loss to the input optical spectrum.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Attenuation [dB] or [lin]: the input/output optical attenuation. Note that
the linear scale refers to the amplitude attenuation, not to the power
attenuation. For instance a 3 dB attenuation corresponds to a square root of
2 linear attenuation. Allowed values: min 0 (dB); max - (dB); default 0 (dB).
In1 Out1
In2 Out2
SPT Behavior
The optical coupler is a linear component therefore its effects on the optical
spectra are fully simulated during SPT simulations. The used model is the
same employed for VBS simulations.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
Output port ratio [-]: represents the power ratio, on a percentage basis,
between each of the inputs and the corresponding cross output, (Pout2/Pin1 or
Pout1/Pin2). Allowed values: min 0; max 100; default 50.
Complementary output port ratio [-]: the complementary power ratio,
between each of the inputs and the corresponding direct output (Pout1/Pin1 or
Pout2/Pin2).
19.3 Splice
This component has a graphical utility only; it has no functions except
allowing optical lines to be connected in arbitrary shapes, by putting a splice
on each vertex. It is used also in the iterated components (see the OptSim
User Manual) because this is a fictitious optical component with one optical
output, useful to draw a period of optical link.
Inputs Outputs
Optical Optical
SPT Behavior
An ideal fiber grating introduces a certain amount of dispersion, without
influencing the optical power spectrum if the bandwidth of the channels is
neglected, like in SPT simulations. Therefore, in SPT simulations, the ideal
fiber gratings do not influence the optical power spectrum. For the behavior
of the user-defined fiber grating, please refer to the optical filter description
(section 10.1).
Parameters
Name: the optional name that appears in the component label.
Type: opens the related dialog windows for the two models, Ideal Fiber
Grating and User-Defined Fiber Grating.
19.4.1 Ideal
This component implements an ideal fiber grating, i.e. a passive device with a
flat modulus frequency response and a group delay corresponding to a given
total dispersion and dispersion slope.
These parameters represent the dispersion value of the fiber at the reference
wavelength (frequency). The dispersion value can be given as a function of
Dtot or β2,tot, so it can be directly related to the standard fiber dispersion
parameter.
Inputs Outputs
Optical Optical
Parameters
Reference Frequency/Wavelength [THz] or [nm]: the reference
frequency for the following dispersion parameters. Allowed values: min 0;
max -; default f0,VBS.
Total Dispersion to compensate at the Reference
Frequency/Wavelength [ps/nm] or [ps2]: the dispersion value to
compensate. It can be either inserted as Dtot = - DL [ps/nm] or as β2,tot = -
β2L [ps2], where L is the length of the fiber span and D (or β2) is the
dispersion coefficient of the fiber span. If a different degree of compensation
has to be tested, then the following criteria must be followed: L = 1/100
comp% Lactual, where comp% is the percentage of compensation needed, and
Lactual, is the actual length of the span. Allowed values: min -; max -; default
0.
Total Dispersion Slope to compensate at the Reference
Frequency/Wavelength [ps/nm2] or [ps3]: the dispersion derivative to
compensate. It is related to the third order dispersion, i.e. the variation of the
dispersion with respect to the wavelength, ∂D/∂λ coefficient (or with respect
to the angular frequency, ∂β2/∂ω coefficient). Its value can be introduced
using either the derivative of the Dtot coefficient with respect to the
wavelength (∂Dtot/∂λ = -(∂D/∂λ)L [ps/nm2]), or the derivative of β2,tot
19.4.2 User-defined
To simulate any kind of experimentally measured fiber grating, a user-defined
optical filter component should be employed. Therefore, the transfer function
of the component can be defined in terms of amplitude and phase (or group
delay) through the definition in an ASCII file. Please read the section 10.1.5
to obtain detailed information on the parameters setting.
Inputs Outputs
Optical Optical
3π
∆τ p = δ PMD L p
8
Moreover, the waveplates length Lp is related to the fiber beat length LB and
the fiber correlation length LC by the following relation [5]:
Lp
Lc = −
L
ln cos π p
LB
The resulting DGD ∆τ at the and of the fiber span is a random variable with a
Maxwellian distribution and the Probability Density Function takes the
following form [6]:
2 ∆τ 2 ∆τ 2
P (∆τ , z ) = exp − 2 with 0 < ∆τ < +∞
π q3 2q
π
where q2 = δ PMD
2
L fiber .
8
The average DGD is ∆τ = δ PMD L fiber , and its variance is
3π
σ ∆τ = − 1 ⋅ δ PMD L fiber .
8
The DGD value is also reported as output using the popup menu command
View Data on the component icon, and it can be also correlated using the
popup menu command View Chart.
SPT Behavior
Since SPT simulations are based on the power spectrum of the optical signal,
the information about polarization can not be managed with the SPT
simulation technique. This component works on the polarization, thus it can
not be simulated with the SPT simulation technique. SPT simulation does not
work with projects containing this component.
Inputs Outputs
Optical Optical
Parameters
Instance [-]: the instance is an integer number that identifies the block.
This number is used by OptSim to generate a data file that contains the block
output signal. Allowed values: min: 1; max: -; default 1.
This value must be used with care. In particular, if the same model includes
several instances of the AOPMD component, this parameter value has to be
different for each AOPMD block. Moreover, when you make a multiple
(parametric) run you need to use a variable also in this parameter field.
For example in multiple runs on "power" parameter: 0-2-4, you have to
define the following table (that appears choosing the menu command
Simulate > Define Runs):
run1: power=0, instance=1
run2: power=2, instance=2
run3: power=4, instance=3
Where the AOPMD block has the variable "instance" as this parameter value.
Take care when using the following particular configurations:
Project with several AOPMD blocks: in this case the instance value has to
be different for each AOPMD block.
Project with a multiple (parametric) run:
− systems with one AOPMD block and with parametric runs that use one
or more AOPMD parameters as variables. In this case the instance
has to be different for each run so you need to use a variable also in
this parameter field
− systems with one or more AOPMD blocks and with parametric runs
that use one or more AOPMD parameters as variables. In this
configuration the AOPMD blocks involved in the parametric run need
different instance values for each run. So you have to add the
corresponding instance as a variable
− systems with parametric runs that do not involve any AOPMD blocks.
In this case the instance value can maintain the same value
independently to the runs
Inclusion of AOPMD in an iterate block:
− if you need to include in an iterate structure an AOPMD that maintains
always the same behavior, it is sufficient to insert an arbitrary value
for the instance
− if you want to reproduce a more realistic simulation in which the PMD
effect induced by AOPMD is not exactly the same at each span, it is
not possible to use the current iteration structure present in OptSim.
The workaround is to unroll the configuration where each different
AOPMD block has different instance value
Length [Km]: specifies the length of the PMD fiber. Allowed values: min: 0;
max: -; default 0.
PMD_coefficient [ps/km1/2]: the PMD coefficient δPMD that is usually
specified in data sheets to quantify the amount of PMD introduced by the
fiber. Allowed values: min: 0; max: -; default 0.1.
Correlation_length [Km]: the fiber correlation length Lc expressed in
kilometers. It indicates the distance over which the State of Polarization of a
signal loses memory of the input State of Polarization due to PMD and
birefringence. Allowed values: min: >0; max: -; default 0.2.
Beat_length [m]: the fiber beat length LB expressed in meters. It is defined
as the distance after which the polarization of the propagating optical signal
recovers its initial state. Allowed values: min: 0; max: -; default 5.
Seed [-]: an integer number that defines the seed for the statistical process
used in the PMD model. Allowed values: min: 0; max: -; default 1.
Note
The AOPMD component allows the simulation of the PMD phenomenon and its
repeatability. In fact it is characterized by this own seed parameter that
ensures its reproducibility in different projects.
SPT Behavior
During SPT simulations the optical frequency shifter shifts the input spectrum
of the spectral amount specified. Note that a circular shift is made: for large
frequency shift the results may not agree with optical power spectra obtained
in VBS simulations.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Frequency Shift [GHz] or [nm]: the value of the applied spectral shift.
Allowed values: min - BWSPT/2; max + BWSPT/2; default 0.
EIN = A(t )e jΦ (t )
SPT Behavior
The phase conjungator is a non-linear device and therefore cannot be reliably
simulated using the SPT technique. During SPT simulations, this component is
considered taking into account its attenuation only, i.e. the efficiency. No
other spectral modifications are considered.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Efficiency: the efficiency η of the device referred to the signal amplitude.
Allowed values: min 0; max 1; default 1.
Phase Shift [rad]: the extra constant phase shift φSHIFT applied to the output
optical signal. Allowed values: min 0; max 2π; default π/2.
SPT Behavior
Since SPT simulations are based on the power spectrum of the optical signal,
the information about polarization can not be managed with the SPT
simulation technique. This component works on the polarization, thus it can
not be simulated with the SPT simulation technique. SPT simulation does not
work with projects containing this component.
Inputs Outputs
Optical Optical
Parameters
Name: the optional name that appears in the component label.
Poincare' Sphere Rotation Angle about Axis S1 [degrees]: the rotation
in degrees around the S1 axis. Allowed values: min -; max -; default 0.
Poincare' Sphere Rotation Angle about Axis S2 [degrees]: the rotation
in degrees around the S2 axis. Allowed values: min -; max -; default 0.
Poincare' Sphere Rotation Angle about Axis S3 [degrees]: the rotation
in degrees around the S3 axis. Allowed values: min -; max -; default 0.
Number of Rotations: the number of actually generated rotations, None,
One, Two or Three.
Rotation Set of graphical buttons that select the rotation order.
Order
First Rotation: sets the first rotation; these buttons are enabled only if the
selected number of rotation is one, two or three.
Second Rotation: sets the second rotation; these buttons are enabled only
if the selected number of rotation is two or three.
Third Rotation: sets the third rotation; these buttons are enabled only if the
selected number of rotation is three.
SPT Behavior
Since SPT simulations are based on the power spectrum of the optical signal,
the information about polarization can not be managed with the SPT
simulation technique. This component works on the polarization, thus it can
not be simulated with the SPT simulation technique. SPT simulation does not
work with projects containing this component.
Inputs Outputs
Electrical Optical
Parameters
Name: the optional name that appears in the component label.
Scrambling Law: selects the scrambling law, Square or Sine.
Internal Clock Rate [GHz]: the frequency of the periodic variation. Allowed values: min 0;
Synchronizati max BWSPT/4; default 10.
on
Phase Delay [ns]: the delay of the internal clock. Allowed values: min 0;
max 1/(Rate); default 0.
19.10 Polarizer
This component simulates a polarizer, i.e. a component that allows to select a
specific polarization of the input optical field, blocking the orthogonal one.
You can select the axis of the polarizer, i.e. the polarization that is allowed to
pass through the polarizer, by specifying this polarization in the Stokes space
representation. More specifically, you should specify the two angles 2α and
2φ required to determine this polarization in the Stokes space using a
spherical coordinate system (azimuthal and zenithal angles).
If you are familiar with the elliptical representation of the state of
polarization, we remind that the angle α corresponds to the Angle of Major
Ellipse Axis respect to the x Geometrical Axis and the angle φ corresponds to
the arc tan of the ratio of Minor over Major Ellipse Axis, with a positive sign
for Clockwise Polarization (clockwise polarization is such that the corkscrew-
rule would generate a vector parallel to the direction of propagation).
SPT Behavior
Since SPT simulation is based on the power spectrum of the optical signal,
the information about polarization is not managed with the SPT simulation
technique. Although, this component works on signal polarization, which is
not considered during SPT, the user can take into account the peak and
average power reduction by using the parameters reported in the Spectral
Estimation section.
Inputs Outputs
Optical Optical
POL1
DELAY τ
Input Output
POL2
The input optical signal is split into two branches and sent to the two
polarizers POL1 and POL2. The two polarizers are complementary: you can
set the POL1 polarization in the Stokes space representation, POL2 is
automatically set to the orthogonal polarization. See the polarizer description
for more details (section 19.10).
More specifically, for the polarizer POL1 you should specify the two angles 2α
and 2φ required to determine this polarization in the Stokes space (on the
Poincarè sphere) using a spherical coordinate system (azimuth and zenith
angles).
After the POL1 polarizer, an optical delayer is inserted, whose delay τ is set
by the rdelay parameter. The signal exiting from POL2 is not delayed. Finally
the two signals are added. As a result the two components experience a
differential delay.
Through the parameter rdelay, you can set the optical relative group delay τ
between the two branches. Notice that the optical delay actually applied in
the simulation is an integer multiple of the simulation time sampling step, so
the delay effectively used τ’ is related with the simulation bandwidth through
the relationship:
τ ⋅ BWVBS 0.8
τ ' = Nearest _ Integer
0.8 BWVBS
where BWVBS is the VBS simulation bandwidth.
SPT Behavior
Since SPT simulations are based on the power spectrum of the optical signal,
the information about polarization can not be managed with the SPT
simulation technique. In SPT simulation this component leaves the input
signal unchanged.
Inputs Outputs
Optical Optical
Parameters
rdelay [ps]: the value of requested differential group delay τ. Allowed
values: min 0; max: Tsim (Total Simulated Time Span simulation
parameter); default 0.
2alpha [degrees]: the value of the azimuth angle of polarizer POL1 axis
with respect to axis S1 on the Poincaré sphere. Allowed values: min -360;
max +360; default 0.
2psi [degrees]: the value of the zenith angle of polarizer POL1 axis with
respect to the equator on the Poincaré sphere. Allowed values: min -90; max
+90; default 0.
Coupler opt_out1
opt_in
opt_out2
τ
Figure 19.3 Ideal dual-arm Mach-Zehnder interferometer block scheme
1
τ=
BW
The interferometer input-outputs transfer functions can be tuned using the
Tuning Frequency parameter.
Note
The actual value of the optical delay used in the time domain simulation must
be an integer multiple of the sampling time, so the delay effectively used τ' is
related with the user required delay τ through the relationship:
τ ⋅ BWVBS 0.8
τ ′ = Nearest _ Integer
0.8 BWVBS
where BWVBS is the VBS simulation bandwidth.
SPT Behavior
In SPT simulations the input Power Spectrum is multiplied by the following
transfer functions:
1
2
H1 ( f ) = {1 + cos[2π ( f − f tune )τ ]} = cos 2 [π ( f − f tune )τ ]
2
1
H2( f ) =
2
{1 − cos[2π ( f − f tune )τ ]} = sin 2 [π ( f − f tune )τ ]
2
Inputs
Optical
Outputs
Optical
Parameters
Name: the optional name that appears in the component label.
Delay [ps]: the value of the requested delay τ. Allowed values: min
[0.8/VBSBW]; max: [(1/20)·Tsim]; default 25.
Corresponding Simulated Delay [ps]: it shows the actual simulated delay
τ according to the current VBS bandwidth set in the Simulated Parameters
dialog windows.
Bandwidth [GHz]: it sets the bandwidth of the passbands in the two
transfer functions. It represents another way to characterize the value of the
optical delay τ. Due to the intrinsic structure of the interferometer, this value
corresponds also to the frequency spacing between passband peak or notches
in both output port transfer functions (see Figure 19.4). Allowed values: min:
[20·(1/Tsim)]; max: [1.25·VBSBW]; default: 40.
Tuning Frequency [THz] or [nm]: it sets the center frequency of a
passband in |H1(f)|2 , the transfer function of output port 1. At the same time
there will be a notch at this frequency on output port 2. Allowed values: min
flowSPT; max fupSPT; default: f0,SPT.
Example
As an example, we report in the following picture the magnitude square of
the interferometer transfer functions having τ = 25 ps and ftun = 194 THz.
BW
|H1(f)|2 |H2(f)|2
BW
ftun
19.13 References
[1] M. Born and E. Wolf, Principles of optics, pp. 554-556. Oxford, England:
Pergamon Press, 1975.
[2] P. Poggiolini and S. Benedetto Theory of polarization spreading
techniques - Part I, IEEE Transactions on Communications, vol. 42, n. 5,
May 1994.
[3] P. Poggiolini and S. Benedetto Theory of polarization spreading
techniques - Part II, IEEE Transactions on Communications, vol. 42, n. 5,
May 1994.
[4] C.D. Poole et al., "Polarization-Mode Dispersion Measurement Based on
transmission Spectra Through a Polarizer" IEEE/OSA Journal of Lightwave
Technology, vol. 12, N0. 6, pp. 917–929, June 1994.
[5] F. Corsi et al., "Polarization Mode Dispersion Characterization of Single-
Mode Optical Fiber Using Backscattering Technique," IEEE/OSA Journal of
Lightwave Technology, vol. 16, N0. 10, pp. 1832–1843, October 1998.
[6] F. Curti et al., "Statistical treatment of the evolution of the principal
states of polarization in single-mode fibers," IEEE/OSA Journal of
Lightwave Technology, Vol. 8, N0. 8, pp. 1162–1165, August 1990
Optical link, implements an ideal optical fiber where no effects on the optical
field are taken into account.
Inputs Outputs
Optical Optical
Electrical Electrical
Inputs Outputs
Logical Logical
Optical
Combiner, simulates an optical combiner.
Electrical
Combiner, simulates an electrical combiner.
Logical
Splitter, simulates a logical splitter.
Split&Combine
Inputs Outputs
Note
This component implements a balanced combiner with the same attenuation
on each input. If an unbalanced combiner is required, simply precede the
combiner with the proper attenuators on each input.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
Attenuation on each output [dB] or [lin]: it is the amplitude attenuation.
Allowed values: min 0 (dB); max -; default 0 (dB).
Note
This component implements a balanced splitter with the same attenuation on
each output. If an unbalanced splitter is required, simply follow the splitter
with the proper attenuators on each output.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
Attenuation on each output [dB] or [lin]: it is the amplitude attenuation.
Allowed values: min 0 (dB); max -; default 0 (dB).
Note
This component implements a balanced combiner with the same attenuation
on each input. If an unbalanced combiner is required, simply put an
attenuator before the combiner.
Inputs Outputs
Parameters
Name: the optional name that appears in the component label.
Attenuation on each output [dB] or [lin]: it is the amplitude attenuation.
Allowed values: min 0 (dB); max -; default 0 (dB).
Note
This component implements a balanced splitter with the same attenuation on
each output. If an unbalanced splitter is required, simply follow the splitter
with the proper attenuators on each output.
Optical
Power Meter, measures the power (mean square value) of an optical signal.
Probe, simulates a generic probe for optical signals. It collects data for
diagrams such as power, phase, instantaneous frequency, power spectrum
and Stokes parameters.
Electrical
Power Meter, measures the power (mean square value) of an electrical
signal.
Q Estimator, measures the Q value: a pattern length for system affected by
ISI can be specified. Besides the Q value, the following information are
provided: optimal threshold, eye closure, average eye opening, eye opening,
tolerance to sampling instant variation, jitter measurement for RZ signal,
diagram of Q value versus sampling instant, diagram of eye closure versus
sampling instant.
BER Estimator, measures the Bit Error Rate for a binary modulation starting
from statistical moments of the received signal. Gaussian and χ2 statistical
distribution approximation are available.
Scope, simulates an oscilloscope for electrical signals. It collects data for
diagrams such as amplitude, eye diagram, histogram at the optimum
sampling instant and power spectrum.
Logical
Logical Signal Probe, prints the logical signal on a chart.
SPT Behavior
Measurement components, even if measuring optical signals, are not taken
into account in SPT simulations. The only outputs of SPT simulations are
optical spectra, power and OSNR measurements over each optical link.
Inputs Outputs
Optical -
Parameters
Name: the optional name that appears in the component label.
Measure Bandwidth: it selects how the optical power meter will perform the
evaluation. With the Whole option the whole BWVBS will be considered for the
power evaluation. With the Limited option only the bandwidth set with the
following parameters will be considered.
Center frequency [THz] or [nm]: it is the center frequency of the
bandwidth where the power of the input optical signal will be evaluated. This
field is enabled only if the Limited option is selected in Measure
Bandwidth. Allowed values: min: flowVBS; max: fupVBS; default: f0,VBS.
Bandwidth [GHz] or [nm]: it is the bandwidth over which the power
evaluation will be performed. This field is enabled only if the Limited option
is selected in Measure Bandwidth. Allowed values: min 0; max BWVBS;
default 50 GHz.
SPT Behavior
Measurement components, even if measuring optical signals, are not taken
into account in SPT simulations. The only outputs of SPT simulations are
optical spectra, power and OSNR measurements over each optical link.
Inputs Outputs
Optical -
Parameters
Name: the optional name that appears in the component label.
Number of Spectrum Points over the Simulation Bandwidth: it is the
number of frequency samples to be displayed over the time domain
Simulation Bandwidth. Starting from this parameter, will be evaluated the
number of points to be used in the FFT with the modified periodogram
method. It implies the minimum spectral frequency resolution. Allowed
values: min: 2; max: 500000; default 1000.
SPT Behavior
Measurement components, even if measuring optical signals, are not taken
into account in SPT simulations. The only outputs of SPT simulations are
optical spectra, power and OSNR measurements over each optical link.
Inputs Outputs
Optical -
Parameters
Name: the optional name that appears in the component label.
Measured time-span: it selects the time window where the measurement
will be performed. With the Whole option the whole Tsim will be considered.
With the Limited option only the time span set with the following parameters
will be considered.
Measure Start Time [ns]: it is the starting time of the measurement
window. The available time window is given by the chosen Total Simulated
Time Span (Tsim). Allowed values: min 0; max Tsim; default 0.
Measure End Time [ns]: it is the ending time of the measurement window.
The available time window is given by the chosen Total Simulated Time
Span (Tsim). It must be greater than Measure Start Time. Allowed values:
min 0; max Tsim; default 0.
Sampling Factor: it is the ratio between the simulation time domain
sampling rate and the sub-sampling rate; basically if this factor is equal to n,
one sample out of n is displayed. Allowed values: min 1; max -; default 1.
Note
Please note that changing the sampling factor, the displayed spectrum may
differ from the actual spectrum of the signal propagated along the link. This is
due to the reduction of the considered bandwidth caused by the subsampling
that may create aliasing effects in the Discrete Fourier Transform computed
by the block.
Electrical -
Parameters
Name: the optional name that appears in the component label.
Bit Rate Nominal [Gbit/s]: it is the bit-rate of the input signal. Allowed values:
min -; max -; default Rb.
Simulated [Gbit/s]: it shows the actual simulated bit-rate with the current
VBS bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it shows the number of sample per bit at the Simulated
bit-rate.
Time Resolution: it specifies the time resolution. With Automatic the
number of visualized points per bit is automatically set to the minimum value
between the Samples per bit value and 64, in order to avoid the generation
of huge output files in simulation involving large bandwidth or large simulated
time window. With Custom the number of visualized points per bit is set in
the following field. This option is useful to reduce the output file dimension,
setting a value less than the Samples per bit value.
Number of visualized points per bit: it is the number of visualized points
per bit. This field is enabled only if the Custom option is selected in the Time
Resolution list. Allowed values: min 1; max Samples per bit;
default Samples per bit.
Note
Please note that changing the number of visualized points per bit, the
displayed electrical spectrum may differ from the actual spectrum of the
simulated signal. This is due to the reduction of the considered bandwidth
caused by the subsampling that may create aliasing effects in the Discrete
Fourier Transform computed by the block.
Electrical -
Parameters
Name: the optional name that appears in the component label.
Measure Bandwidth: it selects how the electrical power meter will perform
the evaluation. With the Whole option the whole BWVBS will be considered for
the power evaluation. With the Limited option only the bandwidth set with
the following parameters will be considered.
Center frequency [GHz]: it is the center frequency of the bandwidth where
the power of the input electrical signal will be evaluated. This field is enabled
only if the Limited option is selected in Measure Bandwidth. Allowed
values: min 0; max BWSIM/2; default BWSIM/4.
Bandwidth [GHz]: it is the bandwidth over which the power evaluation will
be performed. This field is enabled only if the Limited option is selected in
Measure Bandwidth. Allowed values: min 0; max BWSIM/2; default BWSIM/2.
22.6 Q Estimator
This component estimates the Q factor of the received signal. Following its
standard definition, the Q factor is calculated using mean and standard
deviation of the signal samples taken at the optimum sampling instant and
considering the optimum decision threshold, i.e., using the following
expression:
m1 − m0
Q= (1)
σ1 + σ 0
where m1, m0, σ1, σ0 are the mean and standard deviation of the signal
samples when a "1" or a "0" is received. We remind that under the Gaussian
approximation, the bit error probability (BER) is given by:
1 Q
P(e ) = erfc (2)
2 2
Moreover, we remind here that the Q factor is meaningless for any kind of
multilevel digital modulation or for an analog modulation. It only makes sense
for binary digital modulation.
The Q factor standard definition does not consider any pattern dependence,
i.e. any effect related to the intersymbol interference (ISI). Consequently,
this component introduces an alternative evaluation of the Q factor that takes
into account the pattern dependence. If it can be assumed that the decision
variable relative to a given bit interval depends only on the interval itself and
on the previous n and the next n bits, the error probability can be estimated
as:
1 22 n +1 m − Vth
P(e ) = min Vth 2 n+1 ∑ erfc i
(3)
2 i =1 2σ i
where the error probability is averaged on the 22n+1 possible patterns of 2n+1
bits, then optimized in term of the decision threshold Vth.
Finally the global Q factor is estimated by inverting the formula (2). The
resulting Q factor corresponds to the traditional one when n=0, i.e. if no
pattern dependence is considered.
The pattern length should be specified by the Npattern parameter, which
assumes the value Npattern=2n+1. Thus, it must be an odd number. If the
total number of transmitted bits is too small to ensure a reasonable accuracy,
the pattern length is automatically changed to Npattern=1 and a warning is
issued. Reasonable values for the Npattern parameter are 1, 3 and 5.
Moreover the bit-time in OptSim, i.e. the time-duration of the bit, must be an
integer number of time-samples NS (Samples per bit value). The sample-
period depends on the VBS bandwidth, therefore introducing separately the
VBS bandwidth and the reference bit-rate, in general, the requested bit-rate
cannot be simulated because the bit-time is not an exact multiple of the
sample-period. The bit-rate is therefore rounded (the value is displayed by
the Simulated parameter) to the closest value whose bit-time is an exact
multiple of the sample-period.
Note
To obtain significant results, at least 100 bits for each of the considered
patterns should be simulated. Consequently, when simulating less that
approximately 200 bit, it is advisable to set Pattern Length=1. In general, if
Pattern Length=L, the number of simulated bits should be at least 2*M*L,
where M should be in the range [50-100] or greater. If these conditions are
not met, the Q-factor estimation may be inaccurate due to Monte-Carlo
fluctuation of the results. See the white paper dedicated to this issue in the
OptSim User Manual or [1] for further details.
Measurement period can be reduced with respect to the Total Simulated Time
Span simulation parameter Tsim. Measure Start Time and Measure End
Time parameters can be specified to select a sub-window.
Numerical The numerical results displayed by this component are:
Results
Optimal threshold: the decision threshold that maximize the Q value,
at the optimum sampling instant
Q value: it is the evaluated Q value, obtained using the optimum
decision threshold at the optimum sampling time, displayed on a linear
scale
Q value [dB]: it is the evaluated Q value, obtained using the optimum
decision threshold at the optimum sampling time, displayed on a
logarithmic scale (20*log(Qlin))
Note
The maximum value evaluated is 40 dB. When greater values are obtained
they are truncated to 40 dB and a warning message is given.
t j = t 0 + jT .
For each bit, the maximum is found at the instant tjmax, being
max
t 0 + jT < t j < t 0 + ( j + 1)T .
∆t j = t j − t max
j .
t
t0+T t1+T t2+T t3+T t4+T
max max max
t0 t2 t3
Graphical The following graphical results can be enabled or disabled in the Additional
Results Outputs section:
Q-factor vs. sampling instant: it shows a plot reporting the Q factor in
dB vs. the sampling instant inside the bit duration. This measurement is
very useful to study the tolerance to variation of the sampling instant
with respect to the optimum sampling instant and on the horizontal
opening of the received eye diagram.
Eye closure vs. sampling instant: it is about the same as the previous
plot, except that the eye closure parameter is reported on the y axis.
Use of the The Q-estimator component implemented in OptSim is based on an algorithm
component that automatically looks for the optimum sampling instant and the optimum
decision threshold. As you can see from the list of the input parameters, it
does not require any knowledge about the digital data stream that is in input
to the block or, in other words, it does not require any form of
synchronization with the transmitter. The result is a flexible and easy to use
component.
The automatic algorithm performing these features works properly in nearly
all cases of interest, except the following:
when the received eye is totally closed or exceedingly distorted
when the transmitted data stream has a strongly unbalance between the
number of logical “1” and “0”
Whenever these two conditions are not met, warning or error messages are
issued. In general, in these cases, you must be aware that the numerical
results could be unreliable.
Warning and This component generates several kind of warning or error messages
Errors signaling anomalous situations. Here is a brief summary of them.
The Q factor is very large. The received eye diagram is extremely open
and the resulting Q-factor is truncated to a conventional maximum value
of 40 dB.
The Q factor is very low. The received eye diagram is nearly closed or
strongly distorted, so that numerical results may be unreliable.
The number of occurrences of one of the received pattern is low. This
situation can be generated by two different situations:
− the eye diagram is strongly distorted and the automatic algorithm
that optimize the sampling instant and the decision threshold cannot
find a proper value (see the following section)
− the transmitted data stream shows an unbalance between the
number of “1” and “0”. In this case, we advice to choose a proper
pseudo-random sequence degree in the Datasource component
and/or increase the number of simulated bits
The different patterns have a strongly unbalanced number of occurrences.
Similar to the previous error.
The number of occurrences of a given pattern is null. In this case, it is
impossible to evaluate the mean and variance related to that pattern.
Inputs Outputs
Electrical -
1 +∞
Vth
1
P(e ) = [P(e 0 ) + P(e 1)] = ∫ fψ (x 0)dx + ∫ fψ (x 1)dx
2 2 Vth −∞
where fψ(x|0), fψ(x|1) are the pdf of the decision signal when a “0” or “1”
are received, respectively. Vth is the decision threshold, which is
automatically optimized to minimize the BER.
Please refer to section 22.7.1 for further information on BER estimation
technical details.
When a χ2 distribution is chosen, you should also give the distribution degree
of freedom. In the assumption that the prevailing noise comes from the
optical components (typically ASE in EDFA) and that standard Direct
Detection is used, the degree of freedom Nχ should be set approximately to:
2NpolBopt/Bel. Bopt is the bandwidth of the optical filter, Bel is the bandwidth of
the receiver postdetection electrical filter and Npol=2 if the simulation is based
on a two polarization model, while Npol=1 for single polarization.
This component can also take into account pattern dependence. In this case,
it is assumed that the decision variable relative to a given bit interval
depends only on the interval itself and the previous n and the next n bits. The
moments are then evaluated over the 22n+1 possible patterns and then again
fitted to the chosen pdf. Finally, BER is evaluated using an extension of the
previous equation. Evaluating BER taking into account pattern dependence is
fundamental when the received eye-diagram is significantly distorted due to
fiber dispersion, non-linearities or improper receiver filters. In these cases,
neglecting pattern dependence usually gives strongly overestimated BER.
The component requires the Resolved Bit Pattern Length, which is defined
as (2n+1) and should thus be an odd-number. If the total number of
transmitted bits is too small, the pattern length is automatically fixed to 1.
Moreover the bit-time in OptSim, i.e. the time-duration of the bit, must be an
integer number of time-samples NS (Samples per bit value). The sample-
period depends on the VBS bandwidth, therefore introducing separately the
VBS bandwidth and the reference bit-rate, in general, the requested bit-rate
cannot be simulated because the bit-time is not an exact multiple of the
sample-period. The bit-rate is therefore rounded (the value is displayed by
the Simulated parameter) to the closest value whose bit-time is an exact
multiple of the sample-period.
Note
To obtain significant results, at least 100 bits for each of the considered
patterns should be simulated. Consequently, when simulating less that
approximately 200 bit, it is advisable to set Pattern Length=1. In general, if
Pattern Length=L , the number of simulated bits should be at least 2*M*L,
where M should be in the range [50-100] or greater. If these conditions are
not met, the Q-factor estimation may be inaccurate due to Monte-Carlo
fluctuation of the result. See the white paper dedicated to this issue in the
OptSim User Manual, or [1] for further details.
Measurement period can be reduced with respect to the Total Simulated Time
Span simulation parameter Tsim. By the Measure Start Time and Measure
End Time parameters a sub-window may be selected.
Numerical The numerical results displayed by this component are:
Results
BER at optimal decision threshold: estimated BER at optimal
threshold
Equivalent Q at optimal decision threshold: Q related to estimated
BER at optimal threshold, obtained inverting the formula:
1 Q
BER = erfc
2 2
BER at mean decision threshold: estimated BER at mean threshold,
defined as the mean value of signal time samples.
Equivalent Q at mean decision threshold: Q related to estimated BER
at optimal threshold, obtained inverting previous the formula
Note
The minimum value estimated is 10-40. When smaller values are obtained
they are truncated to 10-40 and a warning message is given.
Graphical The following graphical results can be enabled or disabled in the Additional
Results Outputs section:
BER vs. sampling instant: it shoes a plot reporting the BER versus the
sampling instant inside the bit duration
BER vs. decision threshold: a plot reporting the BER versus the
decision threshold
Warning and The warning and errors generated by this component are similar to those
Errors generated by the Q-estimator component. Please see the Q-estimator
Warning and Errors section for further details (section 22.6).
Specific warnings are also given in relation to the use of the χ2
approximation, whenever it is not possible to fit the estimated moments of
the received signals with the moment of a χ2 distribution.
Inputs Outputs
Electrical -
BER vs. Decision Threshold [-]: it enables or disables the graphical plot
reporting the BER versus the Decision Threshold.
BER Evaluation at Relative Decision Threshold [-]: it enables or disables
the evaluation of the BER at the relative decision threshold set in Relative
Threshold Value.
Relative Threshold Value [-]: it sets the threshold as percentage of input
signal maximum excursion (i.e. the range between the absolute minimum
and maximum of the received signal) and it is used to evaluate the BER at
the Relative Decision Threshold. The actual decision threshold used for BER
estimation will be:
Gaussian Approximation
If you choose the Gaussian approximation, the BER is evaluated, starting
from the received signal averages m1, m0 and standard deviation σ1, σ0 as:
1 m1 − Vth V − m0
P(e ) = erfc + erfc th
2σ
4 2σ 1 0
The threshold level Vth is then optimized so to give the minimum P(e).
χ2 Approximation
We remind here that in general a χ2 non-central random variable is the sum
of Nχ Gaussian variables having the same variance and generic average, i.e.
can be put in the form:
Nχ
χ = ∑ψ i2
i =1
where ψi are Gaussian random variables with a given variance σι2 and
average mi. The parameter Nχ is called the χ2 "degree of freedom", while the
parameter:
Nχ
s 2 = ∑ mi
i =1
2 N pol Bopt
Nχ =
Bel
where Bopt is the bandwidth of the optical filter, Bel is the bandwidth of the
receiver postdetection filter and Npol=1 or 2 depending on whether single or
dual polarization is considered.
Even if these assumptions are rarely met in practice, nevertheless the model
can be used in general, provided that the prevailing noise comes from EDFA
ASE.
If you choose the χ2 approximation, the distribution parameters are evaluated
starting from the received signal averages m1, m0, variances σ1, σ0 and the
specified χ2 degree of freedom Nχ, by solving a set of nonlinear equation.
Afterwards, the χ2 pdf are integrated according to:
1 +∞
Vth
Electrical -
Parameters
Name: the optional name that appears in the component label.
Displayed Start [GHz]: it is the lower frequency of the displayed section of spectrum.
Frequency Allowed values: min 0; max: BWVBS/2; default: 0.
Range
Stop [GHz]: it is the upper frequency of the displayed section of spectrum.
Allowed values: min: 0; max BWVBS/2; default: BWVBS/2.
Number of Spectrum Points over the displayed Bandwidth: the number
of frequency samples to be displayed over the chosen bandwidth. Starting
from this parameter, will be evaluated the number of points to be used in the
FFT with the modified periodogram method. Allowed values: min: 2; max:
500000; default: 500.
Electrical -
Parameters
Name: the optional name that appears in the component label.
Bit Rate Baud_rate [Gbaud/s]: it is the baud-rate of the input signal.
Simulated [Gbit/s]: it shows the actual simulated baud-rate with the
current VBS bandwidth set in the Simulation Parameters dialog window.
Samples per bit: it shows the number of sample per symbol at the
Simulated baud-rate.
Measured time-span: it selects the time window where the measurement
will be performed. With the Whole option the whole Tsim will be considered.
With the Limited option only the time span set with the following parameters
will be considered.
Measure Start Time [ps]: it is the starting time of the measurement
window. The available time window is given by the chosen Total Simulated
Time Span (Tsim). Allowed values: min 0; max Tsim; default 0.
Measure End Time [ps]: it is the ending time of the measurement window.
The available time window is given by the chosen Total Simulated Time
Span (Tsim). It must be greater than Measure Start Time. Allowed values:
min 0; max Tsim; default Tsim.
Subcarrier Frequency [GHz]: it is the signal subcarrier frequency. Allowed
values: min 0; max BWSIM/2; default BWSIM/2.
Parameters
TSTART [ps]: it is the starting time of the measurement window. The
available time window is given by the chosen Total Simulated Time Span
(Tsim). Allowed values: min 0; max Tsim; default 0.
TSTOP [ps]: it is the ending time of the measurement window. The available
time window is given by the chosen Total Simulated Time Span (Tsim). It
must be greater than Measure Start Time. Allowed values: min 0; max Tsim;
default Tsim.
Baud_rate [Tb/s]: it is the baud-rate of the input signal.
Delay_sync [ps]: it selects the sampling instant.
KOPT: it defines the plot option:
1: the continuous scattering diagram is plotted
2: the sampled scattering diagram is plotted
Logical -
Parameters
Name: the optional name that appears in the component label.
Measured time-span: it selects the time window where the measurement
will be performed. With the Whole option the whole Tsim will be considered.
With the Limited option only the time span set with the following parameters
will be considered.
Measure Start Time [ps]: it is the starting time of the measurement
window. The available time window is given by the chosen Total Simulated
Time Span (Tsim). Allowed values: min 0; max Tsim; default 0.
Measure End Time [ps]: it is the ending time of the measurement window.
The available time window is given by the chosen Total Simulated Time
Span (Tsim). It must be greater than Measure Start Time. Allowed values:
min 0; max Tsim; default Tsim.
Logical Level: it is the number of bit per symbol of the input signal. Allowed
values: min 1; max -; default 1.
22.12 References
[1] M. C. Jeruchim, Techniques for Estimating the Bit Error Rate in the
Simulation of Digital Communication Systems, IEEE Journal of Selected
Areas in Communications, Vol. SAC-2, no.1, January 1984.
[2] W.Willner, S.Benedetto and L.G. Kazovsky, Optical Communication
Systems.
[3] J.G.Proakis, Digital Communications, New York: McGraw-Hill, 1983 (a
detailed theory on χ2 is given in this book).
[4] A.Humblet and M. Azizoglu “On the bit error rate of lightwave systems
with optical amplifiers”, Journal Lightwave Technol., vol. 9, no. 11, pp.
1576-1582, Nov. 1991.
23 Electrical
Components
Electrical Components
Mathematical
Multiplier of electrical signals.
Advanced
Delay (DELAY), simulates an electrical delay.
Clock Recovery
This category contains several components implementing carrier and clock
recovery subsystems, together with some basic building components such as
Phase Locked Loops (PLL).
Co-Decoder
PCM decoder (RPCMDE), simulates a PCM decoder of the logical signal. The
logical input signal is in absolute value plus sign representation.
PCM encoder (RPCMEN), simulates a PCM encoder of an electrical signal. The
logical output signal is in absolute value plus sign representation.
Parameters
A: multiplier value, it's a real number.
Category: Miscellanea
This component evaluates the sine of the input electrical signal.
Inputs Outputs
Parameters
KFLAG: operational flag.
0 = the output is the time derivative of the input
1 = the output is the time derivative of the input multiplied by the sign of
the input.
Parameters
N: power to which the input signal is raised (integer).
Parameters
A: power to which the input signal is raised.
Note
The input signal must be positive.
Note
The input signal must be positive
Parameters
T_delay [ps]: delay. Allowed values: min 0 ps; max -; default -.
Parameters
Fcentr [THz]; signal central frequency. See section 1.4 for further
information.
Theta [rad]: required phase shift.
Parameters
Fold [THz]: center frequency (section 1.4) of the input signal.
Fnew [THz]: center frequency (section 1.4) of the output signal. Allowed
values: min 0 THz; max 0.4*Simulation Bandwidth; default -.
Parameters
NTYPE: chooses the type of the rectifier. Allowed values: min -2; max +2;
default -; zero is excluded.
Parameters
NTYP: order of the interpolation.
Parameters
N: introduced delay (the filter length is 2*N+1). Allowed values: min 2; max
-; default -.
Parameters
TSAM [ps]: duration of the output pulse. Allowed values: min 3 ·
0.8/Simulation Bandwith; max -; default -.
AMP: amplitude of the output pulse.
Parameters
TSAM [ps]: duration of the output pulse. Allowed values: min
3·0.8/Simulation Bandwidth; max -; default -.
AMP: amplitude of the output pulse.
23.23 Sampler
Component name: SAMPLE
Category: Miscellanea
This component samples an input electrical signal at time instants specified
by a control clock signal. The output signal is set to zero between sampling
instants.
Inputs Outputs
Parameters
Carrier_frequency: see section 1.4 for further information.
N_output_option: switch controlling the output envelope:
0 = the output envelope is raised to the power 1/N
1 = the output is set to the constant value AMP
Output_envelope_amplitude: value of limited output envelope.
m(t )
n
where a(t) may be either or a constant value A.
Inputs Outputs
Parameters
Carrier_frequency: see section 1.4 for further information.
N_output_option: switch controlling the output envelope
0 = the output envelope is raised to N-th power
1 = the output is set to the constant value AMP
Output_envelope_amplitude: value of limited output envelope.
N_power_factor: power of the device.
This block allows the user to simulate nonlinear behavior due to saturation in
electrical component such as amplifiers. Used in combination with a
photodiode it will also accurately simulate photodiode saturation.
Output level
(a.u.)
Linear
response
curve
Saturation level
Saturated
response curve
Linear level
The transfer function is linear below the lin_level level and it saturates
exponentially above it:
in ≤ lin _ level : out = in
in − lin _ level
in > lin _ level : out = sat _ level − delta ⋅ e delta
where:
delta = sat _ level − lin _ level
Both positive or negative values can be used for lin_level and sat_level.
Inputs Outputs
Parameters
lin_level [a.u.]: Output level up until which the response is linear. Allowed
values: min -; max -; default 2.5.
sat_level [a.u]: Output saturation level. Allowed values: min -; max -;
default 5.
E_input
SHOLD NON-LIN DELAY
FILMAT
SHOLD NON-LIN SUM
E_timing_transition E_VCO_input
TIMGEN FILOOP
E_error
E_timing_mid_symbol
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D and G are real constants.
Inputs Outputs
further details
Parameters
Baud_rate [Tbaud/s]: input signal baud rate. Allowed values: any positive
real number; default 10-3 Tbaud/s.
Filter_Option: switch selecting the Matched Filter. Allowed values: min 0;
max 2; default 1.
0 = no filter 1 = rectangular impulse response
2 = sinusoidal impulse response
Nonlinear_Trans_Option: type of non linear transformation. Allowed
values: min 1; max 2; default 1.
1 = absolute value
2 = squaring law
Integration_Order: order of integration (0, 1 or 2) for the loop filter,
integer variable. Allowed values: min 0; max 2; default 1.
Loop_Filter_Par_A: loop filter constant (A in the formulae). Allowed values:
any real number; default 1.
Loop_Filter_Par_B: loop filter constant (B in the formula (3), ignored in
other cases). Allowed values: any real number; default 1.
Loop_Filter_Par_D: loop filter constant (D in form. (1) and (2), ignored in
(3)). Allowed values: any positive real number; default 1.
Loop_Filter_Gain: loop filter gain (G in the formulae). Allowed values: any
positive real number; default 1.
VCO_sens: VCO gain, real variable. Allowed values: any real number;
default 1.
Offset: time offset of the Delay component shown in the diagram at the top
of the page. Allowed values: any real number greater than zero and smaller
than one half of the input signal Baud period; default 0.
FILMAT B
E_VCO_input E_error
E_in_quad
DELAY DELAY
FILMAT B
Timing signals
The Matched filter FILMAT can be selected between the rectangular and the
half-sinusoidal impulse response, or may be suppressed.
The non-linear transformation NON-LIN can be selected between the
absolute-value and square-law.
The loop filter FILOOP may have one of the following transfer functions:
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D and G are real constants.
Inputs Outputs
Parameters
Baud_rate [Tbaud/s]: input signal baud rate. Allowed values: any positive
real number; default 10-3 Tbaud/s.
Filter_Option: switch selecting the Matched Filter. Allowed values: min 0;
max 2; default 1.
0 = no filter
1 = rectangular impulse response
2 = sinusoidal impulse response
Nonlinear_Trans_Option: type of non linear transformation. Allowed
values: min 1; max 2; default 1.
1 = absolute value
2 = squaring law
Integration_Order: order of integration (0, 1 or 2) for the loop filter,
integer variable. Allowed values: min 0; max 2; default 1.
Loop_Filter_Par_A: loop filter constant (A in the formulae). Allowed values:
any real number; default 1.
Loop_Filter_Par_B: loop filter constant (B in the formula (3), ignored in
other cases). Allowed values: any real number; default 1.
Loop_Filter_Par_D: loop filter constant (D in form. (1) and (2), ignored in
(3)). Allowed values: any positive real number; default 1.
Loop_Filter_Gain: loop filter gain (G in the formulae). Allowed values: any
positive real number; default 1.
VCO_sens: VCO gain, real variable. Allowed values: any real number;
default 1.
Offset: time offset of the Delay component shown in the second diagram at
the top of the page. Allowed values: any real number greater than zero and
smaller than one half of the input signal Baud period; default 0.
E_hard_limited
In-phase arm
Mid-phase arm
E_timing_transition
E_VCO_input E_error
TIMGEN FILOOP
E_timing_mid_symbol
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D and G are real constants.
Inputs Outputs
Parameters
Baud_rate [Tbaud/s]: input signal baud rate. Allowed values: any positive
real number; default 10-3 Tbaud/s.
Integration_Order: order of integration (0, 1 or 2) for the loop filter,
integer variable. Allowed values: min 0; max 2; default 1.
Loop_Filter_Par_A: loop filter constant (A in the formulae). Allowed values:
any real number; default 1.
Loop_Filter_Par_B: loop filter constant (B in the formula (3), ignored in
other cases). Allowed values: any real number; default 1.
Loop_Filter_Par_D: loop filter constant (D in form. (1) and (2), ignored in
(3)). Allowed values: any positive real number; default 1.
Loop_Filter_Gain: loop filter gain (G in the formulae). Allowed values: any
positive real number; default 1.
VCO_sens: VCO gain, real variable. Allowed values: any real number;
default 1.
Mid_phase_integration_time: half the Mid-Phase of the integration time.
Allowed values: any real number greater than zero and smaller than one half
of the input signal Baud period; default 0.
E_in_1
B
DELAY DELAY
E_VCO_input E_error
E_in_2
B
E_hard_limited_1 (2)
In-phase arm
Mid-phase arm
Timing signals
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D and G are real constants.
Inputs Outputs
Parameters
Baud_rate: input signal baud rate
Integration_order: switch selecting the Matched Filter:
0 = no filter
1 = rectangular impulse response
2 = sinusoidal impulse response
Loop_Filter_Par_A: loop filter constant (A in the formulae), real variable
Loop_Filter_Par_B: loop filter constant (B in the formula (3), ignored in
other cases), real variable
Loop_Filter_Par_D: loop filter constant (D in form. (1) and (2), ignored in
(3)), real variable
Loop_Filter_Gain: loop filter gain (G in the formulae), real variable
VCO_sens: VCO sensitivity
E_input E_output
NTHPOW PLL2 PHASHF NTHDIV
Parameters
Loop_order: order of the loop.
Carrier_frequency: see section 1.4 for further information.
PLL_natural_frequency: natural frequency of the Second order PLL.
PLL_damping_factor: damping factor of the PLL.
E_phase
x
E_input ZD
pi/2 VCO Filter Combin.
E_VCO
x
E_quad
ZD = 0.25 4 sin (4 f )
The loop filter is a single-pole low pass filter, whose 3 dB bandwidth is set by
the user.
Inputs Outputs
Parameters
N_PSK: number of phases in the CPSK modulation. Allowed values: any
integer value greater or equal to 2; default 4, corresponding to a 4-PSK
carrier recovery.
Filter_Bandwidth: 3-dB bandwidth of the loop filter. Allowed values: any
real number greater than zero, and giving a filter bandwidth inside the
simulation bandwidth, see section 1.4 for further information; default -.
Amp: amplitude of the signal produced by the VCO (zero-to-peak amplitude).
Allowed values: any positive real number; default 1.
VCO_sens: sensitivity of the VCO. Allowed values: any positive real number;
default 1.
E_phase
x
DELAY x
E_input E_clock
VCO Filter Sum Decis.
E_VCO
pi/2 DELAY x
E_quad
x
The Decision algorithm divides the (-π, π) interval in N parts, searches the
interval to which the actual received phase belongs to and produces, as
output, cos(φ) and sin(φ), where φ is the center value of the selected interval.
The Decision process takes place only when a timing pulse is fed to the
component; after each timing pulse, the estimated values of cos(φ) and sin(φ)
are hold until the next timing pulse occurs.
Each arm includes a delay of half the symbol duration, in order to align the
demodulated baseband signal and the phase estimate produced by the
decision algorithm.
The loop filter is a single-pole low pass filter, whose 3 dB bandwidth is set by
the user.
Inputs Outputs
E_clock: clock input signal. See E_VCO: signal produced by the VCO
section 1.6 for further details.
E_phase: base-band demodulated
E_input: electrical input signal signal (in-phase component)
E_quad: base-band demodulated
signals (quadrature component)
Parameters
N_PSK: number of phases in the PSK Modulation. Allowed values: any
integer value greater or equal to 2; default 4, corresponding to a 4-PSK
carrier recovery.
Filter_Bandwidth: 3-dB bandwidth of Loop Filter. Allowed values: any real
number greater than zero, and giving a filter bandwidth inside the simulation
bandwidth, see section 1.4 for further information; default -.
Amp: amplitude of the VCO output signal (zero-to-peak amplitude). Allowed
values: any positive real number; default 1.
VCO_sens: sensitivity of the VCO. Allowed values: any positive real number;
default 1.
VCO_Initial_Frequency [THz]: center frequency of the VCO; it should be
equal (or as close as possible) to the input signal central frequency. Allowed
values: see section 1.4 for further information; default 0.002 THz.
VCO_Initial_Phase [rad]: initial phase of the VCO. Allowed values: any real
number; default 0. rad.
Baud_rate [Tbaud/s]: input signal baud rate. Allowed values: any positive
real number; default 10-3 Tbaud/s.
si = ln (ch(ai ))
where:
ETA
ai = 2 ⋅ c ⋅ xi
B T
and:
ETA is the estimated signal-to-noise ratio
B is the nominal signal level
T is the symbol duration
c is a constant depending on the type of Matched Filter
xi is the sample of the signal at the output of the Matched Filter
The optimum sampling instant within the symbol duration is then determined
as the instant corresponding to the maximum among the accumulated
quantities.
The estimate of the optimum sampling instant is based on a number of
observations (No. of symbols) given as a parameter. Before the first
observation period has expired, no timing signal is produced.
Inputs Outputs
Parameters
Baud_rate: input signal baud rate.
Observation_time: duration of the observation time.
SNR_expected [dB]: expected Signal-to-noise ratio (corresponding to ETA
in the previous formula).
Nominal_input_signal_level: nominal signal input level.
Filter_type: type of Matched Filter:
0 = no matched filter
1 = rectangular impulse response
2 = half-sinusoidal impulse response.
si = ln(ch(a ⋅ xi )) + ln (ch(a ⋅ y i ))
where:
ETA
ai = 2 ⋅ c ⋅ xi
B T
and:
ETA is the estimated signal-to-noise ratio
B is the nominal signal level
T is the symbol duration
c is a constant depending on the type of Matched Filter
xi, yi are the sample of the signals at the output of the Matched Filter for
the in-phase and quadrature components.
The optimum sampling instant within the symbol duration is then determined
as the instant corresponding to the maximum among the accumulated
quantities.
The estimate of the optimum sampling instant is based on a number of
observations (No. of symbols) given as a parameter. Before the first
observation period has expired, no timing signal is produced.
A delay, specified as a parameter, can be introduced between the two
channels, if necessary.
Inputs Outputs
Parameters
Baud_rate: input signal baud rate.
Observation_time: duration of the observation time.
SNR_expected [dB]: expected Signal-to-noise ratio (corresponding to ETA
in the previous formula).
Nominal_input_signal_level: nominal signal input level.
Filter_type: type of Matched Filter:
0 = no matched filter
1 = rectangular impulse response
2 = half-sinusoidal impulse response
Offset_between_channels: the delay introduced on Channel 2.
GAIN
E_input E_multiplier
Multipl. x
E_VCO
VCO
The signal produced by the VCO has a peak amplitude equal to SQRT(2.); the
constant GAIN must be specified as a natural number.
Inputs Outputs
Parameters
Hilbert_delay: delay introduced to perform the Hilbert transform of the
input signal.
Loop_Gain: loop gain.
VCO_frequency: center frequency of the VCO. See section 1.4 for further
informations.
E_input E_multiplier
Multipl. Filter x
E_VCO
VCO
The signal produced by the VCO has a peak amplitude equal to SQRT(2.).
The Filter is a single pole low-pass filter with a user-defined 3-dB bandwidth.
The constants G and B are related to the parameters OMN and DAMP in the
argument list by the following relations:
OMN
G=
2 ⋅ DAMP
OMN ⋅ DAMP
B=
PI
The multiplier operates on the analytic signal associated to the input signal X.
For this purpose the Hilbert transform of the input signal is evaluated using
the component HILBER. The parameter N is the delay (expressed in number
of samples) used by the Hilbert transformer. The lowest frequency
component of the input signal must be larger than I (2 ⋅ DELT ⋅ N ) , thus
the choice of N is based on the spectral properties of the input signal. For
more details see the component HILBER (section 23.20).
Inputs Outputs
Parameters
Hilbert_delay: delay introduced to perform the Hilbert transform of the
input signal.
Loop_bandwidth [rad/ps]: natural frequency of the loop.
Loop_damping_factor: damping factor.
VCO_Frequency [rad/ps]: center frequency of the VCO. See section 1.4 for
further information.
1 + As
H (s ) = G (1)
1 + Ds
1 + As
H (s ) = G (2)
s ⋅ (1 + Ds )
1 + As + Bs 2
H (s ) = G (3)
s2
where A, B, D, and G are real constants.
The PLL order can only take the values 1, 2, and 3.
No check is made on the consistency between the PLL order and the chosen
loop filter transfer function when the PLL order is either 2 or 3.
The VCO rest frequency and the input signal center frequency must be inside
baseband.
The multiplier operates on the analytic signal associated to the input signal X.
For this purpose the Hilbert transform of the input signal is evaluated using
the component HILBER. The parameter N is the delay (expressed in number
of samples) used by the Hilbert transformer. The lowest frequency
component of the input signal must be larger than 1 (2 ⋅ DEL ⋅ N ) , thus the
choice of N is based on the spectral properties of the input signal. For more
details see the component HILBER (section 23.20).
Inputs Outputs
Parameters
Hilbert_Delay: delay introduced to perform the Hilbert transform of the
input signal.
N_PLL_order: order of the PLL (can be 1, 2, or 3), integer variable.
VCO_Frequency: VCO rest frequency (must be outside baseband), real
variable.
VCO_output_amplitude: amplitude. of the VCO out. signal, real var.
VCO_gain: VCO gain, real variable.
N_order_of_integration: order of integration (can be 0, 1, or 2). It gives
the number of poles for s=0 in the loop filter. Integer variable.
Non-lin. E_recovered_clock
E_in_phase
SUM RESON ZCGEN
E_in_quad
E_recovered_sinusoid
Non-lin.
Parameters
Baud_rate: input signal baud rate.
Q_factor: resonator Q-factor.
N_slope_option: if set to +1, the timing pulse is generated when the signal
Z crosses the zero axis with positive slope; if set to -1, the timing pulse is
generated when the signal Z crosses the zero axis with negative slope.
Delay: delay between the effective zero crossing and the generated output
timing pulse.
RECTIF E_recovered_clock
E_in_phase
SUM RESON ZCGEN
E_in_quad
E_recovered_sinusoid
RECTIF
Parameters
N_rectifier_type: chooses the type of the rectifier
2 for Full Wave positive output rectifier
1 for Half Wave positive output rectifier
-2 for Full Wave negative output rectifier
-1 for Half Wave negative output rectifier
Baud_rate: input signal baud rate.
Q_factor: resonator Q-factor.
N_slope: if set to +1, the timing pulse is generated when the signal Z
crosses the zero axis with positive slope; if set to -1, the timing pulse is
generated when the signal Z crosses the zero axis with negative slope.
Time_offset: delay between the effective zero crossing and the generated
output timing pulse.
E_recovered_clock
E_recovered_sinusoid
Parameters
Carrier_frequency: see section 1.4 for further information.
Baud_rate: input signal baud rate.
Q_factor: resonator Q-factor.
N_slope_option: if set to +1, the timing pulse is generated when the signal
Z crosses the zero axis with positive slope; if set to -1, the timing pulse is
generated when the signal Z crosses the zero axis with negative slope.
Time_offset: delay between the effective zero crossing and the generated
output timing pulse; it is expressed in number of sampling intervals and may
be fractional.
MONOSZ E_recovered_clock
E_in_phase
SUM RESON ZCGEN
E_in_quad
E_recovered_sinusoid
MONOSZ
Parameters
Carrier_frequency: see section 1.4 for further information
Baud_rate: input signal baud rate
Q_factor: resonator Q-factor
N_slope_option: if set to +1, the timing pulse is generated when the signal
Z crosses the zero axis with positive slope; if set to -1, the timing pulse is
generated when the signal Z crosses the zero axis with negative slope
Time_offset: delay between the effective zero crossing and the generated
output timing pulse; it is expressed in number of sampling intervals and may
be fractional.
Category: Miscellanea
This component adds jitter on the electrical signal. This is very useful, for
instance, for simulating the clock jitter, which moves the instant when the
signal is sampled. For analysis purposes this effect can be modeled skewing
the electrical signal before the measurement component and considering an
ideal clock not jittered.
This component changes the position of the electrical signal samples as if
were sampled by a jittered clock.
Output Jittered
Electrical Signal
1 / bit-rate
Jittered Clock
On each period the input electrical signal is twisted according to the value of
the jitter signal amplitude multiplied by the jitter factor k_jitter.
Inputs Outputs
Parameters
bit_rate [Gbit/s]: the input signal bit-rate Allowed values: min >0; max -;
default 10.
k_jitter [ps/a.u.]: the jitter factor that is multiplied by the input jitter signal
amplitude to obtain the shifting value. Allowed values: min -; max -; default
1.
sampling_time [ps]: the delay added to the sampling time period (equal to
1/bit-rate). This parameter allows to obtain an higher precision on the applied
jitter. Allowed values: min 0; max -; default 0.
Parameters
Overload_threshold [-]: maximum allowable absolute value for the input
signal. Allowed values: min 1; max -; default 1.
Signal L_output logical level [-]: logical level of output data. Allowed
values: min 2; max -; default 2.
Note
When the absolute value of the input electric signal becomes greater than the
overload threshold an overflow occur. No error is assumed and the converter
behaves as if saturated.
Parameters
Overload_threshold [-]: maximum allowable absolute value for the input
signal to be converted. Allowed values: min 1; max -; default 1.
Signal L_output logical level [-]: logical level of output data. Allowed
values: min 2; max -; default 2.
Note
When the absolute value of the input electric signal becomes greater than the
overload threshold an overflow occur. No error is assumed and the converter
behaves as if saturated.
Parameters
Maximum_output_value: maximum absolute value that will assume the
electrical output signal. Allowed values: min 1; max -; default 1.
Signal L_input logical level [-]: logical level of input data. Allowed values:
min 2; max -; default 2.
Parameters
Maximum_output_value [-]: maximum absolute value that will assume
the electrical output signal. Allowed values: min 1; max -; default 1.
Signal L_input logical level [-]: logical level of input data. Allowed values:
min 2; max -; default 2.
Parameters
Maximum_output_value: it is the maximum output value that the electric
output signal will assume. Allowed values: min 0; max -; default 0.
Compressing_law: it is a compressing law selector. Choosing 0 the A-law
will be applied. Choosing 1 the Mu-law will be applied. Allowed values: min
0; max 1; default 0.
Companding_advantage [dB]: it is the companding advantage factor.
Allowed values: min 0 dB; max -; default 0.
For the A law C = 20*ALOG10(A/(1+ALOG(A))) , for the Mu-law C =
20*ALOG10(Mu/ALOG(1+Mu)).
Note
The conversion is performed by first converting the signal as if quantized
uniformly and then decompressing it with the selected law. The compressing
law parameters A or Mu are evaluated from C with an iterative procedure.
Parameters
Overload_threshold [-]: maximum allowable absolute value for the input
signal. Allowed values: min 0; max -; default -.
Compressing_law [-]: it is a compressing law selector. Choosing 0 the A-
law will be applied. Choosing 1 the Mu-law will be applied. Allowed values:
min 0; max 1; default 0.
Companding_advantage [dB]: it is the companding advantage factor.
Allowed values: min 0 dB; max -; default -.
For the A law C = 20*ALOG10(A/(1+ALOG(A))) , for the Mu-law C =
20*ALOG10(Mu/ALOG(1+Mu)).
Note
The compressing law parameters A or Mu are evaluated from C with an
iterative procedure, whose convergence is guaranteed to be monotone and is
stopped when the absolute error is smaller than 10-2.
Basic
AND
NOT
OR
Exclusive OR (RXOR).
Adder (RSUM), sums of two logical signals. It considers the absolute value
plus signal representation.
Adder (RSUM2), sums of two logical signals. It considers the two’s
complement representation.
Advanced
Parallel to Serial Conversion (PARSEV), converts parallel logical signals into
serial logical signal.
Parallel to Serial Conversion + Line Encoding (PSCLE2), converts parallel
logical signals with user-defined encoding into serial logical signal.
Serial To Parallel Conversion (SEPAR), converts a serial logical signal into
parallel logical signals.
Serial To Parallel Conversion + Line Encoding (SPCON2), converts a serial
logical signal with user-defined encoding into parallel logical signals.
Logical Components
Sign Changer (RCNG), changes the sign of a logical signal. The logical input
signal is assumed to be in absolute value plus sign representation.
Sign Changer (RCNG2), changes the sign of a logical signal. The logical input
signal is assumed to be in two’s complement representation.
Signal Copy (RCHOM4), copies four logical signals into one logical signal.
Signal Copy (RCOPY), copies a component of bits from a logical signal to the
output.
Integer To Logical Converters (RIRC – RIRCX), converts using the binary
representation of the integer value. The logical signal is assumed to be in
absolute value plus sign representation.
24.1.2 NOT
Component name: NOT
Category: Digital Signal Processing Modules
This component does the bitwise logical NOT operation on the logical input
signal.
Inputs Outputs
24.1.3 OR
Component name: OR
Category: Digital Signal Processing Modules
This component places the bitwise logical OR between the two logical input
signals.
Inputs Outputs
24.1.4 Exclusive-OR
Component name: RXOR
Category: Digital Signal Processing Modules
This component places the bitwise logical EXCLUSIVE-OR between the two
logical input signals.
Inputs Outputs
Note
An error condition will be issued if the levels of the three logical signal are not equal.
Note
An error condition will be assumed if the levels of the three logical signals is not the
same. An error will be assumed only if overflow occur: the logical output signal level is
too small to represent the result.
Note
An error condition will be assumed if the levels of the three logical signals is
not the same. An error will be assumed only if overflow occur: the logical
output signal level is too small to represent the result.
LinA: first logical input signal Lout: resulting logical output signal
LinB: second logical input signal
Parameters
Rb [Tb/s]: Baud rate. Allowed values: min 0 Tb/s; max -; default -.
TSHIFT [ps]: instant within the symbol interval at which the sum is
executed. Allowed values: min 0 ps; max 1/Rb; default -.
Parameters
Level: logical level of the input logical signal. Allowed values: min 1; max -;
default 1.
T_delay [ps]: Delay applied to the input logical signal. Allowed values: min
0; max -; default 0.
CLOCK DATA Q Q
↑ 0 0 1
↑ 1 1 0
↓ X Q Q
X X Q Q
X = don’t care
↑ = level changes from 0 to 1
↓ = level changes from 1 to 0
Inputs Outputs
Parameters
Logical level [-]: logical level of input data. Allowed values: min 1; max -; default 1.
↓ X Qt-1 Qt-1
X X Qt-1 Qt-1
X = don’t care
↑ = level changes from 0 to 1
↓ = level changes from 1 to 0
Inputs Outputs
Parameters
Signal Clock Logical level [-]: logical level of the clock input data. Its value
must be kept 1.
Signal T Logical level [-]: logical level of the T input data. Its value must
be equal to the other Q and Q\ logical level. Allowed values: min 1; max -;
default 1.
Signal Q Logical level [-]: logical level of the Q output data. Its value must
be equal to the other T and Q\ logical level. Allowed values: min 1; max -;
default 1.
Signal Q_ Logical level [-]: logical level of the Q\ output data. Its value
must be equal to the other T and Q logical level. Allowed values: min 1;
max -; default 1.
Parameters
Baud_rate: baud rate referred to the output logical signal. Allowed values:
min 0; max Simulation Bandwidth; default -.
Delay: transmission delay between the source of information bits and the
component. Allowed values: min 0; max -; default 0.
Note
The baud rate of the input logical signal MUST be (KI KO ) ⋅ NS , where KI
and KO are the size of input and output logical signals respectively.
Parameters
Input_baud_rate: it is the baud rate referred to the input logical signal.
Allowed values: min 0; max Simulation bandwidth; default 0.
Output_bit_rate: it is the baud rate referred to the output logical signal.
Allowed values: min Input_baud_rate* Input_level; max Input_baud_rate*
Input_level; default -.
Coding_selector: it is a flag to selector the coding law. Allowed values: min
0; max 7; default 0.
Note
An error condition will be assumed and an error message issued if the output
baud rate is not the input baud rate multiplied by KI, where KI is the level of
the input logical signal.
Parameters
Baud_rate: it is the baud rate referred to the input logical signal. Allowed
values: min 0; max Simulation bandwidth; default -.
Delay: transmission delay between the source of information bits and the
component. Allowed values: min 0; max -; default 0.
Note
The baud rate of the output register is (KI/KO)*input_baud_rate, where KI
and KO are the size of input and output logical signal respectively.
Parameters
Input_bit_rate: it is the baud rate referred to the input logical signal.
Allowed values: min Output_baud_rate*Output_level; max
Output_baud_rate*Output_level; default -.
Output_baud_rate: it is the baud rate referred to the output logical signal.
Allowed values: min 0; max Simulation Bandwidth; default -.
Coding_selector: it is a flag to selector the coding law. Allowed values: min
0; max 7; default 0.
Note
An error condition will be assumed and an error message issued if the input
baud rate is not the output baud rate multiplied by KO, where KO is the level
of the output signal.
Note
An error condition will be assumed if the logical signal levels are different.
Note
An error condition will be assumed if the logical signal levels are different.
If FLAG = 1
1 2 ND 1 2 NC 1 2 NB 1 2 NA 0 …... 0
If FLAG = -1 NN = NY-ND-NC-NB-NA
NN+1 NN+ND NN+ND+NC NN+ND+NC+NB
1 ND NY
0 …. 0 1 2 ND 1 2 NC 1 2 NB 1 2 NA
Inputs Outputs
Parameters
Flag: flag to choose upper or lower part of output logical signal. Allowed
values: min -1; max +1; default +1.
Note
An error condition will be assumed when the sum of the logic input signal
levels is greater than the logic output signal level.
If FLAG = 1 NI+NBIT-1
L_input 1 NI NX
L_output 1 NO NY
NO+NBIT-1
If FLAG = -1
NI+NBIT-1
L_input 1 NI NX
L_output NY NO 1
NO+NBIT-1
Inputs Outputs
Parameters
Start in input: starting position of the component of bits inside the logical
input signal.
NBIT: number of bits to be copied.
Start in output: starting position inside the logical output signal.
Flag: direct or inverse copy. Allowed values: min -1; max +1; default +1.
Parameters
Number: Integer number to be converted in binary representation.
Note
An error condition will be assumed if the integer parameter can not be
represented within the logical output signal: the signal level is too small.
Parameters
Number: integer number to be converted in binary representation.
Note
An error condition will be assumed if the integer parameter can not be
represented within the logical output signal: the signal level is too small.
Parameters
Number: integer number to be converted in binary representation.
Operating levels: effective number of level to be used in the output logical
signal.
Note
An error condition will be assumed if the integer parameter can not be
represented within the logical output signal: the number of operating levels is
too small.
Parameters
Number: integer number to be converted in binary representation.
Operating levels: effective number of level to be used in the output logical
signal.
Note
An error condition will be assumed if the integer parameter can not be
represented within the logical output signal: the number of operating levels is
too small.
Note
The levels of logical signals can be different. An error will be assumed only if
overflow occur. Let NA, NB be the logical input signals levels and NO the
logical output signal level. OVERFLOW CAN NOT occur if NO >= NA+NB-1.
The user is advised to use levels so that NO >= NA+NA-1 as far as it is
possible.
Note
The levels of the logical signals can be different. An error will be assumed
only if overflow occur. Let NA, NB be the logical input signals levels and NO
the logical output signal level. OVERFLOW CAN NOT occur if NO >= NA+NB-
1. The user is advised to use levels so that NO >= NA+NA-1 as far as it is
possible.
Parameters
Nshift: the absolute value represents the number of shift to be applied, the
sign the direction. Plus is for a right shift, minus for a left shift. Allowed
values: -NA,-NA+1,...,0, ...,NA-1,NA; default +1. Where NA are the levels of
the input logical signal.
Note
An error condition will be issued if the levels of the two logical signal are not
equal.
Parameters
Nshift: the absolute value represents the number of shift to be applied, the
sign the direction. Plus is for a right shift, minus for a left shift. Allowed
values: -NA,-NA+1,...,0, ...,NA-1,NA; default +1. Where NA are the levels of
the input logical signal.
Note
An error condition will be issued if the levels of the two logical signal are not
equal.
1 1 NX = Level of L_input
NA = Level of L_output_1
2 2 NB = Level of L_output_2
NX = NA+NB
L_output_2
NB+1 NB NB
1
L_output_1
NX NA
Inputs Outputs
Note
The level of the logical input signal NX must be the sum of logical output
signal levels (NA and NB), i.e. NX = NA+NB.
1 1 NX = Level of L_input
NA = Level of L_output_1
NB = Level of L_output_2
NC = Level of L_output_3
L_output_3
NX = NA+NB+NC
NC+1 NC NC
1
L_output_2
NC+NB
NB
NC+NB+1
1
L_output_1
NX NA
Inputs Outputs
Note
The level of the logical input signal NX must be the sum of logical output
signal levels (NA, NB and NC), i.e. NX = NA+NB+NC.
1 1 NX = Level of L_input
NA = Level of L_output_1
NB = Level of L_output_2
NC = Level of L_output_3
L_output_4 ND = Level of L_output_4
NX = NA+NB+NC+ND
ND+1 ND ND
L_output_3
ND+NC
NC
ND+NC+1
1
ND+NC+NB L_output_2
ND+NC+NB+1 NB
L_output_1
NX NA
Inputs Outputs
Note
The level of the logical input signal NX must be the sum of logical output
signal levels (NA, NB, NC and ND), i.e. NX = NA+NB+NC+ND.
Parameters
Rb [Tb/s]: Baud rate. Allowed values: min 0 Tb/s; max 0.6*Simulation
Bandwith; default -.
TSYNC [ps]: Time instant within the symbol period in which the bits are
sampled. Allowed values: min 0 ps; max 1/Rb; default -.
NDEL [ps]: Delay expressed in number of symbols. Allowed values: min 0
ps; max -; default -.
This category groups several blocks which have been maintained only for
compatibility with older OptSim version, in order to allow projects developed
with previous version to be loaded also in the current version. Anyway, we
strongly encourage avoiding the use of these components in new projects,
and using the new LIB components instead.
Ideal clock generator (CLOCK2).
Parameters
Clock_Relative_delay [ps]: clock offset.
Parameters
Interval_between_Pulses [ps]:period of the Delta train.
Amplitude [V]: peak amplitude of the generates pulse.
First_Pulse_Delay [ps]: pulses time offset.
Parameters
Period [ps]: waveform period.
Initial_Time_Offset [ps]: initial waveform offset.
Amplitude_High [V]: waveform amplitude maximum value.
Amplitude_Low [V]: waveform amplitude minimum value.
Parameters
Period [ps]: waveform period.
Initial_Time_Offset [ps]: initial waveform offset.
Slope [V/s]: slope of the sawtooth shape.
Parameters
Period [ps]: waveform period.
Initial_Time_Offset [ps]: initial waveform offset.
Duty_cycle: square wave duty cycle.
Amplitude_High [V]: waveform amplitude maximum value.
Amplitude_Low [V]: waveform amplitude minimum value.
Parameters
Random_Generator_Seed: Integer number used as a seed to start random
number generation.
Mean [V]: mean value of the generated Gaussian distribution.
Standard_Deviation [V]: standard deviation value of the generated
Gaussian distribution.
Parameters
Random_Generator_Seed: integer number used as a seed to start
pseudorandom bit generation.
Period [ps]: bit period.
Probability_1: any value between 0 and 1.
26 Compound
Components
DQPSK Transmitter
RZ-DQPSK Transmitter
• MZM_phase
This component models a Mach-Zehnder modulator used to create the
transmitter’s phase-modulated output signal. It is configured via the
parameters MZM_Vpi – Vπ of the modulator; MZM_Von – the
modulator’s offset voltage; MZM_excess_loss_dB – the modulator’s
insertion loss; MZM_extinction_ratio_dB – the modulator’s extinction
ratio; and MZM_chirp_factor – the modulator’s chirp factor.
Inputs Outputs
Parameters
Vdrive_lo [Volts]: Minimum value of the driver electrical signal (binary zero
value). Allowed values: min -; max -; default 5.0.
Vdrive_hi [Volts]: Maximum value of the driver electrical signal (binary one
value). Allowed values: min -; max -; default –5.0.
filter_poles: Order of the low-pass Bessel filter. Allowed values: min 1; max
10; default 1.
filter_BW_GHz [GHz]: Bandwidth of the low-pass Bessel filter. Allowed
values: min 0; max BWVBS/4; default 10.0.
wavelength_nm [nm]: Wavelength of the cw optical source. Allowed
values: min flowSPT; max fupSPT; default 1550.0.
power_dBm [dBm]: Output power of the cw optical source. Allowed values:
min -; max -; default 0.0.
MZM_Vpi [Volts]: Vπ of the Mach-Zehnder Modulator. Allowed values: min
0; max -; default 5.0.
MZM_Von [Volts]: Offset voltage of the Mach-Zehnder Modulator. Allowed
values: min -; max -; default 5.0.
MZM_excess_loss_dB [dB]: Insertion loss of the Mach-Zehnder Modulator.
Allowed values: min 0; max -; default 0.0.
MZM_extinction_ratio_dB [dB]: Extinction ratio of the Mach-Zehnder
Modulator. Allowed values: min 0; max -; default 1000.0.
MZM_chirp_factor: Chirp factor of the Mach-Zehnder Modulator. Allowed
values: min -; max -; default 0.0.
Inputs Outputs
Parameters
Vdrive_lo [Volts]: Minimum value of the driver electrical signal (binary zero
value). Allowed values: min -; max -; default 5.0.
Vdrive_hi [Volts]: Maximum value of the driver electrical signal (binary one
value). Allowed values: min -; max -; default –5.0.
filter_poles: Order of the low-pass Bessel filter. Allowed values: min 1; max
10; default 1.
filter_BW_GHz [GHz]: Bandwidth of the low-pass Bessel filter. Allowed
values: min 0; max BWVBS/4; default 10.0.
wavelength_nm [nm]: Wavelength of the cw optical source. Allowed
values: min flowSPT; max fupSPT; default 1550.0.
power_dBm [dBm]: Output power of the cw optical source. Allowed values:
min -; max -; default 0.0.
MZM_Vpi [Volts]: Vπ of the Mach-Zehnder Modulator. Allowed values: min
0; max -; default 5.0.
signal value for binary zeroes; and Vdrive_hi – the signal value for
binary ones.
• LPF_TX_P and LPF_TX_Q
These components implement low-pass Bessel filters for modeling the
non-ideal characteristics of the binary-to-electrical conversion. They
are configured via the parameters filter_poles – the order of the Bessel
filters; and filter_BW_GHz – the filter bandwidths.
• Laser_source
This component models the transmitter’s cw laser source. It is
configured via the parameters wavelength_nm – the laser’s center
wavelength; and power_dBm – the cw laser power.
• DQPSK_mod
This component implements the DQPSK modulation portion of the
transmitter. It is configured via the parameters MZM_Vpi – Vπ of the
various Mach-Zehnder modulators (MZMs) inside this component;
MZM_excess_loss_dB – the MZMs’ insertion losses;
MZM_extinction_ratio_dB – the MZMs’ extinction ratios; PM_Vpi – Vπ of
the phase modulator used to generate the additional phase shift
required for the quadrature signal; and PM_excess_loss – insertion
loss of the phase modulator. The component Phase_shift_bias provides
the Vπ/2 input required by the phase modulator.
Inputs Outputs
Parameters
Vdrive_lo [Volts]: Minimum value of the driver electrical signal (binary zero
value). Allowed values: min -; max -; default 5.0.
Vdrive_hi [Volts]: Maximum value of the driver electrical signal (binary one
value). Allowed values: min -; max -; default –5.0.
filter_poles: Order of the low-pass Bessel filter. Allowed values: min 1; max
10; default 1.
filter_BW_GHz [GHz]: Bandwidth of the low-pass Bessel filter. Allowed
values: min 0; max BWVBS/4; default 10.0.
wavelength_nm [nm]: Wavelength of the cw optical source. Allowed
values: min flowSPT; max fupSPT; default 1550.0.
Inputs Outputs
Parameters
Vdrive_lo [Volts]: Minimum value of the driver electrical signal (binary zero
value). Allowed values: min -; max -; default 5.0.
Vdrive_hi [Volts]: Maximum value of the driver electrical signal (binary one
value). Allowed values: min -; max -; default –5.0.
filter_poles: Order of the low-pass Bessel filter. Allowed values: min 1; max
10; default 1.
filter_BW_GHz [GHz]: Bandwidth of the low-pass Bessel filter. Allowed
values: min 0; max BWVBS/4; default 10.0.
wavelength_nm [nm]: Wavelength of the cw optical source. Allowed
values: min flowSPT; max fupSPT; default 1550.0.
power_dBm [dBm]: Output power of the cw optical source. Allowed values:
min -; max -; default 0.0.
MZM_Vpi [Volts]: Vπ of the Mach-Zehnder Modulator. Allowed values: min
0; max -; default 5.0.
MZM_excess_loss_dB [dB]: Insertion loss of the Mach-Zehnder Modulator.
Allowed values: min 0; max -; default 0.0.
MZM_extinction_ratio_dB [dB]: Extinction ratio of the Mach-Zehnder
Modulator. Allowed values: min 0; max -; default 1000.0.
PM_Vpi [Volts]: Vπ of the Phase Modulator. Allowed values: min 0; max -;
default 5.0.
PM_excess_loss [dB]: Insertion loss of the Phase Modulator. Allowed
values: min 0; max -; default 0.0.
Inputs Outputs
Parameters
ofilter_order: Order of the Super-Gaussian optical filter. Allowed values:
min 1; max -; default 1.
ofilter_BW_GHz [GHz]: Bandwidth of the Super-Gaussian optical filter.
Allowed values: min 0; max BWVBS; default: 50.0.
channel_frequency_THz [THz]: Frequency of the optical channel to detect.
Allowed values: min flowSPT; max fupSPT; default: 194.0.
MZI_bitrate_Gbps [Gbps]: Bit-rate of the optical channel to detect.
Allowed values: min -; max -; default: 10.0.
Inputs Outputs
Parameters
ofilter_rolloff: Rolloff parameter of the Raised-Cosine optical filter. Allowed
values: min 0; max 1; default 0.5.
ofilter_alpha: Alpha parameter of the Raised-Cosine optical filter. Allowed
values: min 0; max 1; default 1.0.
ofilter_BW_GHz [GHz]: Bandwidth of the Raised-Cosine optical filter.
Allowed values: min 0; max -; default: 50.0.
channel_frequency_THz [THz]: Frequency of the optical channel to detect.
Allowed values: min flowSPT; max fupSPT; default: 194.0.
MZI_bitrate_Gbps [Gbps]: Bit-rate of the optical channel to detect.
Allowed values: min -; max -; default: 10.0.
27 Additional
Components
This category groups several models and features that have been added later
on to the standard components and compound components.
Iteration Loop
Ideal Electronic Dispersion Compensator (EDC)
M-QAM Modulator
M-QAM Demodulator
Quadrature Mix
FFT OFDM
IFFT OFDM
Delay 2 Signals
ln 2
−
σf 2
H( f ) = e 2
where σ is the rms impulse response width and σ = 0.1874/Bw where Bw
is the 3 dB optical bandwidth. The total rms impulse response width s is given
by
σ 2 = σ 2 mat + σ 2 mod
where the pulse broadening σ mat due to material dispersion and the pulse
σ mod = σ src LD
1
σ mat = γ cutback
1
Bim
L
where D is the fiber dispersion, L is the fiber length, σ src is the linewidth of
the source, Bim is the intermodal bandwidth of the fiber, and γ cutback is the
cutback factor that takes into account the mode coupling and mixing effects.
In addition to modeling the fiber itself, the linear multimode fiber model block
also models the connector and coupling losses at each end of the fiber ribbon
in the optical bus. These losses are specified explicitly as the connector loss
parameter.
Inputs Outputs
Optical Optical
Parameters
length [km]: length of fiber [Km]. Allowed values: min 0+; max -;
default 0.1.
atten [dB/km]: attenuation of fiber. Allowed values: min 0; max -;
default 14.4.
disp [ps/nm/Km]: dispersion of fiber. Allowed values: min -; max -;
default 120.
linewidth [MHz]: source linewidth. Allowed values: min 0+; max -;
default 10.
band_im [MHz*km]: intermodal bandwidth of fiber. Allowed values:
min 0+; max -; default 160.
gamma_cutback [-]: cutback gamma parameter. Allowed values: min 0+;
max -; default 0.75.
connector_loss [dB]: optical loss at connectors. Allowed values: min 0;
max -; default 0.
References
[1] S. D. Personick, "Baseband linearity and equalization in fiber optic digital
communication systems," Bell Syst. Tech. J., vol. 52, pp. 1175-1195,
Sep. 1973.
[2] D. G. Duff, "Computer-aided design of digital lightwave systems," IEEE
Journal on Selected Areas in Communications, vol. SAC-2, no. 1, pp. 171-
185, January 1984.
Now RSoft Design Group’s most prominent tools in the system and device
levels, OptSim and BeamPROP, can be run together in a powerful co-
simulation.
Users may incorporate optoelectronic components modeled in BeamPROP or
another device-level tool in an OptSim simulation. This unique multi-level
simulation capability provides the best accuracy and efficiency in system
performance analysis involving complex designs.
The BEAMPROP optical device library block is used to represent any optical
WDM component designed and modeled at the device level in BeamPROP.
When the OptSim simulation reaches the BEAMPROP library block, the
specified BeamPROP simulation is launched. When a BeamPROP script is
called from OptSim, device-level simulations are performed over the specified
wavelength range to compute the transfer functions for each of the
input/output port pairs. A data file is then generated to provide OptSim with
these results for its system-level simulation of the link.
If the BeamPROP simulation results have already been computed, OptSim
uses the results from BeamPROP’s output file directly. In essence, this file
specifies a set of wavelength-dependent transfer functions for each of the
input/output port pairs.
At this stage, each input/output port pair must be represented by an
individual BEAMPROP block in OptSim.
Inputs Outputs
Optical Optical
Parameters
BeamPROP_dat_File_Name: Name of BeamPROP output file to be used for
the simulation.
In_port: Designated input port for components with multiple (n) input ports.
Allowed values: min 1; max n; default 1.
Out_port: Designated output port for components with multiple (m) output
ports. Allowed values: min 1; max m; default 1.
BeamPROP_scr_File_Name: Name of BeamPROP script to be used for the
co-simulation.
The first row contains the number of input ports (n), followed by the number
of output ports (m), the number of wavelengths specified in the left column,
and the lowest and highest specified wavelength. This file is the output
created by a BeamPROP simulation based on the structure of the waveguide
simulated.
If the BeamPROP output file is not available, the OptSim simulation will pause
while launching the simulation of the specified BeamPROP script. When the
simulation is finished, the results will be passed to OptSim and the OptSim
simulation will continue. If the output results from BeamPROP are already
there, OptSim will use the results directly without launching BeamPROP.
∑c e
− j 2πfnTeq
H eq ( f ) = n (1)
n=− N
where 2N+1 is the total number of taps and Teq is the time delay between the taps.
• ASE noise nASE(t), introduced by the optical amplifiers (EDFAs). It is modeled as additive
white Gaussian noise, characterized by its one-sided power spectral density N0.
• Thermal noise nsh(t) introduced by the electrical amplifier. It is modeled as additive white
Gaussian noise. The variance of noise [mV2] on the decision variable is equal to:
N0+∞
σ th2 = Rt2 ∫
2
H L ( f ) df (2)
−∞ 2
where HL(f) is the transfer function of the post-detection filter, N0/2 is the “input-referred”
bilateral power spectral density [(pA)2/Hz] and Rt is the overall RX transfer function, which
is generally a transimpedance [Ohm] that transforms the input current [mA] in an output
voltage [mV] 1.
• Shot noise nsh(t) introduced by photodetectors, modeled as a non stationary additive
white Gaussian random process with zero mean and signal-dependent variance. The
variance of shot noise on the decision variable is equal to [4]:
±
H AMZ (f)=
1
2
[
1 ± γ e j (2πfTAMZ +δϕ ) ] (4)
where TAMZ is the optical (macroscopic) path difference within the interferometer (ideally equal
to the inverse of the symbol rate), δϕ is a phase-mismatch due to an additional microscopic
path difference inside the interferometer and γ <1 represents a finite extinction ratio. The ±
sign reflects the two interferometer output ports. The AMZ is then followed by a balanced
photo-detector (BPD) device.
1
Since the overall RX performance does not depend on the value of Rt, in the
simulation program it is assumed to be equal to 1 and neglected.
The implemented algorithm considers the following possible imperfections of the receiver
components (see [3-5]):
• AMZ frequency offset ∆f, corresponding to the detuning between the laser frequency and
the Asymmetric Mach-Zehnder central frequency. It is related to the phase mismatch δϕ
through the relationship:
δϕ
∆f = RS (5)
2π
where Rs is the symbol rate (equal to the bit-rate in case of binary modulation).
• AMZ extinction ratio, related to the parameter γ through the following relationship:
(1 + γ )2
ε dB = 10 ⋅ log10 2
(6)
(1 − γ )
• AMZ delay error, defined as the difference between the actual AMZ delay TAMZ and the
symbol time T=1/RS.
• BPD phase imbalance, defined as the propagation delay difference ∆τ between the two
arms of the BPD.
• BPD amplitude imbalance β, due to the difference between the responsivity values of the
two photodetectors:
R1G1 − R2G2
β= (7)
R1G1 + R2G2
The considered noise sources are the same as for the direct-detection receiver.
3. The overall BER at a given decision threshold sth and sampling instant ts is evaluated as:
N
1
BER( sth , t s ) =
N
∑ P (s
k =1
0 ,1
e ,k th , t s + kT ) (8)
4. Joint minimization of (8) over the couple (sth,ts) is performed to obtain the BER at
optimum threshold and sampling instant.
v (t ) [
= F −1 {R1G1 E ( f ) H o ( f ) * E * ( − f ) H o* ( − f ) H L ( f )} = ]
(12)
= ∫∫ E ( f 1 ) K DD ( f 1 , f 2 )E * ( f 2 ) exp[ j 2π ( f 1 − f 2 )t ]df 1 df 2
for the direct detection receiver, and as:
v (t ) [
= F −1 {R1G1 E ( f ) H o ( f ) H AMZ
+ +*
( f ) * E * ( − f ) H o* ( − f ) H AMZ (− f ) H L ( f ) ]
[ −
− R2G2 E ( f ) H o ( f ) H AMZ −*
( f ) * E * ( − f ) H o* ( − f ) H AMZ ( − f ) H L ( f )}= ] (13)
K DD ( f 1 , f 2 ) = R1G1 H o ( f 1 ) H L ( f1 − f 2 ) H o* ( − f 2 ) (14)
v (t ) = ∑ bi (t ) + n i (t ) τ i
2
(16)
i
where bi(t) and vi(t) are the coefficients of the series expansion of the Fourier transform S(f)
and NASE(f) of the noiseless received signal s(t) and of the ASE noise random process nASE(t),
respectively :
where K(f1,f2) is as in (14) and (15) for direct detection and differential receivers respectively.
Using numerical integration algorithms (see [8], Chapter 4), the solution of equation (18) can
be reduced to the eigenvalue and eigenvector problem for an Hermitian matrix:
f MAX 2k
i−k
where x i = f MAX . The integration interval [-fMAX, fMAX] must be chosen in a way that the
k
integrand evaluated in f >fMAX is sufficiently small and does not affect the result anymore2. The
2
The criterion used in the simulation program consists in choosing fMAX such that both following
conditions are satisfied: |Ho(±fMAX)|<max{|Ho(f)}/100 (i.e. the extreme values in the matrix K are
sufficiently small) and fMAX < 0.25 k/TAMZ (i.e. one period of the AMZ transfer function is represented by
at least 5 frequency points).
choice of the number of integration points (2k+1) is a compromise between accuracy and
computation complexity3.
Equation (18) can thus be rewritten as:
2k
∑φ
i =0
m ( x i ) K ( x n , x i ) = τ mφ m ( x n ) n, m = 0,...,2k (20)
with Ki,j=K(xi,xj), ϕij=ϕj(xi), τij=τiδi,j. Since K is Hermitian, the problem (22) is straightforward
to solve using the Jacobi routines in [8], Chapter 11.
Since nASE(t) is a white random process, it can be shown that the noise expansion coefficients
ni(t) are statistical independent Gaussian random variables with zero mean and variance N0τi.
Thus (16) represents a weighted sum of nonzero-mean squared Gaussian random variables,
whose moment generating function (MGF) is equal to [5]:
z bi 2 τ i
exp −
n val −1
1 + zτ i N 0
h(t, z ) = ∏ (22)
i =0 (1 + zτ i N 0 )M
with M=1 or 2 for single or double polarization representation, respectively. Ideally, the
product in (22) should span from 0 to ∞ . In practice, only a finite number nval of eigenvalues
give a non-negligible contribution to the MGF evaluation4.
If also additive electrical noise sources are present (like shot or thermal noise), which are
statistically independent from the optical ASE noise, the overall moment generating function of
the decision variable can be evaluated by multiplying (22) by the MGF of a Gaussian random
process with zero mean and variance σ2(t)=(σsh(t))2+(σth)2, i.e.:
1
hel (t , z ) = exp − σ 2 (t ) z 2 (23)
2
Using the steepest descent approximation method [7], the probability of error can be
evaluated as:
3
In the simulation program, a value of k equal to 50 has been adopted. It has been verified that, choosing
k as high as 100, would yield almost coincident results in most practical cases.
4
It has been verified that choosing nval as high as 30 is more than sufficient in all practical cases, thus we
adopted it in the simulation program.
exp[θ (t , z 0 )]
∞
Pe0,k ( sth , t ) = ∫ PDF (s, t ) ds ≈
sth
k
2πθ " (t , z 0 )
(24)
exp[θ (t , z1 )]
sth
1 2 2
exp[− zs th ]h(t ,− z ) exp − 2 σ (t ) z
θ (t , z ) = ln (25)
z
θ ” is the second derivative of θ with respect to z and z0 and z1 are respectively the positive
and negative roots of the equation:
∂θ (t , z )
ϑ ' (t , z ) = =0 (26)
∂z
Equation (26) can be efficiently solved using the bisection method described in [8], Chapter 9.
Input signal
Noiseless optical signal s(t) at the input of the receiver, obtained through a VBS noiseless
simulation (single or double polarization).
Parameters
Modulation format (integer, range [1,4], default 1)
1: IMDD or Duobinary (the receiver is as in Fig. 1)
2: DPSK or other binary differential modulation format (the receiver is as in Fig. 2)
3: DQPSK, upper arm (the receiver is as in Fig. 3)
4: DQPSK, lower arm (the receiver is as in Fig. 3)
Samples per bit Ns (integer, range [1,Inf), default 40)
Number of samples per bit of the input optical signal. In general, it should be as high as 30 for
reliable estimation. If it is greater than 60, than sub-sampling is applied.
Number of bits Nb (integer, range [1,Inf), default 128)
Number of bits used for the BER evaluation (usually between 64 and 512, typically 128). The
total number of simulated bits has to be equal to at least Nb+10.
Bit rate RB (real, range (0,Inf), default 10., unit Gbit/s)
Bit rate [Gbit/s] used in the simulation.
N0+∞
σ th2 = ∫
2
H L ( f ) df
−∞ 2
Photodiode 1 avalanche gain G1 (real, range [1.,Inf), default 1., unit linear)
Avalanche gain of APD (linear). If a PIN is used, it must be set to 1.
Photodiode 1 excess noise factor x1 (only for APD) (real, range [0.,Inf), default 0. )
Excess noise factor of APD.
Photodiode 2 responsivity R2 (only for differential receivers) (real, range (0.,Inf),
default 1., unit A/W)
Responsivity of the lower arm photodiode.
Photodiode 2 avalanche gain G2 (only for differential receivers) (real, range [1.,Inf),
default 1., unit linear)
Avalanche gain of APD (linear) in the lower arm. If a PIN is used, it must be set to 1.
Photodiode 2 excess noise factor x2 (only for lower arm APD in differential receivers)
(real, range [0.,Inf), default 0. )
Excess noise factor of APD.
Optical filter type (kopt) (integer, range [-2,10], default 2)
-2: From file ‘optfil.DAT’ (baseband frequency in THz – double sided, square amplitude
in linear scale, phase in radians, at the beginning there can be comment lines starting
with #, and the data file has to be placed in the same folder as the project file or in
the result directory xv_sim_<project name>)
-1: Matched (rectangular in time)
0: No filter
1≤n≤8 : Supergaussian of order n
9: Lorentzian
n=10: Raised cosine
Optical filter bandwidth (real, range (0.,Inf), default 30., unit GHz)
If kopt≥0: -3dB optical filter bandwidth in GHz
If kopt=-1: inverse of the time duration of the rectangular pulse (in GHz)
If kopt=-2: scaling factor, i.e. the real number that specifies the expansion/stretch
factor in the frequency scale of the transfer function stored in the data file. This
parameter should be set to 1. when scaling is not required.
Optical filter roll-off (ony for kopt =10) (real, range [0.,1.], default 0.5)
Roll-off factor of the raised-cosine optical filter.
Electrical filter type (kel) (integer, range [-2,6], default 5)
-2: from file ‘elfil.DAT’ (frequency in GHz - only positive, square amplitude in linear
scale, phase in radians) that resides in the same folder as the project file or in the
result directory xv_sim_<project name>)
-1: Matched (rectangular in time)
0: No filter
1: Lorentzian
2≤n≤6 : Bessel of order n
Electrical filter bandwidth (real, range (0.,Inf), default 10., unit GHz)
If kel≥1: -3dB optical filter bandwidth in GHz
If kel=-1: inverse of the time duration of the rectangular pulse (in GHz)
If kel=-2: scaling factor, i.e. the real number that specifies the expansion/stretch
factor in the frequency scale of the transfer function stored in the data file. This
parameter should be set to 1. when scaling is not required.
AMZ frequency offset (real, range [-50,.50.], default 0.)
Detuning ∆f between the laser frequency and the Asymmetric Mach-Zehnder central frequency
(expressed in %RB).
AMZ extinction ratio (real, range (0.,Inf), default 1000., unit dB)
Extinction ratio ε in dB of AMZ.
BPD phase imbalance (real, range (-Inf,Inf.), default 0.)
Propagation delay difference ∆τ between the two arms of the BPD (expressed in %T, with
T=1/RS).
AMZ delay mismatch (real, range (-Inf,Inf.), default 0.)
AMZ delay error, defined as:
δT TAMZ − T
⋅ 100 = ⋅ 100
T T
where TAMZ is the delay introduced by the AMZ and T is the inverse of the symbol rate.
∑c
− j 2πfnTeq
H( f ) = N e
n=− N
References
[1] G. Bosco and R. Gaudino, “Towards new semi-analytical techniques for BER estimation in
optical system simulation”, National Fiber Optics Engineers Conference NFOEC 2000,
Denver (CO), Tuesday Sess. E1, USA, Aug. 2000.
[2] J.S. Lee and C.S. Shim, “Bit error rate analysis of optically preamplified receivers using an
eigenfunction expansion method in optical frequency domain”, IEEE Journal of Lightwave
Technology, vol. 12, pp. 1224-1229, 1994.
[3] A.H. Gnauck, P.J. Winzer, “Optical Phase-Shift-Keyed Transmission,” IEEE Journal of
Lightwave Technology, vol. 23, n. 1, pp. 115-130, Jan. 2005.
[4] L. Kazovsky, S. Benedetto and A. Willner, Optical Fiber Communication Systems, Artech
House, 1996.
[5] G. Bosco and P. Poggiolini, “On the Joint Effect of Receiver Impairments on Direct-
Detection DQPSK systems”, to be published in IEEE Journal of Lightwave Technology in
2006.
[6] A.M. Mathai, S.B. Prevost, Quadratic forms in random variables, New York, Marcel
Dekker, 1992, Chapter 3.
[7] C. W. Helstrom, “Distribution of the Filtered Output of a Quadratic Rectifier Computed by
Numerical Contour Integration”, IEEE Trans. on Inform. Theory, vol.IT-32, no. 4, Jul.
1986, pp. 450-463
[8] W.H. Press, S.A. Teukolsky, W.T. Vetterling, Numerical recipes in C, Cambridge University
Press, 1992.
Inputs Outputs
Optical Optical
Parameters
Center Frequency or Wavelength [THz] or [nm]: the center frequency f0
of the filter transfer function. Allowed values: min (fsim_THz-BWsim_THz)/2;
max (fsim_THz+BWsim_THz)/2; default fsym_THz.
Squared Pulse Duration [ps]: the squared pulse duration of the matched
filter. In a digital transmission environment it is usually given by the inverse
of the system bit rate. Allowed values: min >0; max -; default: 100.
Amplitude Plot: Toggle to plot or not to plot. Allowed values: yes or no;
default: no
Inputs Outputs
Electrical -
Parameters
Start Time [ps]: the start time for measurement. Allowed values: min 0;
max simulation time; default 0.
End Time [ps]: the stop time for measurement. Allowed values: min 0; max
simulation time; default simulation time.
Samples per Bit: Number of samples per bit.
PRBS Polynomial Degree: the degree of PRBS polynomial. Allowed values:
min 5; max 28; default 12.
PRBS Polynomial Number: the number of PRBS polynomial. Allowed
values: min 1; max 18; default 1.
Inputs Outputs
Electrical -
Parameters
Start Time [ps]: the start time for measurement. Allowed values: min 0;
max simulation time; default 0.
End Time [ps]: the stop time for measurement. Allowed values: min 0; max
simulation time; default simulation time.
Samples per Bit: Number of samples per bit.
Pattern Length: the length of PRBS polynomial. Allowed values: min 1; max
5; default 3.
PRBS Polynomial Degree: the degree of PRBS polynomial. Allowed values:
min 5; max 28; default 12.
PRBS Polynomial Number: the number of PRBS polynomial. Allowed
values: min 1; max 18; default 1.
Window of Q Evaluation [% of Bit Rate]: the window for evaluation of
the Q-factor. Allowed values: min 0; max 100; default 0.2.
The OptSim rate equation laser block, in order to simulate the optical
behavior of the laser, solves the following rate equations:
dN I
= − GS − γ e N
dt q
dS
= (G − γ ph )S + Rsp
dt
S
Pl =
θ
1 2 1 3 1 4 1 5
Pout = Coupl Pl + Cnl 2 Pl + Cnl 3 Pl + Cnl 4 Pl + Cnl 5 Pl
2 6 24 120
with:
Γvgr a
G= (N − N 0 ) − G p S
Vact
2
1 N N −12
γe = = A+ B + C 10
τe Vact VVact
1
γ ph =
τ ph
1
τ out =
rate
Bβ sp
Rsp = N2
Vact
c
v gr =
µ gr
τ out
θ=
hf 0
where:
For a more detailed description of the rate equation parameters see the
corresponding section on the Rate Equation Laser help.
To avoid non-real solutions for the laser rate equations, the following
transformations are applied:
V
nϕ T
N = Nee
Pl = (m + δ )
2
where:
kT
ϕT =
q
N e : charge density
δ : small constant greater than zero
N e nϕT dV I N e nϕT N
V V
2 Γv gr a
e = − e − 1 − e − Rw 2 (ΘI 1 ) − θ (m + δ ) [(ΘI1 ) − N 0 ] − G P
nϕ T dT q τ n τ n Vact
2(m + δ )
dm
=−
(m + δ )2 + 1 Bβ sp N 2 + (m + δ )2 Γv gr a [(ΘI ) − N ] − G
1 0 P
dt τp θ Vact Vact
Pl = (m + δ )
2
where:
1
τn =
A
2
N N −12
Rw 2 ( N ) = B + C 10
Vact V
act
N = ΘI1
τn
Θ=2
q
Mapping the above equations on the electrical circuit of Figure 1.1 //ET, they
can be solved with SPICE (or any other EDA tool compatible with such a
grammar) without introducing any approximation.
V dV
I D = I S e nϕ T − 1 + (Cd + C j )
dt
where:
V
I
Cd = tt s e nϕ T
nϕT
C j0
Cj = m
V
1 −
ϕ0
I = I T 1 + I D 2 + I C 2 + Br1 + Bs1
I T 1 = I D1 + I C1
dm m
C ph + = Br 2 + B s 2
dt R ph
C ph = 2τ p
R ph = 1Ω
Ne nVϕ
I D1 = e T − 1
Θ
Ne nVϕ 1 nϕ T dV
V
ID2 = e T
− 1 + 2τ n e
Θ nϕT dT
1
τn =
A
2τ n
Θ=
q
tt = 2τ n
Ne
I C1 = I C 2 =
Θ
Br1 = q Rw 2 ( N ) = q Rw 2 (ΘIT 1 )
2 Γv a
Bs1 = q θ (m + δ ) gr [(ΘI1 ) − N 0 ] − GP
Vact
τ p 1 Bβ sp 2 τ p 1 Bβ sp
Br 2 = N = (ΘIT 1 )2
θ (m + δ ) Vact θ (m + δ ) Vact
Γv a
Bs 2 = τ p (m + δ ) gr [(ΘI1 ) − N 0 ] − GP − δ
Vact
δ =0
N
Ne = th
V forward
nϕT
e
where:
Vact
N th = N 0 +
c
τ ph Γ a
µg
Finally, B pf allows viewing the laser output power in form of a voltage at the
node pf.
The above electrical circuit is the model of the laser junction. The packaging
effects are modeled through the bonding inductance Lb , the parasitic
The Laser SPICE Interface is a new capability of the OptSim Best Fit Laser
Toolkit, accessible under the tab named Laser Spice Circuit.
Using the Best Fit Laser Toolkit it is possible to get the rate equations
parameters starting from the laser datasheet. Then the Laser SPICE interface
allows generating the circuit to simulate the fitted laser in the electrical
domain. This circuit can be used with any EDA tool compatible with the SPICE
3 grammar.
To generate the laser circuit file, click the button on the toolbar with a blue
laser on it (the color blue indicating the electrical domain) or select the menu
Spice – Generate. On the File Selection window choose the laser name.
A file with the chosen name followed by the extension .cir is generated. It
contains two distinct electrical sub-circuits, the first one modeling the laser
junction only (designated by the chosen file name followed by the extension
j), and the second one modeling the laser junction and the packaging
(designated by the chosen file name).
In the laser junction sub-circuit, the nodes pj and nj represent the electrical
connections, while the voltage at the node pf represents the output power.
In the packaged laser sub-circuit, the nodes pj and nj represent the electrical
connections, the voltage at the node pf represents the output power and the
voltage at the node os indicates the current flowing through the laser
junction. This is important when using the results from an EDA tool simulation
of the packaged laser: To reproduce the exact driving conditions of the rate
equation laser in an OptSim simulation, this current must be used for driving
the laser diode.
Parameters
Laser Forward Voltage [V]: Allowed values: min 0+; max -; default 1.7 V.
Lb Bonding Inductance [nH]: Allowed values: min 0+; max -; default 3
nH.
Cp Parasitic Capacitance [pF]: Allowed values: min 0+; max -; default 1.8
pF.
Rs Series Resistance [Ohm]: Allowed values: min 0+; max -; default 25
Ohm.
Generate test circuit file: This option allows the user to generate a test file
named Test followed by the name chosen for the laser, to test the laser on
DC operation with a current ranging from 0 to 100 mA.
References
[1] Pablo V. Mena et al, “Rate-Equation-Based Laser Models with a
Single Solution Regime”, JOURNAL OF LIGHTWAVE TECHNOLOGY,
VOL. 15, NO. 4, APRIL 1997 717
Electrical Electrical
Parameters
Gain [dB]: Small-signal amplifier gain Allowed values: min 0; max -;
default 20.
Linear_level_AU [a.u.]: Electrical signal output level up until which the
response is linear. Allowed values: min -; max -; default 10.
Saturation_level_AU [a.u.]: Electrical output signal saturation level.
Allowed values: min -; max -; default 20.
Optical Electrical
Parameters
Dark_current_nA [nA]: Dark current of the PIN photodiode Allowed values:
min 0; max -; default 0.1.
Photodiode_bandwidth_GHz [GHz]: Electrical bandwidth of the
photodiode Allowed values: min -; max -; default 20.
Responsivity [-]: Responsivity of the photodiode Allowed values: min 0;
max -; default 0.8.
Linear_level_mW [mW]: Optical input power up until which the photodiode
response is linear. Allowed values: min 0; max -; default 10.
Saturation_level_mW [mW]: Optical input saturation level. Allowed
values: min Linear_level_mW; max -; default 20.
To use this model, select it from the GUI Toolbox and drop it in the design area. Right clicking
this model icon shows an option to look inside where the user lays out the desired section of
the main project that needs to be repeated. The type of input and output signals is the same.
Parameters
NumReps: number of repetitions.
References
[1] P. V. Mena, J. J. Morikuni, S.-M. Kang, A. V. Harton, and K. W.
Wyatt, “A comprehensive circuit-level model of vertical-cavity surface-
emitting lasers,” Journal of Lightwave Technology, 17, 2612 (1999).
[2] J.W. Scott, R.S. Geels, S.W. Corzine and L.A. Coldren, “Modeling
temperature effects and spatial hole burning to optimize vertical-cavity
surface-emitting laser performance”, IEEE J. Quantum Electron., vol. 29,
pp. 1295-1308, May 1993.
[3] M. H. Crawford, K. D. Choquette, H. Q. Hou, R. J. Hickman, K. M. Geib
and B.E. Hammons, “Visible VCSELs: Recent advances and applications”
in Proc. 1997 Dig. LEOS Summer Topical Meetings-Vertical_Cavity Lasers,
1997, pp.17-18
[4] M. H. Crawford, K. D. Choquette, R. J. Hickman and K. M. Geib, “InAlGaP
vertical cavity surface emitting lasers (VCSEL’s): Processing and
performance”, in Proc. Int. Conf. InP Related Mater., 1997, pp. 32-35
[5] Y. Ohiso, K. Tateno, Y. Kohama, A. Wakatsuki, H. Tsunetsugu and T.
Kurokawa, “Flip-chip bonded 0.85-µm bottom-emitting vertical-cavity
laser array on AlGaAs substrate”, IEEE Photon. Technol. Lett., vol. 8, pp.
1115-1117, Sept. 1996
[6] B. J. Thibeault, K. Bertilsson, E.R. Hegblom, E. Strzelecka, P.D. Floyd, R.
Naone and L. A. Coldren, “High-speed characteristics of low-optical loss
oxide-apertured vertical-cavity lasers”, IEEE Photon. Technol. Lett., vol.
9, pp. 11-13, Jan. 1997
Parameters
Number of samples: Number of samples per bit
Inputs Outputs
Electrical Electrical
Parameters
Compensating dispersion [ps/nm]: total amount of chromatic dispersion
to compensate expressed in ps/nm: min 0.0; max -; default 0.0
Inputs Outputs
Electrical Electrical
Parameters
Transimpedance (Rt) [Ohm]: min (0.0; max -; default 5.0e5
Gain: TIA gain that can be calculated from Rt: min -; max -; default
113.9794
Input Referred Noise density (IRND) [pA/sqrt(Hz)]: min 0.0; max -;
default 20.0
Standard deviation (sigma): TIA noise standard deviation that can be
calculated from IRND: min 0.0; max -; default 1.13137e-05
The output signal amplitude can be defined with the following criteria:
• RMS amplitude: root mean square amplitude calculated on
all the output symbols
• Average amplitude: linear average amplitude calculated on all
the output symbols
• Maximum amplitude: output symbols maximum amplitude
• Minimum distance: minimum amplitude difference between
output symbols
Inputs Outputs
QAM_symbol_Q: in-quadrature
component of the M-QAM modulated
electrical signal
Parameters
baud_rate [Gbaud/s]: M-QAM symbols rate expressed in Giga baud per
second: min (0.0; max -; default 2.5
amplitude_criterion: sets the amplitude criterion adopted for the
amplitude: rms_amplitude, average_amplitude, maximum_amplitude,
and minimum_distance
amplitude [Arbitrary Unit]: value of the envelope of the output signal: min
(0.0; max -; default 1.0
delay [ps]: delay in ps before sampling logical input signal: min 0.0; max -;
default 0.0
The input signal amplitude can be defined with the following criteria:
• RMS amplitude: root mean square amplitude calculated on
all the input symbols
• Average amplitude: linear average amplitude calculated on all
the input symbols
• Maximum amplitude: input symbols maximum amplitude
• Minimum distance: minimum amplitude difference between
input symbols
Inputs Outputs
Parameters
baud_rate [Gbaud/s]: M-QAM symbols rate expressed in Giga baud per
second: min (0.0; max -; default 2.5
amplitude_criterion: sets the amplitude criterion adopted for the
amplitude: rms_amplitude, average_amplitude, maximum_amplitude,
and minimum_distance
amplitude [Arbitrary Unit]: value of the envelope of the input signal: min
(0.0; max -; default 1.0
delay [ps]: delay in ps before sampling logical input signal: min 0.0; max -;
default 0.0
ein_I eout_quadmix_I
cos(2 ⋅ π ⋅ f c ⋅ t + ϕ )
90°
ein_Q eout_quadmix_Q
Inputs Outputs
Parameters
frequency [GHz]: local oscillator (carrier) frequency: min -; max -; default
10.0
phase [Degrees]: local oscillator (carrier) phase: min -; max -; default 0.0
input
PBS output1
90 deg
hybrid output2
x pol
output3
90 deg
output4
hybrid
E LO (t ) = PLO ⋅ e j 2πf 0t ⋅ e jΦ LO ( t ) + nRIN (t ) y pol
Splitter
Laser 1->2
E RX (t ) E p (t )
90 degree
hybrid
E LO (t ) E q (t )
E p (t ) 1 jε 1 e jϕc E RX (t )
= e jϕ c ⋅
E q (t ) 2 e 1 E LO (t )
3
ϕc = π −ε ε: angle inaccuracy
4
Inputs Outputs
Optical Optical
Parameters
Ex [deg]: inaccuracy angle of 90 degree hybrid on the x polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
Ey [deg]: inaccuracy angle of 90 degree hybrid on the y polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
plo [dBm]: local oscillator power. min -; max -; default 0.
philo [rad]: local oscillator phase. min 0; max 3.14; default 0.
RIN [dB/Hz]: relative intensity noise of local oscillator. min -; max -;
default -150
NPOLES: number of poles of Bessel LPF used to emulate bandwith limitation
of LO RIN min 1; max -; default 5
BW_RIN [GHz]: -3 dB bandwidth of Bessel LPF used to emulate bandwith
limitation of LO RIN min 0.; max -; default 40
input
PBS Output1,2
90 deg
hybrid Output3,4
x pol
Output5,6
90 deg
hybrid Output7,8
Splitter
Laser 1->2
E RX (t ) E p (t )
90 degree
hybrid
E LO (t ) E q (t )
E p (t ) 1 jε 1 e jϕc E RX (t )
= e jϕ c ⋅
(
Eq t 2) e 1 E LO (t )
3
ϕc = π −ε ε: angle inaccuracy
4
Inputs Outputs
Optical Optical
Parameters
Ex [deg]: inaccuracy angle of 90 degree hybrid on the x polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
Ey [deg]: inaccuracy angle of 90 degree hybrid on the y polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
plo [dBm]: local oscillator power. min -; max -; default 0.
philo [rad]: local oscillator phase. min 0; max 3.14; default 0.
RIN [dB/Hz]: relative intensity noise of local oscillator. min -; max -;
default -150
NPOLES: number of poles of Bessel LPF used to emulate bandwith limitation
of LO RIN min 1; max -; default 5
BW_RIN [GHz]: -3 dB bandwidth of Bessel LPF used to emulate bandwith
limitation of LO RIN min 0.; max -; default 40
Signal Input
PBS Output1
90 deg
hybrid Output2
x pol
Output3
90 deg
hybrid Output4
y pol
E RX (t ) E p (t )
90 degree
hybrid
E LO (t ) E q (t )
E p (t ) 1 jε 1 e jϕc E RX (t )
= e jϕ c ⋅
E q (t ) 2 e 1 E LO (t )
3
ϕc = π −ε ε: angle inaccuracy
4
Inputs Outputs
Parameters
Ex [deg]: inaccuracy angle of 90 degree hybrid on the x polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
Ey [deg]: inaccuracy angle of 90 degree hybrid on the y polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
Signal Input
PBS Output1,2
90 deg
hybrid Output3,4
x pol
Output5,6
90 deg
Output7,8
hybrid
y pol
Provided that the LO power is large with respect to the power of the
received signal, the interference terms are small, and, except for bias
terms, the output are proportional to the in-phase and quadrature
components of x and y component of received optical signals.
Inserting photodetectors at the four output ports 1,3,5,7 or 2,4,6,8 we
can obtain four electric current signals proportional to the mentioned four
components of received optical signal, allowing coherent detection. The
output ports 1-2,3-4,5-6,7-8 can be used in a balanced photodetector
configuration.
• Single ended 90 degree hybrid
The considered 90 degree hybrids are ideal components except for the
possible insertion of an angle inaccuracy that alters the component
behavior as described in Fig. 2
E RX (t ) E p (t )
90 degree
hybrid
E LO (t ) E q (t )
E p (t ) 1 jε 1 e jϕ c E RX (t )
= e jϕ c ⋅
E q (t ) 2 e 1 E LO (t )
3
ϕc = π −ε ε: angle inaccuracy
4
Inputs Outputs
Parameters
Ex [deg]: inaccuracy angle of 90 degree hybrid on the x polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
Ey [deg]: inaccuracy angle of 90 degree hybrid on the y polarization. Ideal
hybrid correspond to Ex=0. min 0; max 90; default 0.
QAM
time
QAM_baud_period
OFDM
time
Inputs Outputs
Parameters
baud_rate [Gbaud/s]: QAM symbol rate: min (0.0; max -; default 2.5
subcarriers_number: number of subcarriers used for the OFDM modulation,
corresponding also to the number of input QAM symbols used to encode one
output OFDM symbol: min -; max -; default 0.0
cyclic_prefix [Fraction_baud_period]: length of the OFDM cyclic prefix
obtained copying circularly the portion of the OFDM symbol at its beginning:
min 0.0; max 1.0); default 0.25
auto_sync_training_sequence: controls the automatic synchronization of
the sampling instant at the beginning of the OFDM symbol using the
transmitted tranining sequence: No, Yes; default: Yes
training_sequence_amplitude: amplitude of the transmitted training
sequence: min (0.0; max -; default 1.0
restore_QAM_amplitude_and_phase: controls the automatic restoration
of the OFDM symbol amplitude and phase if auto_sync_training_sequence is
selected: No, Yes; default: Yes
QAM
time
QAM_baud_period
OFDM
time
The user can specify the number of subcarriers used for the OFDM
modulation, and a cyclic prefix to mitigate the effect of the channel
dispersion
A training sequence can also be transmitted before the first OFDM
symbol, useful to automatically synchronize the receiver sampling
instant with the beginning of the OFDM symbols.
Inputs Outputs
Parameters
baud_rate [Gbaud/s]: QAM symbol rate: min (0.0; max -; default 2.5
subcarriers_number: number of subcarriers used for the OFDM modulation,
corresponding also to the number of input QAM symbols used to encode one
output OFDM symbol: min -; max -; default 0.0
cyclic_prefix [Fraction_baud_period]: length of the OFDM cyclic prefix
obtained copying circularly the portion of the OFDM symbol at its beginning:
min 0.0; max 1.0); default 0.25
transmit_training_sequence: controls the transmission of the training
sequence before the first OFDM symbol: No, Yes; default: Yes
training_sequence_amplitude: amplitude of the training sequence, if
transmitted: min (0.0; max -; default 1.0
Inputs Outputs
Optical
Parameters
NS: Samples per symbol: min 2; max -; default samples.
NDEG: Degree of the 4 PRBS: min 1; max 11; default 11.
NPOL1: Polynomial number of the PRBS transmitted on the first data
sequence default 1.
NPOL2: Polynomial number of the PRBS transmitted on the second data
sequence default 2.
NPOL3: Polynomial number of the PRBS transmitted on the third data
sequence default 3.
NPOL4: Polynomial number of the PRBS transmitted on the fourth data
sequence default 4.
NOTE: the 4 PRBS must be different one to the other one, i.e., NPOL1
≠ NPOL2 ≠ NPOL3 ≠ NPOL4
References
[1] D. N. Godard, "Self recovering equalization and carrier tracking in two-
dimensional data communication systems," IEEE Transactions on
Communications, vol. COM-28, No. 11, Nov. 1980, pp. 1867-1875.
[2] S. J. Savory et al., “Electronic compensation of chromatic dispersion using
a digital coherent receiver,” Optics Express, Vol. 15, No. 5, 5 March
2007.
[3] A. J. Viterbi and A. M. Viterbi, “Nonlinear estimation of PSK modulated
carrier phase with application to burst digital transmission,” IEEE
Transactions on Information Theory, vol. IT-29, No. 4, Jul. 1983, pp. 543-
551.
Inputs Outputs
Optical
Parameters
Samples_per_Symbol: it is an integer number corresponding to the number
of SpS use for the simlation; min 2; max -; default 10.
PRBS_degree: degree of the four PRBS used at the transmitter; min 1;
max 28; default 11.
N_training_PRBS: number of PRBS used as training sequence. For instance,
if PRBS_degree=11 and N_training_PRBS=5 the duration of training
sequance is (2^11-1)*5
Filtering_SpS: number of samples per symbol used for the FIR filters. This
number can be 1 or 2. If Filtering_SpS=2 Samples_per_Symbol must be an
even number. min 1; max 2; default 2.
Num_taps_filter: number of taps used for FIR filters; min 1; max -;
default 11.
LMS_factor: it is the convergence factor of the LMS algorithm. It is a
positive real number smaller than 1; min >0; max 1.; Default 0.001
N_PRBS_px: PRBS identifier for px data flow
N_PRBS_qx: PRBS identifier por qx data flow
N_PRBS_py: PRBS identifier por py data flow
N_PRBS_qy: PRBS identifier por qy data flow
NOTE: the 4 PRBS must be different one to the other one, i.e.,
N_PRBS_px ≠ N_PRBS_qx ≠ N_PRBS_py ≠ N_PRBS_qy
References
[1] Monson H. Hayes Statistical Digital Signal Processing and Modeling, Wiley,
1996
[2] Simon Haykin Adaptive Filter Theory, Prentice Hall, 2002
[3] Simon S. Haykin, Bernard Widrow (Editor) Least-Mean-Square Adaptive
Filters, Wiley, 2003
[4] Bernard Widrow, Samuel D. Stearns Adaptive Signal Processing, Prentice
Hall, 1985
TS TS TS
update DD update update DD update update
Nsym,TS
TS overhead Payload
Nsym,frame
Ntaps
taps Decision
matrix
M M
TS DD
LMS LMS
Nsym,TS µTS µDD
TS: Training Sequence
DD: Decision Driven
LMS: Least Mean Square
Inputs Outputs
Electrical
Parameters
Samples_per_Symbol: number of samples per symbol used for the
simulation; min 2; max -; default samples.
PRBS_degree: must match the transmitter four PRBS degree; min 5;
max 128; default 11.
PRBS_poly_num: must match the transmitter PRBS polynomial number;
min 1; max 176; default 1.
num_sym_frame: Duration of the frame expressed as number of symbols.
If this number is equivalent to a time duration larger than the overall
simulation the frame is truncated; min 1; max totbit; default totbit/5.
num_sym_ts: number of symbols used as training sequence within every
frame; min 1; max num_sym_frame; default num_sym_frame /10.
tracking_DD: flag enabling the decision-driven post-training tracking; option
on, off; default off.
samples_per_symbol_LMS: number of samples per symbols used by the
LMS algorithm; min 1; max 2; default 1.
samples_per_symbol_filter: number of samples per symbols used to apply
the equalization implemented through FIR filters; min 1; max 2; default 1.
clock_shift: clock shift in number of samples with respect to the
automatically recovered clock; min 0; max Samples_per_Symbol-1; default
0.
num_taps_filter: number of taps used for the implementation of the FIR
filters; min 1; max -; default 15.
LMS_factor_training: LMS convergence factor used during the training
phase; min (0; max 1; default 1e-3.
LMS_factor_tracking: LMS convergence factor used during the tracking
phase; min (0; max 1; default 1e-3.
num_sym_IQ_delay: must match the transmitter delay between the PRBS
of the in-phase and quadrature signals in terms of number of symbols;
min 1; max -; default 100 .
References
[1] Monson H. Hayes Statistical Digital Signal Processing and Modeling, Wiley,
1996
[2] Simon Haykin Adaptive Filter Theory, Prentice Hall, 2002
[3] Simon S. Haykin, Bernard Widrow (Editor) Least-Mean-Square Adaptive
Filters, Wiley, 2003
[4] Bernard Widrow, Samuel D. Stearns Adaptive Signal Processing, Prentice
Hall, 1985
( )
K
=∑
2
M (n)
SQRT y k − µ ' n, k
k =1
µ 'n ,k
where is the mean value of the square-root of the k-th signal sample
yk for the n-th trellis branch and K is the number of samples per bit used by
the processor.
Inputs Outputs
Parameters
Starting sample (integer, range [0,Inf), default 0)
First sample of the input electrical signal considered by the algorithm.
Previous samples are discarded.
PRBS degree (integer, range [5,28], default 15)
Degree of the pseudo-random bit sequence used at the transmitter.
Polynomial number (integer, range [1,13], default 1)
Number of polynomial used to generate the PRBS.
Number of samples per bit of the input signal (integer, range [1,total
number of simulated samples], default 20)
Number of samples per bit of the input signal, like set in the simulation
parameters.
Number of samples per bit used in the MLSE processor (integer, range
[1, number of samples per bit of the input signal], default 4)
Number of samples per bit used in the MLSE processor.
Trellis memory (integer, range [1,14], default 4)
Logarithm in base 2 of the number of trellis states.
Number of resolution bits of the A/D converter (integer, range [0,Inf),
default 0)
Number of resolution bits of the A/D converter (“0” means infinite resolution).
Number of training bits (integer, range [1,(total simulated samples-
starting sample)/(number of samples per bit)], default 32767)
Number of bits used for trellis training (i.e. used for the evaluation of
the µ 'n ,k mean values used for the evaluation of the branch metrics).
Number of MLSE bits (integer, range [1,(total simulated samples-starting
sample)/(number of samples per bit)], default 32767)
Number of bits processed by the MLSE processor.
Length of decision window (integer, range [1,Inf), default 3)
Length of decision window of the Viterbi algorithm (expressed in multiples of
the trellis memory n, where 2n is the number of trellis states).
Delay offset (integer, range (-Inf,Inf), default 0)
Additional offset (in number of samples) between PRBS sequence and
electrical signal after the automatic synchronization procedure.
Output Data
Bit error rate, evaluated through direct error counting after MLSE processing.
References
[1] D.E. Crivelli, H.S. Carrer, M.R. Hueda, “On the performance of reduced-
state Viterbi receivers in IM/DD optical transmission systems,” Proc. of
ECOC 2004, paper We.4.P.083.
[2] P. Poggiolini et al.,“1,040 km Uncompensated IMDD Transmission over
G.652 Fiber at 10 Gbit/s using a Reduced-State SQRT-Metric MLSE
Receiver”, Proc. of ECOC 2006, Paper Th4.4.6.
The output field Eout as a function of the frequency f , the input field
Ein , the input voltage Vin , and the modulator parameters is:
r r EL
− dB π (Vin − Von ) ( f − fT ) ⋅ e jϑ
Eout ( f ) = Ein ( f ) ⋅ 10 20 ⋅ cos +π (1)
2 Vπ BW
with:
1
ϑ = α log(Pout ) (2)
2
where:
r r EL
− dB π (Vin − Von ) (λ − λT ) ⋅ e jϑ
Eout (λ ) = Ein (λ ) ⋅ 10 20 ⋅ cos +π (3)
2 Vπ n ⋅ ∆L
where:
SPT Behavior
In SPT simulations the input Power Spectrum is multiplied by the following
transfer function:
−
ELdB
−
PR dB
1 ( f − fT )
H ( f ) = 10
2 10
⋅ 10 10
⋅ 1 + cos 2π
2 BW
( ELdB + PRdB )
− ( f − fT )
= 10 10
⋅ cos 2 π
BW
where:
Inputs Outputs
Electrical: Electrical input signal Vin Optical: the modulated optical signal
required to modulate the input Eout at the output of the modulator
Optical signal Ein
Optical: Optical input signal Ein,
injected into the model, to be
modulated
References
[1] F.Koyama and K. Iga, Frequency chirping in external modulators, J.
Lightwave Technol., vol.6, no.1, pp.87-33, Feb.1988.
[2] Govind P. Agrawal, Nonlinear Fiber Optics, 2nd edition, Academic Press,
San Diego, 1995
[3] Chris Sarantos and Nadir Dagli, Fellow, IEEE, “A Photonic Analog-
to-Digital Converter Based on an Unbalanced Mach-Zehnder
Quantizer”, IEEE, 2007
Inputs Outputs
1 B
16-channel 2DPSK receiver, 202 baseband phase locked loop, 390
16-channel 2DPSK transmitter, 92 baseband rectifier clock recovery, 392
2 BeamPROP-OptSim interface, 457
2DPSK receiver, 201 beat length, 108
2DPSK transmitter, 91 BER estimator, 341
2dpsk_rx, 201 BER techniques, 193
2dpsk_tx, 91 Best Fit Laser Toolkit, 16
A bias, 57
A/D converter (absolute value + sign), 398 binary differential receiver, 460
A/D converter (two's complement), 399 binary pulse amplitude demodulator, 236
adder birefringence, 98
logical (absolute value + sign), 406 Blind Receiver for Coherent Polarization
logical (two's complement), 407 Multiplexed QPSK, 514, 522
modulo 2, 407 BPADEM, 236
all-order PMD, 309 BYADD, 407
AMPDE2, 217
C
amplifier
carrier and clock recovery subsystems
50 ohm, 243
baseband phase locked loop, 390
electrical, 243
baseband rectifier clock recovery subsystem,
optical, 148
392
semiconductor optical, 159
derivative-product clock recovery subsystem,
user-defined, 245
392
amplitude modulator, 130 early-late-gate symbol synchronizer, 369
AMPMO2, 216 early-late-gate symbol synchronizer (2
dimensions), 371
analog amplitude modulator, 216
electrical N-th power device, 366
analog mo-demodulators first order baseband phase-locked loop, 388
analog amplitude modulator, 216 frequency divider, 365
coherent amplitude demodulator, 217 I-F rectifier clock recovery, 395
electrical envelope, 222 in-phase mid-phase symbol synchronizer (1
frequency demodulator, 219 dimension), 374
frequency modulator, 218 in-phase mid-phase symbol synchronizer (2
phase demodulator, 221 dimensions), 376
phase modulator, 220 N-phase decision feedback loop, 382
AND, 405 N-th order Costas loop, 380
AOPMD, 309 N-th order power loop for carrier
synchronization, 379
APD photodiode, 183 optimum symbol (ML) synchronizer (1
attenuator dimension), 384
optical, 304
Index
EXPXY, 359 H
F HILBER, 363
FFT OFDM, 510 Hilbert transformer, 363
fiber grating, 307 I
ideal, 307
ideal balanced 2DPSK receiver, 198
user-defined, 308
ideal clock generator, 62
fiber link, 96
ideal dual-arm Mach-Zehnder interferometer,
fibers
320
fiber link, 96
I-F rectifier clock recovery, 395
filter
Bessel, 259 IFFT OFDM, 512
butterworth, 255 IMPSY1, 374
Chebychev, 263 IMPSY2, 376
electrical, 250
matched, 267, 471 impulse generator, 63
optical, 166 in-phase mid-phase symbol synchronizer (1
raised cosine, 269 dimension), 374
single-pole low-pass, 250 in-phase mid-phase symbol synchronizer (2
two-pole low-pass, 251 dimensions), 376
user-defined, 271
integer to logical signal conversion (absolute
filter (optical) value + sign), 420
Bessel, 170
multiple-stage Lorentzian, 166 integer to logical signal conversion (two's
raised cosine, 168 complement), 420
supergaussia, 169 integer to logical signal conversion with levels
user defined, 170 (absolute value + sign), 422
first order baseband phase-locked loop, 388 integer to logical signal conversion with levels
flip-flop D type, 408 (two's complement), 421