Close
About
FAQ
Home
Collections
Login
USC Login
Register
0
Selected
Invert selection
Deselect all
Deselect all
Click here to refresh results
Click here to refresh results
USC
/
Digital Library
/
University of Southern California Dissertations and Theses
/
Reconfigurable high-speed processing and noise mitigation of optical data
(USC Thesis Other)
Reconfigurable high-speed processing and noise mitigation of optical data
PDF
Download
Share
Open document
Flip pages
Contact Us
Contact Us
Copy asset link
Request this asset
Transcript (if available)
Content
RECONFIGURABLE HIGH-SPEED PROCESSING AND NOISE
MITIGATION OF OPTICAL DATA
by
Amirhossein Mohajerin Ariaei
A Dissertation Presented to the
FACULTY OF THE USC GRADUATE SCHOOL
UNIVERSITY OF SOUTHERN CALIFORNIA
In Partial Fulfillment of the
Requirements for the Degree
DOCTOR OF PHILOSOPHY
(ELECTRICAL ENGINEERING)
August 2018
Copyright 2018 Amirhossein Mohajerin Ariaei
ii
Dedication
To my lovely parents, Soheila Hashemi and Mohammad Mohajerin Ariaei,
and my dear sister Mahshid Mohajerin Ariaei
who always love and support me
unconditionally.
iii
Acknowledgments
I am very grateful to many people for all their supports and advices during my PhD. I
would like especially thank Professor Alan Willner who was a great advisor and
teacher. I would like to thank Prof. Willner because of his patient, supports, invaluable
technical and scholarly discussions and because of all his professional and priceless
life-long lessons. I will always be thankful and have him in mind.
I would like to thank the dissertation committee Professor Stephan W. Haas and
Professor Wei Wu for their supports, invaluable feedback, insightful comments, and
contributions. I am also very thankful to Professor Todd Brun and Professor Joseph
D. Touch for their guidance during my qualification exam. I would like to thank
Professor Jawad Salehi who provided so many insights and tremendous support as my
M.Sc. advisor at Sharif University of Technology. I would like to thank the amazing
and helpful staffs of the electrical engineering department at USC Diane Demetras,
Gerrielyn Ramos, and Corine Wong. My especial thanks to Tim Boston who always had
been helpful, and the whole EE department missed him so much in Spring 2018. I
would like to thank Professor Robert Scholtz who has been a great teacher and role
model for me during my academic journey.
I would like to thank Dr. Nisar Ahmed, Dr. Yongxiong Ren, Dr. Guodong Xie,
Dr. Jeng-Yuan Yang, Dr. Hao Huang, and Dr. Yan Yan for their help and mentorship.
My especial thanks to Dr. Mohammad Reza Chitgarha and Dr. Salman Khaleghi for
their help, supports, insightful suggestions, and great advises. I have learned a lot from
you. Thank you! I would also thank my colleagues, Ahmed Almaiman, Yinwen Cao,
Changjin Bao, Long Li, and Zhe Zhao for their support and help. I am very grateful to
Fatemeh Alishahi and Ahmad Fallahpour who were not only my colleagues but also
great friends who never let me down. Thank you.
I am very grateful to Dr. Morteza Ziyadi not only as an especial colleague who
helped me through the challenges and difficulties of PhD journey, but also because he
has been a great friend who shaped many of my best memories at USC. I would like
iv
to thank him for his supports, insightful discussions, and collaborations. Thank you
very much Morteza. I have been blessed with great friends Dr. Fatemeh Alishahi, Dr.
Keyvan Rezaei Moghadam, Dr. Parisa Mansourzifard, Kaveh Rezaei Moghadam, and
Dr. Nooshin Tajik who helped and supported me in my life at USC.
Finally, I would love to express my deepest gratitude to my parents and my sister
whom without their love, supports, encouragements, and patient I could not reach at
this stage.
v
Table of Contents
Dedication ................................................................................................................... ii
Acknowledgments ..................................................................................................... iii
List of Figures ........................................................................................................... vii
Abstract .................................................................................................................. xiii
Chapter 1 Introduction to Optical Signal Processing ........................................... 15
1.1 Nonlinear Processes ...................................................................................... 15
1.1.1 Four-Wave Mixing ............................................................................. 16
1.1.2 Three Wave Mixing ........................................................................... 17
1.1.3 Materials and Devices ........................................................................ 18
1.2 Enabling Technologies ................................................................................. 19
1.2.1 Optical Wave Encoding ..................................................................... 19
1.2.2 Coherent Detection............................................................................. 20
1.2.3 Optical Frequency Comb ................................................................... 21
1.3 Basic Enabling Operations ........................................................................... 22
1.3.1 Wavelength Conversion ..................................................................... 22
1.3.2 Wavelength Multicasting ................................................................... 23
1.3.3 Optical Multiplexing .......................................................................... 25
1.3.4 Conversion-Dispersion Optical Delays .............................................. 28
Chapter 2 All Optical Phase Noise Mitigation ...................................................... 30
2.1 Introduction .................................................................................................. 30
2.2 Concept ......................................................................................................... 31
2.3 Experimental Setup....................................................................................... 35
2.4 Results .......................................................................................................... 36
2.7 Conclusion .................................................................................................... 40
Chapter 3 Simultaneous All-optical Phase Noise Filtering and Automatically
Locked Tunable Homodyne Reception ................................................ 41
3.1 Introduction .................................................................................................. 41
3.2 Concept ......................................................................................................... 42
3.3 Experimental Setup....................................................................................... 43
3.4 Results .......................................................................................................... 44
3.5 Demonstration of Tolerance to VCSEL-Wavelength-Drift and DFB-
High-Phase-Noise in an All-Optical Homodyne Receiver ........................... 45
Chapter 4 Tunable Nonlinear Phase-Noise Mitigation and Automatic
Frequency/Phase Locking for a Homodyne Receiver using Optical
Mixing of Nonlinearly Generated Higher Harmonics ........................ 51
vi
4.1 Introduction .................................................................................................. 51
4.2 Concept ......................................................................................................... 52
4.3 Experimental Setup....................................................................................... 54
4.4 Experimental Results .................................................................................... 55
Chapter 5 Multiplexing and Transmission of QPSK-to-16QAM Channels using
Wave Mixing for Aggregation and Noise Mitigation .......................... 58
5.1 Introduction .................................................................................................. 58
5.2 Concept ......................................................................................................... 59
5.5 Experimental Setup....................................................................................... 60
5.3 Experimental Results .................................................................................... 62
Chapter 6 Optical inter-channel interference mitigation for spectrally
overlapped 16QAM data channels using nonlinear wave mixing ..... 65
6.1 Introduction .................................................................................................. 65
6.2 Concept ......................................................................................................... 66
6.3 Experimental Setup....................................................................................... 68
6.4 Results .......................................................................................................... 70
Chapter 7 Optical Mitigation of Inter-Channel Crosstalk for Multiple
Spectrally Overlapped WDM Channels using Nonlinear Wave
Mixing ..................................................................................................... 74
7.1 Introduction .................................................................................................. 74
7.2 Concept ......................................................................................................... 75
7.3 Experimental Setup....................................................................................... 77
7.3 Experimental Results .................................................................................... 78
References ................................................................................................................. 80
vii
List of Figures
Figure 1.1 (a) Degenerate and (b) Non-degenerate four-wave mixing
(FWM) schemes for generation of phase conjugate signal copy.
ZDW: zero dispersion wavelength. ........................................................... 17
Figure 1.2 (a) Cascaded sum and difference frequency generations (cSFG-
DFG) and (b) Second harmonic generation and DFG (cSHF-DFG) for
wavelength conversion in a PPLN device. QPM: quasi-phase
matching. ................................................................................................... 18
Figure 1.3 Advanced modulation formats using amplitude and phase
domains, with independent polarization and wavelength multiplexing .... 20
Figure 1.4 Coherent signal detection using the coherent receiver. ..................... 21
Figure 1.5 Optical frequency comb features....................................................... 21
Figure 1.6 (a) Wavelength conversion in a PPLN waveguide, (b) pump
configurations, (c) amplitude and phase of the generated signals in the
cSFG-DFG processes. ............................................................................... 23
Figure 1.7 Various configurations for N-fold signal multicasting using
multi-pumps. ............................................................................................. 24
Figure 1.8 Illustration of signal multicasting of a signal onto multiple
frequencies using coherent frequency comb. ............................................ 25
Figure 1.9 Optical coherent multiplexing of coherent signals using optical
frequency combs. ...................................................................................... 26
Figure 1.10 Conventional method for generating optical QAM......................... 27
Figure 1.11 Generating optical QAM through QPSK modulated signals. ......... 27
Figure 1.12 An example of coherent multiplexing in a nonlinear device to
generate higher order amplitude and phase formats. ................................. 28
Figure 1.13 Tunable Conversion/dispersion based optical delay. ...................... 29
Figure 2.1 The concept of the phase noise mitigation scheme including four
cascaded nonlinear stages for conjugate generation, third order
viii
harmonics generation, phase quantization, and parametric
amplification in saturation regime. ............................................................ 32
Figure 2.2 The PSD of the phase-differentiator.................................................. 34
Figure 2.3 Experimental setup of phase and amplitude noise mitigation.
PM: Phase modulator, HNLF: Highly nonlinear fiber, PPLN:
Periodically poled Lithium Niobate, SLM: Spatial light modulator,
PD: Photo-detector, VOA: Variable optical attenuator. ............................ 36
Figure 2.4 Power and gain profiles of HNLF-2. Operation in saturation
regime results in amplitude noise squeezing. ............................................ 36
Figure 2.5 The constellation diagrams of the input and output of phase
quantizer for different levels of phase noise. ............................................ 37
Figure 2.6 The input and output constellation diagrams of the phase noise
mitigation systems for 30 Gbaud (a, b) and 20 Gbaud (c, d) QPSK
signals impaired by different values of phase noise. ................................. 38
Figure 2.7 (a) Percentage of phase noise range reduction for various levels
of phase noise for 30 Gbaud QPSK signal. (b) EVM of the input noisy
signal, phase quantizer output, and parametric amplifier output for
various levels of phase noise for a 30 Gbaud QPSK signal. ..................... 39
Figure 2.8 BER versus OSNR for two different phase noise levels (δφ ~43°,
~35°). ......................................................................................................... 39
Figure 3.1 The concept of the homodyne phase noise mitigation scheme.
Low frequency components of the phase noise (e.g., laser phase
noise) of an incoming signal is filtered by multiplying the signal to its
delayed conjugate. Simultaneously, the noise mitigated signal is
automatically locked to a “local” laser, P1, and sent to a PD for
detection. ................................................................................................... 43
Figure 3.2 (a) Experimental setup. DCF: Dispersion compensating fiber,
SLM: Spatial light modulator (b) Optical spectra of PPLN-1, PPLN-2. .. 44
Figure 3.3 (a) The I and Q eye diagrams of the detected 20-Gbaud QPSK
signals with three different power levels of induced phase noise and
for three different noise-bandwidths. (b) Phase noise reduction factor
for three levels of induced phase noise with different noise-
bandwidths. ............................................................................................... 45
ix
Figure 3.4 The Conceptual block diagram of the all-optical homodyne
system. ....................................................................................................... 47
Figure 3.5 (a) Experimental setup. DCF: Dispersion compensating fiber,
SLM: Spatial light modulator (b) Optical spectra of PPLN-1, PPLN-2. .. 48
Figure 3.6 Wavelength drifts of the VCSEL source. The random drifts of
<0.05 nm are occurred at different times with a few seconds
difference. .................................................................................................. 48
Figure 3.7 The constellation diagrams of 20-Gbaud QPSK signals from the
back to back VCSEL source without wavelength drifts compensation
on the top, and the all-optical wavelength drifts compensated output
from PPLN-2 on the bottom. ..................................................................... 49
Figure 3.8 constellation diagrams of 20 and 32-Gbaud QPSK signals from
the back to back DFB source without phase noise mitigation on the
left , and the all-optical phase noise mitigated output from PPLN-2 on
the right. .................................................................................................... 50
Figure 3.9 The I and Q eye diagrams for 20 and 32 Gbaud QPSK signals
modulated on DFB laser source and an uncooled VCSEL. ...................... 50
Figure 4.1 The concept of the homodyne nonlinear phase noise mitigation
scheme. ...................................................................................................... 54
Figure 4.2 (a) Experimental setup. PM: Phase modulator, HNLF: Highly
nonlinear fiber, PPLN: Periodically poled lithium niobate, SLM:
Spatial light modulator, PD: Photo detector, VOA: Variable optical
attenuator. (b) 20 Gbaud optical spectra (c) 32 Gbaud optical spectra ..... 55
Figure 4.3 The I and Q eye diagrams of 20 and 32 Gbaud QPSK signals
without noise, with noise level-1, and with noise level-2. The results
are shown for the homodyne detection scheme without phase noise
mitigation, and, the proposed homodyne phase noise mitigation
method. ...................................................................................................... 56
Figure 4.4 BER versus for homodyne detection scheme with/without phase
noise mitigation and for two different noise levels. (a) 20Gbaud (b) 32
Gbaud. ....................................................................................................... 57
x
Figure 5.1 The concept of the simultaneous optical aggregation and phase
noise mitigation system. The incoming two QPSK signals
contaminated with phase noise, e.g. phase noise of lasers with large
linewidths, are received. The conjugate copies of signals with one
symbol delay are generated using PPLN-1 waveguide and a dispersive
medium. In PPLN-2, each signal multiplied with its conjugate copy to
filter out the phase noise, and in a simultaneous process, they
coherently multiplexed to generate the output 16 QAM signal. ............... 60
Figure 5.2 (a) Experimental setup. PM: Phase modulator, PPLN:
Periodically poled lithium niobate, DCF: Dispersion compensated
fiber, SLM: Spatial light modulator, PD: Photo detector, VOA:
Variable optical attenuator. (b) 20 Gbaud optical spectra after
nonlinear stages (PPLN-1, 2). ................................................................... 61
Figure 5.3 The constellations diagrams for 10-20 Gbaud signals (a) The
input and output constellation diagrams when only a noisy QPSK
channel is sent to the system. (b) The aggregated 16 QAM signals
from QPSK channels with phase noise levels of around 50 and 35
degree. The phase noise bandwidth of the channel with higher phase
noise varies from 300 to 5000 MHz and the phase noise bandwidth of
the other channel is 300 MHz. .................................................................. 63
Figure 5.4 (a) The EVM reduction of 20 Gbaud aggregated 16QAM signal
with different phase noise bandwidths and for different levels of
phase noise, ~50, 45, and 35 degree (Noise levels 1 to 3); compared to
the case of 5000 MHz noise bandwidth (b) The BER performance of
the 16QAM aggregated signal for input signals with phase noise of
~50 and ~35 degree and different phase noise bandwidths, before and
after 100 Km transmission. ....................................................................... 64
Figure 6.1 Conceptual block diagram of the proposed optical inter-channel
interference (ICI) mitigation method. In a PPLN waveguide, three
copies of the signals are generated. In a programmable optical filter,
the signals are selected, delayed, and multiplied with complex
coefficients by adjusting their amplitudes and phases. In PPLN
waveguides at ports 1 and 2, the signals are coherently multiplexed to
mitigate the ICI. ......................................................................................... 68
Figure 6.2 (a) Experimental setup. PC: polarization controller, PPLN:
periodically poled lithium niobate, LCoS: liquid crystal on silicon, (b)
optical spectra after PPLN-1 and PPLN-2 waveguides. .................................. 69
xi
Figure 6.3 Experimentally recorded signal constellation diagrams with one tap
and two taps optical ICI mitigation and without (w/o) optical ICI
compensation. The diagrams are obtained for overlapped channels of
20Gbaud 16QAM signal and QPSK signal at different channel spacings Δf. ... 71
Figure 6.4 BER measurements with and without optical ICI compensation
method for QPSK/16QAM overlapped channels and for different channel
spacings ∆𝒇 (a,b) 20Gbaud QSPK and 16QAM channels (c,d) 25Gbaud
QPSK and 16QAM channels. ....................................................................... 72
Figure 6.5 Experimentally measured signal constellation diagrams with one tap
and two taps optical ICI mitigation and without(w/o) optical ICI
compensation for overlapped channels of 20Gbaud 16QAM signals at
different channel spacings Δf. ....................................................................... 73
Figure 6.6 BER measurements with and without optical ICI compensation
method for 16QAM/16QAM overlapped channels and for different
channel spacings ∆𝒇 (a,b) 20Gbaud 16QAM channel-1 and channel-2
(c,d) 25Gbaud 16QAM channel-1 and channel-2. .................................... 73
Figure 7.1 (a) Conceptual diagram of the proposed optical inter-channel
interference (ICI) mitigation method. The signal conjugate copies are
generated in the first PPLN waveguide. In an optical programmable
filter, the conjugate copies of data channels are separated in two sets
of even and odd channels and are adjusted with desired complex taps.
The original signals with the even and odd sets of conjugate copies
are directed to optical ICI compensation modules. (b) Conceptual
diagram of the optical ICI compensation module for even channels.
This module is composed of two PPLNs and a LCoS filter for
amplitude, phase and delay adjustments. The target signals are mixed
with neighboring channels and their delayed variants to mitigate the
ICI. ............................................................................................................. 76
Figure 7.2 (a) Experimental setup. PC: polarization Controller, PPLN:
periodically poled lithium niobate, LCoS: liquid crystal on silicon, (b)
optical spectra of conjugate copies of 20 Gbaud overlapped QPSK
signals (A) and ICI mitigated channels 1,3,5 and 7 (odd channels)
after the last nonlinear stage (B). .............................................................. 78
Figure 7.3 Experimentally measured signal constellation diagrams of
channels 1, 3, and 6 with(w.) and without(w/o) optical ICI mitigation
method for 20 Gbaud overlapped QPSK signals and at channel
spacings, Δf of 17.5 GHz, 20 GHz and 25 GHz. ...................................... 78
xii
Figure 7.4 BER measurements with(w.) and without(w/o) optical ICI
compensation method for QPSK overlapped channels (1,3 and 6) and
for different channel spacing conditions. (a) ∆f =17.5 GHz (b) ∆f = 20
GHz. .......................................................................................................... 79
xiii
Abstract
The exponential trend of bandwidth demanding applications such as cloud computing,
photos and video sharing, data storage systems and recent technological advances in
high-speed data networks have created a demand for higher speeds in data processing
and transmission. Over the past few years, the increase in system capacity has been
realized by a combination of coherent technologies and advanced modulation formats.
Coherent detection and advanced optical modulations with the use of high-speed
digital signal processing (DSP), encode information on the four optical domains of
wavelength, amplitude, phase, and polarization [1–5]. In particular, coherent
transceivers enable systems that can support spectrally efficient modulation formats to
transmit and receive many bits of information in one symbol time [6–10].
High speed optical signal processing has been one of the main research areas in
photonics for more than a decade. Optical signal processing has been of great interest
due to its inherent ultrafast THz bandwidth and its potentially phase-preserving
nature [1,11–14]. One of the main interest for using optical signal processing
techniques is the optical methods such as the ones based on nonlinear processes which
do not need to switch every bit as electronic transistors do. Optical amplifiers, can
amplify very high-speed signals (Tb/s) without processing the signal at the bit level.
Considering wavelength conversion technique as an another example, by using a pump
laser and a nonlinear element, ultra-high-speed optical data can be transferred from
one wavelength to another as optical signals pass through the optical elements [1,14].
More importantly, advances in photonic integrated circuits (PICs), nonlinear materials
and devices with higher efficiencies are the important factors for any practical
realization of optical signal processing systems in the future [15–17].
The current dissertation explores the potential of ultra-high speed optical systems
to assist DSP in processing and noise mitigation of huge amounts of data. In general,
phase or amplitude noise can cause a degradation in the received data signal-to-noise
xiv
ratio, resulting in a system power penalty. In specific, phase noise originating from
the interaction of ASE noise and Kerr nonlinearity can pose a key limitation in such
systems. This dissertation demonstrates optical systems to mitigate the noise in the
optical domain which can provide several advantages such as avoiding the impact of
optical-to-electronic conversion and supporting in-line signal processing for high baud
rate signals. By utilizing various forms of photonic nonlinear interactions, different
functionalities are demonstrated which includes; phase quantization, phase noise
filtering, simultaneous aggregation and noise mitigation of optical data, simultaneous
phase noise mitigation and automatically locked homodyne reception, and optical
inter-channel-interference mitigation for multiple spectrally overlapped channels.
15
Chapter 1 Introduction to Optical Signal
Processing
Diverse fields such as astronomy, biology, genome sequencing, physics, sociology, and
environmental science have all significantly benefit from the ability to utilize large amounts
of data. The amount of data being collected is reaching the point beyond our current ability to
process and extract all the useful information. Moreover, this data traffic is expected to grow
since fiber optic networks provide a platform for high bandwidth communications. Optical
data processing can assist electronics and DSP by take Terabytes of data and rapidly identify
key patterns and extract useful key features. The extracted Gigabytes features can then be
processed by electronics with a lower speed.
This chapter introduces the basic building blocks and nonlinear processes which
become the foundations of making optical signal processing systems and tasks in the
next chapters. We will discuss the four-wave mixing and the three-wave mixing
nonlinear processes as well as the realization of tunable delays in optical systems. We
will also introduce some main optical signal processing functions, such as optical
multicasting, coherent superposition, and higher-harmonics generation.
1.1 Nonlinear Processes
Nonlinear processes such as Kerr nonlinear effect have ultra-fast sub-ps response
times and they can be used for various wave mixing and manipulations of signals with
high speed and bandwidth of >THz [1,18]. In general, “optical wave mixing” is a
process of nonlinear interaction of different optical waves with the same or different
wavelengths generating an optical wave at a new wavelength [1]. The nonlinear wave-
mixing processes are governed by (i) conservation of energy, and (ii) phase matching
conditions, which are based on the conservation of momentum [1,19,20]. We will
overview the χ
(2)
and χ
(3)
nonlinear processes. The third-order nonlinear process of
four-wave mixing (FWM) can be achieved in third-order susceptibility χ
(3)
materials [19]. In nonlinear elements with χ
(2)
nonlinearities two waves can be mixed
16
to generate mixing terms such as second harmonic generation (SHG), sum frequency
generation (SFG), difference frequency generation (DFG), and a cascading of such
mixing products, i.e., (SFG-DFG) and (SHG-DFG) [1,21].
1.1.1 Four-Wave Mixing
Four-wave mixing (FWM) is the wave mixing process that takes place in χ
(3)
materials.
In this process, three input waves mix under the phase-matching conditions in a χ
(3)
nonlinear medium to generate a fourth wave. Figure 1.1 shows the schematic spectra
of two different types of FWM, namely, degenerate and non-degenerate FWM [1].
In the degenerate FWM process, Figure 1.1 (a), a continuous wave pump at
frequency fpump and a data signal with frequency fsig are combined and sent into a χ
(3)
nonlinear material, e.g., a highly nonlinear fiber (HNLF) [1]. If the pump is located
around the zero-dispersion-wavelength (ZDW) of the nonlinear device, the phase
matching conditions are then satisfied, and the frequency of the newly created wave
follows:
𝑓 𝑐𝑜𝑛𝑣 = 2𝑓 𝑝𝑢𝑚𝑝 − 𝑓 𝑠𝑖𝑔 (1.1)
The generated signal is a wavelength converted with conjugate phase of the
original optical data channel [1]. In the degenerate FWM, if the signal is used as the
pump, then the converted signal is proportional to the square of the signal. Therefore,
it does not have the phase information of the original signal, and does not preserves
the intensity of the original signal. In the non-degenerate FWM scheme, shown in
Figure 1.1(b), two pumps at fpump1 and fpump2 located around the ZDW of the nonlinear
medium, and the input signal are combined and sent into the nonlinear medium. Phase-
matching conditions are satisfied for FWM between the pumps and signals, if the
nonlinear element has flat dispersion slope around the ZDW, resulting in the
generation of the following mixing products [1]:
𝑓 𝑐𝑜𝑛𝑣 1
= 2𝑓 𝑝𝑢𝑚𝑝 1
− 𝑓 𝑠𝑖𝑔 (1.2)
17
𝑓 𝑐𝑜𝑛𝑣 2
= 𝑓 𝑝𝑢𝑚𝑝 1
+ 𝑓 𝑝𝑢𝑚𝑝 2
− 𝑓 𝑠𝑖𝑔 (1.3)
𝑓 𝑐𝑜𝑛𝑣 3
= 𝑓 𝑠𝑖𝑔 + 𝑓 𝑝𝑢𝑚𝑝 2
− 𝑓 𝑝𝑢𝑚𝑝 1
(1.4)
Figure 1.1 (a) Degenerate and (b) Non-degenerate four-wave mixing (FWM) schemes for generation
of phase conjugate signal copy. ZDW: zero dispersion wavelength.
Note that multiple wave-mixing interactions occur simultaneously in the two-pump
scheme resulting in both phase-conjugating and non-phase-conjugating wavelength
conversions [1].
1.1.2 Three Wave Mixing
Considering optical signals at two frequency (e.g., f1, f2 ), they can mix in a χ
(2)
nonlinear material under phase-matching conditions to generate a new signal at the
sum frequency fSFG= f1 + f2 and difference frequency fDFG= f1 - f2. Note that if only one
pump is injected to the nonlinear element, instead of sum frequency, the second
harmonic term with fSFG= 2f1 is generated.
If the two input signals are in the same frequency band, e.g., ~1550 nm, the
SHG/SFG term is at ~775 nm, which is in a different frequency band. In practice, there
are motivations in keeping the generated signals in the same frequency band. A
diagrams of the cascaded SFG and DFG (cSFG-DFG) and cascaded SHG and DFG
(cSHG-DFG) processes, are shown in Figures 1.2 (a) and (b).
sig
f
conv
f
ZDW
pump
f
*
Degenerate FWM
sig
f
1 conv
f
1 pump
f
*
2 conv
f
2 pump
f
*
3 conv
f
Non-Degenerate FWM
ZDW
18
Figure 1.2 (a) Cascaded sum and difference frequency generations (cSFG-DFG) and (b) Second
harmonic generation and DFG (cSHF-DFG) for wavelength conversion in a PPLN device. QPM: quasi-
phase matching.
1.1.3 Materials and Devices
Nonlinear wave mixing can be realized in various types of materials and devices.
Basically, different nonlinear devices provide different figures of merit, e.g, high
nonlinear efficiency, loss, bandwidth, low dispersion for phase matching, low two-
photon absorption, and latency [1,22,23]. Silica, silicon, lithium niobate,
chalcogenide, semiconductors are only few examples of materials which are used for
wave mixing process in signal processing [1].
In this thesis, highly nonlinear fibers (HNLF) are used for FWM nonlinear
process. In general, optical fibers can be used as efficient nonlinear mixing devices
when the core is engineered to become very small to confine the beams, and dispersion
profile is engineered in the design. Simultaneous wave mixings accompanied by a
large bandwidth of conversion and low-dispersion performance can be obtained as a
result of such designs for the HNLFs [1,24]. However, one of the draw-backs of
nonlinear wave mixing process for optical signal processing tasks in χ
(3)
optical
devices is that for a wide phase-matching bandwidth, a lot of mixing terms can be
generated that may cause cross-talk on the desired signal and occupy the optical
spectrum.
(2)
SFG
f
dummy
f
pump
f
SFG
QPM
(2)
DFG
f
signal
f
idler
Cascaded SFG and DFG
QPM: Quasi-phase matching
(2)
SHG
f
pump f
SFG
QPM
(2)
DFG
f
signal
f
idler
*
Cascaded SHG and DFG
19
As for the χ
(2)
nonlinear elements; a periodically poled lithium niobate (PPLN)
waveguide device can be used as a platform for the nonlinear wave mixing. PPLNs
can provide high nonlinear efficiency and relatively low propagation loss. It should be
noted that the quasi-phase matching (QPM) wavelength in PPLN devices is very
similar to ZDW in HNLFs, e.g., pumps need to be placed symmetrically around the
QPM wavelength for realizing efficient wave mixings [1].
1.2 Enabling Technologies
In this section, we will look at some of the enabling technologies such as optical signal
modulation techniques to make use of multiple domains of the optical signal and
coherent detection method to decode the optical amplitude and phase information. We
will also overview the optical comb technology as one of the potential building block
for optical signal processing system.
1.2.1 Optical Wave Encoding
An optical wave has different orthogonal domains that can be used to encode the
information [25]. In general, n bits of information can form M = 2
n
different
information symbols. Each of these M symbols can be mapped to a specific amplitude
and phase values, as shown in Figure 1.3. Moreover, the polarization of the wave (X
and Y polarization) besides the amplitude and phase of an optical wave, can be utilized
to encode the information on both polarization states and create polarization
multiplexed (PM) signals. In addition, wavelength division multiplexing technique can
be used to encode the information on different wavelengths. Altogether, by utilizing
different optical domain, we can increase the number of transmitted bits/second/Hertz
which is also known as spectral efficiency.
20
Figure 1.3 Advanced modulation formats using amplitude and phase domains, with independent
polarization and wavelength multiplexing
As shown in the constellation diagrams in Figure 1.3, different wavelengths and
polarizations are orthogonal to each other and different constellations can be chosen
for different channels. Phase shift keying (PSK), amplitude shift keying (ASK), and
quadrate amplitude modulation (QAM) are some of the modulation formats that could
be utilized to encode the data on different phase and amplitude of an optical wave. As
an example, on channel λ1 in Figure 1.3, a 16-QAM signal is encoded on the X-
polarization of the optical wave and an 8-PSK signal is encoded independently on the
Y-polarization. Or considering channel λ2 in Figure 1.3, quadrature-phase-shift-keying
(QPSK) modulation format is used for both X and Y polarization. Because these
symbols have complex number formats, instead of using amplitude and phase, one can
also define a symbol based on real (Re) and imaginary (Im) parts of the symbol. In
communication systems, the Re part is known as the in-phase (I) and the Im part is
known as the quadrature (Q) component of the data symbol.
1.2.2 Coherent Detection
Coherent systems including coherent detection can be used to recover both amplitude
and phase of an optical signal. To recover the I and Q components of a modulated
optical signal, a coherent receiver is used. In general, an incoming signal would be
mixed with a local oscillator (LO) in a 90° hybrid as shown in Figure 1.4. The mixed
Polarization-X
Polarization-Y
Re
Im
Im
Re
Wavelength
1
Re
Im
Im
Re
Wavelength
2
21
signals then will be detected by using photodiodes for the I/Q information on the X/Y
polarizations. In the coherent receiver, digital signal processing is implemented to
measure the error-vector magnitude (EVM) as a quality metric of the optical
signal [26].
Figure 1.4 Coherent signal detection using the coherent receiver.
1.2.3 Optical Frequency Comb
Optical frequency combs provide multiple frequency comb-lines which are coherent
and phase/frequency locked with each other. Furthermore, each comb-line has narrow
linewidth, Fig 1.5. These features enable the manipulation of the amplitude and phase
for each individual comb-line and then processing of multiple comb-lines together
without significant phase noise degradation. At extremely high rates, different comb-
lines can be combined/mixed in a nonlinear optical element to implement different
optical signal processing functions [27].
Figure 1.5 Optical frequency comb features.
90
o
Hybrid
Local
Oscillator
Signal In
I(t)
Q(t)
E
S
(t) = E
S
(t) e
j(ω
S
t +
S
(t))
Analog-Digital
Converter
E
LO
Carrier Phase / Polarization
DSP Unit
Carrier Recovery
CD
-1
PMD
-1
PDL
-1
Nonlinearities
Goals:
➢ Low noise
➢ Broadband
➢ High coherence
➢ Flat spectrum
Optical Frequency Comb
Tunable
cw laser
22
Optical frequency combs feature multiple comb lines that are equidistant in the
frequency domain and are phase-coherent with respect to each other. In this thesis, we
use a comb source and select multiple narrow-linewidth coherent comb lines instead
of utilizing separate lasers to generate and process multiple signals. The complexity of
systems utilizing a bank of lasers are considerably reduced using comb source lines.
1.3 Basic Enabling Operations
In this section, basic operations as fundamental blocks for optical signal processing
system will be reviewed. These functions will frequently be used in the following
chapters.
1.3.1 Wavelength Conversion
Wavelength conversion can be achieved using FWM in a HNLF or silicon waveguide
or the cascaded 𝜒 (2)
::𝜒 (2)
processes of sum frequency generation followed by
difference frequency generation (cSFG-DFG) in a PPLN waveguide. These wave-
mixing interactions are governed by the two rules: conservation of energy and
conservation of momentum. As shown in Fig. 1.6, in the cSFG-DFG wavelength
conversion process, the input signal (at 𝑓 𝑠𝑖𝑔𝑛𝑎𝑙 ) and a continuous wave pump laser (at
𝑓 𝑝𝑢𝑚𝑝 ) are symmetrically placed around the quasi-phase matching (QPM) frequency
of the PPLN waveguide. As a result, the phase matching condition for SFG is satisfied
and the signal will be first copied to the frequency 𝑓 𝑠𝑖𝑔𝑛𝑎𝑙 + 𝑓 𝑝𝑢𝑚𝑝 = 2𝑓 𝑄𝑃𝑀 . This new
SFG signal can also be used in another nonlinear process to generate a new signal copy
at a frequency close to the original frequency. To achieve this, a dummy laser (at
fdummy) can also be sent into the PPLN waveguide, in which it can interact with the
SFG signal and produce a copy at 𝑓 𝑐𝑆𝐹𝐺 −𝐷𝐹𝐺 = 𝑓 𝑠𝑖𝑔𝑛𝑎𝑙 + 𝑓 𝑝𝑢𝑚𝑝 –𝑓 𝑑𝑢𝑚𝑚𝑦 through the
DFG process. According to phase matching condition the electrical field of the signal
copy is proportional to 𝐸 𝑠𝑖𝑔𝑛𝑎𝑙 𝐸 𝑝𝑢𝑚𝑝 𝐸 𝑑𝑢𝑚𝑚𝑦 ∗
, in which (.)
∗
denotes the complex
conjugate. This multiplication can enhance the phase noise and linewidth of the signal
23
copy. As shown in Fig. 1.6(c), if phase coherent comb-based lasers are used, the phase
noise of the signal copy will remain the same as the original signal and thus cascading
wavelength conversion stages does not worsen the phase noise significantly.
Figure 1.6 (a) Wavelength conversion in a PPLN waveguide, (b) pump configurations, (c) amplitude
and phase of the generated signals in the cSFG-DFG processes.
1.3.2 Wavelength Multicasting
Optical wavelength multicasting similar to wavelength conversion could be realized
by using nonlinearities to create multiple copies of the input data signal at different
wavelengths [28,29]. There are different methods based on different materials,
nonlinear processes, numbers of pumps and pump wavelength configurations for
multicasting [1]. Figure 1.7 shows conceptual spectra for two multicasting
techniques [1,30]. In Figure 1.7, to generate N signal copies, the input data signal is
sent to a nonlinear device with N discrete pumps to create N copies of the input signal
in a degenerate FWM process. In the literature, there are many examples of optical
multicasting. For example, the results of multicasting 16-QAM signals in a PPLN
waveguide were reported in [29]. Moreover, similar to the wavelength conversion,
various configurations can provide either phase conjugated or non-phase conjugated
copies.
(2)
SFG
f
signal
f
dummy
f
SFG
QPM
(2)
DFG
f
cSFG-DFG
f
pump
…
Signal
Copy
Original
Signal
Electrical Fields
(Phase Matching)
Frequencies
(Energy Conservation)
Comb-Based Lasers
Phase Noise
(φ
SFG
= φ
signal
= φ
pump
≡ φ
comb
)
Sum Frequency
Generation (SFG)
E
SFG
~ E
signal
E
pump
φ
SFG
~ φ
signal
+ φ
pump
f
SFG
= f
signal
+ f
pump
φ
SFG
~ 2φ
comb
Difference Frequency
Generation (DFG)
E
DFG
~ E
SFG
(E
dummy
)*
φ
DFG
~ φ
SFG
– φ
dummy
f
DFG
= f
SFG
– f
dummy
φ
DFG
~ φ
comb
Cascaded SFG-DFG
E
cSFG-DFG
~ E
signal
E
pump
(E
dummy
)*
φ
cSFG-DFG
~ φ
signal
+ φ
pump
– φ
dummy
f
cSFG-DFG
= f
signal
+ f
pump
– f
dummy
φ
cSFG-DFG
~ φ
comb
Dummy
Signal
Signal Copy
PPLN Waveguide
Pump
(a) (b)
(c)
24
Figure 1.7 Various configurations for N-fold signal multicasting using multi-pumps.
In another technique, signal multicasting can be realized by utilizing a frequency
comb. The input signal will be copied onto multiple wavelengths in a process known
as wavelength multicasting. Wavelength multicasting can be achieved using FWM in
a HNLF or the cascaded processes of sum frequency generation followed by difference
frequency generation in a PPLN waveguide. In the cSFG-DFG process, the input
signal (at fs) and a continuous wave (CW) pump (at fP1) are symmetrically placed
around the quasi-phase matching (QPM) frequency of the PPLN waveguide. As a
result, the phase matching condition for SFG is satisfied and the signal will be first
copied to the frequency fSIG+fP1=2fQPM. The new signal can also be used in a nonlinear
process to generate multiple new copies at wavelengths close to the original frequency.
To achieve this, an optical frequency combs (frequency lines at fDi’s) can also be sent
into the PPLN waveguide where each of them will interact with the SFG signal and
produce multiple copies at fCi=fSIG+fP1–fDi through the DFG process. Therefore, each
copy (fCi) and its corresponding comb finger (fDi) will also be symmetric around the
QPM frequency. Thus, a set of signal replicas can be obtained. The power of each
copy is proportional to the power of its corresponding comb finger. Moreover, the
phase of each copy is the summation of its corresponding comb finger and the original
signal. Since all comb fingers are coherent and all signal replicas are also phase locked.
Requires N probe pumps for N-
fold multicasting
f
Signal
Requires (N+1)/2 pumps for N-
fold multicasting
Signal
f
Multi-Pump multicasting configurations
25
Figure 1.8 Illustration of signal multicasting of a signal onto multiple frequencies using coherent
frequency comb.
1.3.3 Optical Multiplexing
There have been different techniques for optical multiplexing of WDM data channels
into a single channel. These include but not limited to the use of cross phase
modulation (XPM) in HNLFs [31], FWM in HNLFs and waveguides [22,32], and
cSFG-DFG in PPLN devices [33].
In the coherent multiplexing, optical frequency comb as well as the same cSFG-
DFG process that was previously used to multicast the signal to multiple copies can
be used to multiplex multiple coherent signals to a single signal. The phase locked
signal and their comb fingers are sent into a PPLN waveguide along with a CW pump
laser (at fP2≈fP1). Therefore, each signal mixes with its corresponding phase-
manipulated comb fingers through the SFG process and gives rise to a new signal at
frequency fCi+fDi=2fQPM. The multiplexed signal at 2fQPM will then interact with the
added pump laser at fP2 via the DFG process and create a multiplexed signal at original
frequency fSIG= 2fQPM-fP2. The coherency of original signals as well as comb fingers
are necessary to realize coherent addition.
Optical
Frequency
Comb
Nonlinear
Medium
Amplitude
Programmable
Filter
Input
Signal
Coherent
Copies
α
1
α
2
α
3
l
f
f
C3
f
C2
f
C1
f
D1
f
D2
f
D3
f
s
QPM
f
26
Figure 1.9 Optical coherent multiplexing of coherent signals using optical frequency combs.
As an example, we will describe in the following how a higher order QAM signal
generation can be realized by using coherent multiplexing building block [33].
Quadrature-amplitude-modulation (QAM) format has captured significant interest due
to its spectral efficiency compared with other modulation formats. Conventional
method for generating optical quadrature amplitude modulation (QAM) format
includes a pair of arbitrary waveform generators (AWG) to synthesize multilevel
electrical signals –one for in-phase and one for quadrature –, two linear RF amplifiers
and an I/Q modulator. Two multilevel electrical signals are modulated on the CW laser
using the I/Q modulator. Although it is possible to generate extremely large QAM with
this method (e.g., 512-QAM, 1024-QAM), the maximum baud rate is quite limited (<
6 Gbaud). AWG bandwidth, linearity of the RF amplifiers, and the linear range of the
optical field transfer function in the modulator limit the size and baud rate of generated
QAM.
Optical
Frequency
Comb
Nonlinear
Medium
Amplitude
Programmable
Filter
Coherent
Multiplexed
signal
α
1
α
2
α
3
l
f
f
C3
f
C2
f
C1
f
D1
f
D2
f
D3
f
s
QPM
f f
C3
f
C2
f
C1
BPF
f
f
s
Coherent
signals
27
Figure 1.10 Conventional method for generating optical QAM.
QPSK (4-QAM) can be generated at high baud rate (e.g., 50Gbaud) since the I/Q
modulator needs to be derived by only a pair of binary electrical signals which can be
synthesized simply at ~ 50Gbaud. Higher-order QAM, on the other hand, can be
potentially generated by combining multiple QPSK signals with appropriate amplitude
and phase weights. (Fig. 1.11) For instance, in order to generate a 64-QAM signal,
three QPSK signals with 1, 1/2 and 1/4 amplitude weights are needed to be combined.
When a QPSK signal is added, each constellation point of original QAM becomes four
points. Therefore, in general, n QPSK signals can be combined to generate 4
n
-QAM
(Fig. 1.11).
Figure 1.11 Generating optical QAM through QPSK modulated signals.
Since mathematical addition is occurred in optical domain, the original QPSK
signals are needed to be phase-coherent. The coherency between the fingers of a
narrow linewidth optical frequency comb, can potentially be used in an efficient
coherent multiplexing process. The optical interaction between multiple fingers of a
CW
Laser
I/Q Modulator
MZM-I
MZM-Q
Electrical
Amplifier
Electrical
Amplifier
Multi-level
Electrical Signal
I:
Q:
8-level Signal
8-level Signal
Optical QAM
64-QAM
Multi-level
Electrical Signal
π/2
+ +
=
1
st
QPSK
2
nd
QPSK
3
rd
QPSK
QPSK Mod.
Coherent
Multiplexing
Higher-order
QAM
• Each constellation point becomes four points.
• n QPSK signals generate 4
n
-QAM
QPSK Mod.
QPSK Mod. 1
1/2
1/4
Σ
28
frequency comb and corresponding lower-level signals in a nonlinear device can
tunably generate the data constellation points of a higher-level QAM. First, a set of
coherent and narrow linewidth fingers located at different frequencies (i.e., S1, S2 and
S3) are modulated to an QPSK format separately using conventional method. To
generate a QAM of size 4
n
, the amplitude of signals are need to be adjusted to 1, 1/2,
1/4,…, 1/2
n-1
. The relative phases between the signals are compensated by adding
appropriate constant phase on each signal. These QPSK signals along with another set
of comb fingers located at D3, D2 and D1 are injected into a nonlinear device -e.g.,
periodically-poled-lithium-niobate (PPLN) or highly-nonlinear-fiber (HNLF). Since
these two sets of comb fingers are equidistance, each signal interacts with its
corresponding comb finger. All product terms of these wave mixing processes land on
the same frequency. Since the original comb finger sets are coherent, the output signal
is thus the coherent multiplexed version of the original QPSK signals.
Figure 1.12 An example of coherent multiplexing in a nonlinear device to generate higher order
amplitude and phase formats.
1.3.4 Conversion-Dispersion Optical Delays
The chromatic-dispersion-based delays exploit the wavelength-dependent speed of
light in a dispersive medium coupled with tunable wavelength conversion to achieve
continuously tunable delays, as shown in Fig. 1.13 [1].
Amplitude Weights:
[ 1 , 0.5 , 0.25 ]
Frequency Comb-1
f
QPSK
Modulated
Frequency Comb-2
f
e.g.,
64-QAM
f
CW
Comb
f
Modulated
Comb
Pump
Nonlinear Device
(Phase Coherent
Addition)
QPSK
Modulation
S
1
S
2
S
3
D
3
D
2
D
1
P
• Nonlinear device multiplies each QPSK signal with a comb finger and the pump.
• All three multiplication signal land on the same frequency.
• Comb-based signals are phase-coherent and vector addition happens.
• Vector addition can be used to generate complex QAM symbols from basic QPSK symbols.
QAM
29
Figure 1.13 Tunable Conversion/dispersion based optical delay.
This technique consists of the following blocks:
Wavelength Conversion: The original signal is wavelength converted using optical
nonlinear effects, such as FWM in an HNLF or cSFG-DFG in a PPLN waveguide. The
final wavelength depends on the choice of the input pump lasers that mix with the
original signal. If a comb source is used for pump lasers, appropriate fingers of the
comb can be selected to generate a wavelength converted copy of the signal at any
wavelength.
Dispersion Induced Delay: Once a wavelength converted copy of the signal is sent
into a highly dispersive element, a relative delay corresponding to each wavelength
will be induced on it. The signal can then be wavelength converted back to the original
wavelength. The use of dispersive element could lead to intra-channel dispersion (i.e.
dispersion over the bandwidth of the converted signal). If the delay/dispersion is small,
this dispersive element does not significantly distort the signal, if not, it can be
dispersion-compensated afterward [1].
Dispersive
Element
t
Δτ
Δτ
Wavelength
Conversion #2
λ
in
λ
1
λ
2
λ
1
λ
2 λ
in
λ
in
t
t t
Wavelength
Conversion #1
Relative Group Delay (s)
λ
1
λ
2
Δλ
Δτ
Wavelength dependent delay via chromatic dispersion Δτ = D Δλ
30
Chapter 2 All Optical Phase Noise Mitigation
2.1 Introduction
Advanced modulation formats such as quadrature-phase-shift-keying (QPSK) are
gaining importance in high capacity optical networks [4,5]. Since both amplitude and phase
are modulated in QPSK systems, phase or amplitude noise can cause a degradation in optical
signal-to-noise ratio resulting in a system power penalty. Specifically, phase noise which is
originally Self phase modulation (SPM) of the ASE noise because of the Kerr nonlinearity,
can pose a key limitation in such systems [37–39].
Nonlinear phase noise of a high data rate signal can be reduced by taking advantage of
electronic-based parallel data-processing techniques [39,40]. However, there might be
advantages to mitigate the phase noise in the optical domain, such as avoiding the impact of
optical-to-electronic conversion and supporting in-line signal processing for high baud rate
signal [40]. Different approaches have been demonstrated for optical regeneration of QPSK
signals using different variations of phase-sensitive-amplification to achieve a level of “phase
squeezing” [15,41,42]. However, these methods tend to require coherency between a signal
and a pump, typically requiring phase-based feedback loops and injection locking lasers [41–
45]
In this chapter, we experimentally demonstrate a baud-rate-tunable phase noise
mitigation scheme. Phase noise is mitigated in an all-optical phase quantizer by combining the
signal, the conjugate copy and their delayed variant third harmonics. In the proposed method,
the signal and the harmonics are automatically phase-locked multiplexed, avoiding the need
for phase-based feedback loops and injection locking to maintain the coherency between
different components of the quantization process.
Phase noise might be partially converted to amplitude noise in the quantizer stage [15].
Such converted amplitude noise can be compensated, by means of an amplitude limiter.
Therefore, a combined phase/amplitude noise mitigation can further improve the system
performance.
31
2.2 Concept
The conceptual block diagram of the proposed approach for the phase noise mitigation
is shown in Fig. 2.1. A QPSK signal contaminated with phase noise is coupled with a
CW pump, and injected into a nonlinear wave mixer to generate a phase conjugate
copy of the signal [46]. A nonlinear element with either second or third-order
nonlinear susceptibility (𝜒 (2)
or 𝜒 (3)
), can be used in this stage to generate the copy.
The signal and the conjugate copy are then sent into a 𝜒 (3)
medium to generate the
third-order harmonics of the signal and the conjugate copy. In fact, these two stages
can be combined in one stage by using a 𝜒 (3)
element with high amount of nonlinear
efficiency. The signal, its third-order harmonic, and their conjugates are sent into an
optical programmable filter to apply appropriate delays and adjust amplitudes and
relative phases. The adjusted signals and harmonics with a CW pump are injected into
a 𝜒 (2)
medium to create a staircase phase transfer function. In this stage, the product of
the signal and its delayed conjugate, and the product of the delayed third-order signal
harmonic and the conjugate third-order harmonic are phase-locked multiplexed. This
nonlinear process builds a staircase phase transfer function of the input signal, which
results in squeezing of the phase noise of the input signal. The residual phase noise
which is not squeezed in the phase quantization process converts to amplitude noise
due to the non-uniform amplitude profile of the staircase phase transfer function [15].
The amplitude noise is suppressed by utilizing an optical parametric amplifier operated
in the saturation regime. If needed, the output signal can be wavelength converted back
to the input signal wavelength using common all-optical wavelength conversion
methods.
32
Figure 2.1 The concept of the phase noise mitigation scheme including four cascaded nonlinear stages
for conjugate generation, third order harmonics generation, phase quantization, and parametric
amplification in saturation regime.
In the following, the mathematical descriptions of nonlinear stages to create the phase
quantization function are presented.
A periodically-poled lithium niobate (PPLN) waveguide is used in the first nonlinear
stage. The wavelength of the first pump laser 𝑃 1
(𝑡 ) is set to the quasi-phase matching (QPM)
wavelength of PPLN-1. This pump interacts with itself through a second-harmonic-generation
process, and creates the mixing term 𝑃 1
2
(𝑡 ) at 2𝜔 𝑃 1
. The SHG term then mixes with the
input signal 𝑆 𝑖𝑛
(𝑡 ) with frequency 𝜔 𝑖𝑛
through the difference-frequency-generation (DFG)
process to create a phase conjugate replica at 2𝜔 𝑃 1
− 𝜔 𝑖𝑛
with the electric field proportional
to 𝑃 1
2
(𝑡 )× 𝑆 𝑖𝑛
∗
(𝑡 ) . The PPLN-1 output is sent into an in-line spatial light modulator (SLM)
to filter out the pump. The signal and its conjugate are mixed through a four-wave-mixing
(FWM) process in a highly nonlinear fiber (HNLF) to generate the third-order harmonics at
−2 𝜔 𝑃 1
+ 3 𝜔 𝑖𝑛
and 4 𝜔 𝑃 1
− 3 𝜔 𝑖𝑛
. The electric field of the generated third-order
harmonics are proportional to 𝑃 1
∗
2
(𝑡 )× 𝑆 𝑖𝑛
3
(𝑡 ) and 𝑃 1
4
(𝑡 )× 𝑆 𝑖𝑛
∗
3
(𝑡 ). In a liquid crystal
on silicon (LCoS) programmable filter, one symbol delay(𝑇 𝑠 )is applied between: (i) the signal
and its conjugate copy, and (ii) the generated third-order harmonics. In addition, amplitudes
and relative phases are adjusted in the LCoS. In the second PPLN waveguide, the signal and
Nonlinear Element
(PPLN-1)
Nonlinear Element
(HNLF-1)
Amplitude/
Phase
Filter
Nonlinear Element
(PPLN-2)
Phase Conjugate
Generation
Adjusting
Delays,
Coefficients
-order
Harmonics
Generation
Phase-Locked
Multiplexing
+
Phase Quantization
Noisy QPSK
Signal
-Φ
s
Φ
s
ω
P1
3Φ
s
-3Φ
s
Four Wave Mixing
ω
P1
SHG+ DFG Mixing
P
1
-Φ
s
Φ
s
-Φ
s
(t-T
s
) Φ
s
(t)
ω
P1
3Φ
s
(t-T
s
) -3Φ
s
(t)
P
2
SFG Mixing
DFG Mixing
ω
out
ω
P2
Higher Harmonics Generation
Nonlinear Element
(HNLF-2)
Parametric
Amplification
(Saturation
Regime)
Phase-
Quanitzed
ω
P3
P
3
ω
out
Output QPSK
PPLN-1: HNLF-1: PPLN-2: HNLF-2:
to
HNLF-2
QPM
P
1
P
2 P
3
33
its delayed conjugate are mixed through a sum-frequency-generation (SFG) nonlinear process
and generate a new signal at 2𝜔 𝑃 1
with the electric field of 𝐸 1
(𝑡 )proportional to
𝐸 1
(𝑡 ) ∝ 𝑃 1
2
(𝑡 − 𝑇 𝑠 )× [𝑆 𝑖𝑛
(𝑡 )× 𝑆 𝑖𝑛
∗
(𝑡 − 𝑇 𝑠 )] (2.1)
Similarly, in a parallel process, the third-order harmonics are mixed through an SFG
process and generate a new signal at 2𝜔 𝑃 1
with the electric field of 𝐸 2
(𝑡 ) proportional to
𝐸 2
(𝑡 ) ∝ 𝑃 1
∗
2
(𝑡 − 𝑇 𝑠 )× 𝑃 1
4
(𝑡 )× [𝑆 𝑖𝑛
(𝑡 − 𝑇 𝑠 )× 𝑆 𝑖𝑛
∗
(𝑡 )]
3
(2.2)
The challenge of combining the signals 𝐸 1
(𝑡 ) and 𝐸 2
(𝑡 ) at 2𝜔 𝑃 1
results from the fact
that they need to be phase-locked and have the same phase reference. According to Eqs. (1)
and (2) the phases of 𝐸 1
(𝑡 ) and 𝐸 2
(𝑡 ) can be expressed as:
Φ
𝐸 1
(𝑡 ) = 2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 )+ ∆Φ
𝑖𝑛
(𝑡 ) (2.3)
Φ
𝐸 2
(𝑡 ) = −2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 )+ 4Φ
𝑃 1
(𝑡 )− 3 ∆Φ
𝑖𝑛
(𝑡 )
= 2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 )+ 4∆Φ
𝑃 1
(𝑡 )− 3 ∆Φ
𝑖𝑛
(𝑡 ) (2.4)
where Φ
𝑖𝑛
(𝑡 ) and Φ
𝑃 1
(𝑡 ) refer to the phases of the input signal and the pump,
respectively. Moreover, ∆ is defined as a difference operator for one symbol interval 𝑇 𝑠 ,
i.e.,∆Φ(𝑡 ) ≜ Φ(𝑡 )− Φ(𝑡 − 𝑇 𝑠 ). In order to study the coherence of 𝐸 1
(𝑡 ) and 𝐸 2
(𝑡 ), the
phase noise of the input signal, Φ
𝑖𝑛
𝑁 (𝑡 ), and the phase noise of the pump, Φ
𝑃 1
𝑁 (𝑡 ), need to
be considered in Eqs.(3) and (4), except for the first similar term 2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 ).
Φ
𝐸 1
(𝑡 ) = 2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 )+ ∆Φ
𝑖𝑛
𝐷 (𝑡 )+ ∆Φ
𝑖𝑛
𝑁 (𝑡 ) (2.5)
Φ
𝐸 2
(𝑡 ) = 2Φ
𝑃 1
(𝑡 − 𝑇 𝑠 )+ 4∆Φ
𝑃 1
𝑁 (𝑡 )+ ∆Φ
𝑖𝑛
𝐷 (𝑡 )− 3 ∆Φ
𝑖𝑛
𝑁 (𝑡 ) (2.6)
where Φ
𝑖𝑛
𝐷 (𝑡 ) refers to the QPSK data and ∆Φ
𝑖 𝑛 𝐷 (𝑡 ) = Φ
𝑖𝑛
𝐷 (𝑡 )− Φ
𝑖𝑛
𝐷 (𝑡 − 𝑇 𝑠 )
represents the encoded version of the original signal and has the same format as differential
phase shift keying modulation. Eq.2.6 uses the fact that for QPSK data exp(−𝑗 3Φ
𝑖𝑛
𝐷 (𝑡 )) =
exp(𝑗 Φ
𝑖𝑛
𝐷 (𝑡 )); and the term −3∆Φ
𝑖𝑛
𝐷 (𝑡 ) is replaced by ∆Φ
𝑖𝑛
𝐷 (𝑡 ).
Assuming that the phase noise of the pump, Φ
𝑃 1
𝑁 (𝑡 ), stems from the laser linewidth
(Δ𝜐 ), and with a fluctuation which is significantly slower than the symbol rate (𝛥𝜐 ≪ 1/𝑇 𝑠 ),
we conclude that ∆Φ
𝑃 1
𝑁 (𝑡 ) = Φ
𝑃 1
𝑁 (𝑡 )− Φ
𝑃 1
𝑁 (𝑡 − 𝑇 𝑠 ) is negligible.
By performing Fourier transform of ∆Φ
𝑃 1
𝑁 (𝑡 ), the result is:
34
ℱ(∆Φ
𝑃 1
𝑁 (𝑡 )) ≜ 2j exp (−𝑗𝜔𝑇 /2)sin(
ω𝑇 𝑠 2
) × Φ
𝑃 1
𝑁 (𝜔 ) (2.7)
Because the power spectral density (PSD) of the differentiator,|𝑠𝑖𝑛 (𝜔 𝑇 𝑠 /2) |
2
,rejects
the lower frequency components of Φ
𝑃 1
𝑁 (𝜔 ) (as shown in Fig. 2.2), the PSD of the
∆Φ
𝑃 1
𝑁 (𝑡 ) contains almost no power assuming 𝛥𝜐 ≪ 1/𝑇 𝑠 .
Figure 2.2 The PSD of the phase-differentiator.
A CW pump with an electric field of 𝑃 2
(𝑡 ) is injected into the PPLN-2 waveguide to
convert the SFG signal at 2 𝜔 𝑃 1
to the new signal 𝑆 𝑜𝑢𝑡 (𝑡 ). The electric field of 𝑆 𝑜𝑢𝑡 (𝑡 ) is
proportional to:
𝑆 𝑜𝑢𝑡 (𝑡 ) ∝ 𝑒 𝑗 2Φ
𝑃 1
(𝑡 )
𝑒 𝑗 ∆Φ
𝑖𝑛
𝐷 (𝑡 )
[𝑒 𝑗 ∆Φ
𝑖𝑛
𝑁 (𝑡 )
+ 𝑚 𝑒 −𝑗 3∆Φ
𝑖𝑛
𝑁 (𝑡 )
] (2.8)
where the value of 𝑚 is defined as 𝑚 ≜ |𝐸 2
|/|𝐸 1
|, and can be adjusted in the LCoS
filter [10]. The term inside the brackets in Eq. 2.8 demonstrates the function of
squeezing the phase noise of the input signal. If the value of 𝑚 in the squeezing
function is set to zero by bypassing the higher order harmonics generation stage in Fig.
2.1, the term inside the brackets in Eq. 2.8 is simplified to exp(𝑗 ∆Φ
𝑖𝑛
𝑁 (𝑡 )). This
function is a differentiator on the phase domain which can mitigate phase noise with a
low-bandwidth PSD; such as phase noise coming from laser linewidth. As mentioned
before, ∆Φ
𝑖𝑛
𝐷 (𝑡 ) = Φ
𝑖𝑛
𝐷 (𝑡 )− Φ
𝑖𝑛
𝐷 (𝑡 − 𝑇 𝑠 ) represents the encoded version of the
original signal and has the same format as differential phase shift keying modulation.
If we need to recover the original data signals in the receiver side, the effects of
PSD
35
differentiation has to be undone in either the transmitter, receiver or an intermediate
device [47]. The inversion can be done before or after the differentiation which
respectively are referred to pre-processing and post-processing functions in the
system [47]. A system with differentiation process is affected differently in terms of
the error and in general differential processing makes every input side error into two
successive output side errors [47]. Note that pre-processing does not cause error
propagation in the system. On the other hand, post-processing results in error
propagation in the system until resynchronization is occurred [47]. Thus, pre-
processing is preferred for compensating the effects of the differential processing in
general [47].
2.3 Experimental Setup
Figure 2.3 shows the experimental setup for verifying the phase noise mitigation
scheme. A 20/30-Gbaud QPSK data (PRBS 2
31
-1) is generated using a CW laser at
1550 nm. The signal is phase modulated with an ASE source followed by a variable
optical attenuator (VOA) and a photo-diode (PD) to induce phase noise with different
power levels. The noisy signal is amplified to 23 dBm and coupled with a 26 dBm
amplified CW pump around 1551.5nm and injected into PPLN-1 waveguide. A ~450m
HNLF-1 (ZDW at 1551.5nm) is used to generate the third-order harmonics. A 23 dBm
amplified CW pump at 1543nm is injected into the PPLN-2 waveguide to multiplex
the signals. The generated signal is amplified to ~10 dBm and coupled with a 30 dBm
amplified CW pump, around 1556.3nm, and injected into a 700m dispersion stable
HNLF-2 (ZDW at 1551.5 nm). The pump is phase modulated with 4.5 GHz PRBS
(2
15
-1) data to suppress stimulated Brillouin scattering.
36
Figure 2.3 Experimental setup of phase and amplitude noise mitigation. PM: Phase modulator, HNLF:
Highly nonlinear fiber, PPLN: Periodically poled Lithium Niobate, SLM: Spatial light modulator, PD:
Photo-detector, VOA: Variable optical attenuator.
2.4 Results
Figure 2.4 shows measured gain and power profiles of HNLF-2. A quite wide range of
signal power from about 0 to 10dBm can be squeezed within 0.4dB deviation. Finally, in order
to measure the phase noise range, EVM, and BER of the system, the output signal is captured
by a coherent detector, without applying any DSP equalizer.
Figure 2.4 Power and gain profiles of HNLF-2. Operation in saturation regime results in amplitude
noise squeezing.
The system performance is assessed using 20 and 30-Gbaud QPSK signals. Figure 2.5
shows the constellation diagrams of the 20-Gbaud input noisy signal and the output of the
phase quantizer. The results are obtained for different values of phase noise. In order to
90
λ
s
2 nm 1 nm
VOA
Broadband
ASE Source
λ
p1
1 nm
PPLN-1
Waveguide
RF Amp. PD
SLM
Filter
HNLF-1
5 nm
SLM
Filter
9 nm
λ
p2
1 nm
1 nm 1 nm
20 Gbaud PRBS (2
31
-1)
20 dBm
26 dBm
20 dBm
23 dBm
PPLN-2
Waveguide
PM
HNLF-2
λ
p3
1 nm
Coherent
Receiver
PM
AWG PPG
4.5GHz
PRBS (2
15
-1)
Phase Quantizer
Amplitude
Squeezer
RF Amp.
23 dBm
1542 1544 1546 1548 1550 1552 1554 1556 1558 1560
-80
-70
-60
-50
-40
-30
-20
-10
HNLF-1
Output
-Φ
s
Φ
s
3Φ
s -3Φ
s
1542 1544 1546 1548 1550 1552 1554 1556 1558 1560
-80
-60
-40
-20
0
PPLN-1
Output
-Φ
s
Φ
s
QPM
1540 1542 1544 1546 1548 1550 1552 1554 1556 1558 1560 1562
-80
-60
-40
-20
0
To
HNLF-2
3Φ
s
-Φ
s
Φ
s
-3Φ
s
PPLN-2 Output
QPM
1560 1551.5 1543
nm
1552.6
10 dB/D
1550.4
nm
1551.5 1553.7 1549.3
10 dB/D 10 dB/D
1548 1550 1552 1554 1556 1558 1560 1562 1564
-70
-60
-50
-40
-30
HNLF-2
Pump
Output
1560
nm
10 dB/D
1556 1552
0
5
10
15
20
25
30
5
8
11
14
17
20
23
-20 -15 -10 -5 0 5 10 15 20
Gain (dB)
Output power (dBm)
Signal's optical power into HNLF-2 (dBm)
37
compare the phase noise range between the constellation diagrams, the parameter δφ is
defined to quantify the phase deviation from the corresponding expected value, showing the
standard deviation of the phase. In addition, the parameter δρ is defined as the percentage of
amplitude deviation from the expected value, showing the relative standard deviation of the
amplitude. As can be seen in Fig. 2.5, phase noise is reduced by phase quantizer, in particular
for higher levels of noise. In Fig. 2.5(a), δφ is reduced by ~49%. The value of δρ is, however,
increased by ~56%. This indicates that phase noise is partially converted to amplitude noise
in the phase quantizer.
Figure 2.5 The constellation diagrams of the input and output of phase quantizer for different levels of
phase noise.
The amplitude noise can be suppressed by utilizing a parametric amplification in
the saturation regime. Figure 2.6 shows the constellation diagrams of 30-Gbaud noisy
QPSK signals, and the corresponding outputs of the phase quantizer and the parametric
amplifier. As can be seen, phase noise and amplitude noise are decreased in the output
of parametric amplifier. In Fig. 2.6(a), δφ is reduced by ~40%, and δρ is reduced by
16% in the output of parametric amplifier compared to the input signal. As a result,
the EVM between the input signal and the final output is decreased by ~36% in Fig.
2.6(a). Similar results are obtained for the constellation diagrams at 20-Gbaud QPSK
signals in Fig 2.6(c,d).
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
180
150
120
60
300
330
30
270
240
210
δ
φ
=54.2° , δ
ρ
=25.5% δ
φ
=27.9°, δ
ρ
=39.8% δ
φ
=49.9°, δ
ρ
=25.4%
EVM=29.7% EVM=22.6% EVM=25.9%
100%
0%
100%
0%
δ
φ
=37.3°, δ
ρ
=22.8% δ
φ
=23.1, δ
ρ
=35.7% δ
φ
=31°, δ
ρ
=21.3%
EVM=17.8% EVM=15.2% EVM=15.9%
δ
φ
=26.2°, δ
ρ
=39% δ
φ
=44.8°, δ
ρ
=23% δ
φ
=24.9°, δ
ρ
=37.2%
EVM=20.7% EVM=22.5% EVM=17.9%
δ
φ
=21.2°, δ
ρ
=36.8% δ
φ
=28°, δ
ρ
=20.1% δ
φ
=19.8°, δ
ρ
=35.2%
EVM=14.4% EVM=14.5% EVM=14%
Input Noisy Signal Phase Quantizer Output Input Noisy Signal Phase Quantizer Output Input Noisy Signal Phase Quantizer Output
a)
d)
b)
e)
100%
0%
100%
0%
100%
0%
100%
0%
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
180
150
120
60
300
330
0
30
270
240
210
0
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
f)
c)
38
Figure 2.6 The input and output constellation diagrams of the phase noise mitigation systems for 30
Gbaud (a, b) and 20 Gbaud (c, d) QPSK signals impaired by different values of phase noise.
Figure 2.7(a) shows the percentage of phase noise reduction for 30-Gbaud input
noisy QPSK signal, in the output of phase noise quantizer and the output of parametric
amplifier. The amount of reduction is increased for higher levels of phase noise and it
becomes gradually constant for the phase noise with δφ > 50°. Moreover, the amount
of phase noise reduction in the parametric amplifier output is less than the amount of
reduction in the phase quantizer output. This fact indicates that the amplitude
saturation stage adds some phase noise to the output of the phase quantizer. Figure
2.7(b) shows the EVMs of the 30-Gbaud input noisy QPSK signal, and the
corresponding outputs of the phase quantizer and the parametric amplifier for various
levels of phase noise. The EVM is improved in the system output;parametric amplifier
output, and remained between 13% to 18% for different levels of phase noise.
δ
φ
=45.1° , δ
ρ
=21.6% δ
φ
=22.2°, δ
ρ
=37.5% δ
φ
=27.5°, δ
ρ
=18.2%
EVM=22.5% EVM=17.7% EVM=14.3%
v v v
90
180
150
120
60
300
330
30
270
240
210
100%
0%
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
v v v
90
180
150
120
60
300
330
30
270
240
210
100%
0%
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
δ
φ
=36°, δ
ρ
=20.4% δ
φ
=19.9°, δ
ρ
=38.6% δ
φ
=24.8°, δ
ρ
=18.2%
EVM=18.5% EVM=16.1% EVM=13.8%
v v v
90
180
150
120
60
300
330
30
270
240
210
100%
0%
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
δ
φ
=42.3°, δ
ρ
=18.2 δ
φ
=23.1°, δ
ρ
=25% δ
φ
=25.5°, δ
ρ
=17.8%
EVM=20.3% EVM=17.4% EVM=14.1%
v v v
90
180
150
120
60
300
330
30
270
240
210
100%
0%
v v v
90
150
120
60
300
330
30
270
240
210
v v v
90
150
120
60
300
330
0
30
270
240
210
δ
φ
=34.1°, δ
ρ
=19.3% δ
φ
=20°, δ
ρ
=24.8% δ
φ
=24.5°, δ
ρ
=20.4%
EVM=16.7% EVM=14.8% EVM=13.5%
30 Gbaud QPSK Signal
Input Noisy Signal Phase Quantizer Output Amplitude Squeezer Output
20 Gbaud QPSK Signal
Input Noisy Signal Phase Quantizer Output Amplitude Squeezer Output
a)
b)
c)
d)
39
Figure 2.7 (a) Percentage of phase noise range reduction for various levels of phase noise for 30 Gbaud
QPSK signal. (b) EVM of the input noisy signal, phase quantizer output, and parametric amplifier
output for various levels of phase noise for a 30 Gbaud QPSK signal.
Figure 2.8 shows the BER curves of the phase noise mitigation scheme for two
different levels of phase noise, δφ ~43° and ~35°. Near 1.5 dB OSNR gain is achieved
in the phase quantizer output at BER of 10
-3
. By suppressing the amplitude noise in
the parametric amplifier, near 3 dB OSNR gain is achieved at BER 10
-3
.
Figure 2.8 BER versus OSNR for two different phase noise levels (δφ ~43°, ~35°).
EVM%
δ
Ф
decrease %
a)
20 25 30 35 40 45 50 55
10
15
20
25
30
Input noisy siganl
Phase quantizer output
Parametric amplifier output
30 35 40 45 50 55
20
30
40
50
60
Phase quantizer output
Parametric amplifier output
b)
δ
Ф
Input signal (degree) δ
Ф
Input signal (degree)
1
9 11 13 15 17 19
-log
10
(BER)
OSNR (dB)
Noise Level1_Input Noisy Signal
Noise Level1_Phase Quantizer Output
Noise Level1_Parametric Amplifier Output
Noise level2_Input Noisy Signal
Noise Level2_Phase Quantizer Output
Noise Level2_Parametric Amplifier Output
2
3
4
5
40
2.7 Conclusion
In conclusion, an all-optical nonlinear phase noise mitigation is demonstrated based on the
phase-locked multiplexing of signal harmonics and amplitude saturation which avoids the
need for phase-based feedback loops and injection locking. The phase locking process
converts the signal format to differential phase shift keying. Further studies are needed to
investigate a phase-locked method without format conversion. Moreover, here, we focused on
the QPSK signal contaminated only with the phase noise. For the signal with amplitude noise,
it might be possible to suppress the amplitude noise in an extra nonlinear stage using the
amplitude saturation method, and then mitigate the remained phase noise by utilizing the
proposed phase noise mitigation method.
41
Chapter 3 Simultaneous All-optical Phase
Noise Filtering and Automatically Locked Tunable
Homodyne Reception
3.1 Introduction
Homodyne systems have a long history of achieving highly sensitive reception in optical
communication systems [48–50]. Similar to coherent systems in general, homodyne systems
also enable the recovery and digital signal processing of phase-encoded data signals (e.g.,
phase-shift-keying, PSK). Such phase-encoded systems can also achieve high spectral
efficiency, such as in quadrature PSK (QPSK) and 16 quadrature-amplitude-modulation
(QAM) [48–50].
However, efficient PSK homodyne systems present difficult challenges, including: (1)
the need to phase and frequency lock the incoming signal to the local oscillator (LO). (2) the
need for phase noise mitigation, which is a key limitation for phase-based communication
systems [37–41]. There may be a desire to meet these two challenges using optical signal
processing.
Approaches for phase and frequency locking for homodyne reception include (a)
transmitting the carrier along with the data signal which occupies some part of the spectrum
or polarization state [51], (b) using a laser LO in the receiver coupled with a phase locked loop
(PLL) and signal processing algorithms to ensure locking [48], and (c) using optical signal
processing and an optical feedback loop [52]. Feedback loop approaches tend to be fairly
complex and require time to achieve a stable “lock” [53].
In terms of phase-noise, important sources of phase noise at the coherent receiver are
the laser linewidth of the transmitter laser and of the receiver local oscillator [37,54]. Note that
additional sources can include nonlinear phase noise created by amplified spontaneous
emission interacting with the nonlinear Kerr effect [37–41]. The traditional method to mitigate
the effect of laser linewidth is to use an optical or electrical phase-locked loop (PLL) that
42
synchronizes the frequency and phase of the local oscillator (LO) with that of the transmitter
laser. The PLLs are very sensitive to the propagation delay inside their feedback loop, which
can limit this method for a laser with higher linewidth [5]. Advances in high-speed very large-
scale integration enable high-speed and efficient signal processing algorithms to help partially
mitigate the laser-linewidth-induced phase noise. Compared to a PLL-based receiver, these
algorithms can be more tolerant to the laser phase noise since they use digital compensation
techniques [55,56]. A desirable goal might be to reduce the effect of linewidth-based phase
noise in coherent receivers through high-data-rate optical approaches, which might produce
higher speed and wider dynamic range performance. Using the optical method, the stringent
requirement on laser linewidth may be significantly lower than for other phase-recovery
approaches [46].
A laudable goal might be to attempt to meet simultaneously both these challenges of
automatic frequency/phase locking and linewidth-based phase-noise mitigation in a
homodyne system.
3.2 Concept
The conceptual block diagram of the homodyne phase noise mitigation scheme is
shown in Fig. 3.1. A QPSK signal, which is received on a laser with large line-width
or contaminated with phase noise, along with a CW pump, P
1
, are injected into a
nonlinear wave mixer to generate the conjugate copy of the original signal. The signal
and its conjugate copy are then sent into the delay stage to apply one bit delay between
them. The signals with a second CW pump are injected into a second nonlinear stage.
In this stage two tasks will be done simultaneously: 1) phase noise mitigation by
coherently mixing the product of the signal and its delayed conjugate and 2) phase-
locked multiplexing by the “local” laser 𝑃 1
. We denote the noise mitigated signal by
𝑅 (𝑡 ) ∝ 𝑒𝑥𝑝 (𝑗 Φ
𝑠 (𝑡 )− Φ
𝑠 (𝑡 − 𝑇 )). Since the phase of the original signal consists of
both data and the phase noise, then Φout (t) = [Φ
D
(t) − Φ
D
(t − T)] + [φ
N
(t) −
φ
N
(t − T)]. The first bracket is only a decoding of the original data. The second
bracket can be viewed as a filter on the phase noise which is able to filter out low
43
frequency components of the phase noise (e.g., laser phase noise). Phase-locked
multiplexing can be done by multiplexing the noise mitigated signal 𝑅 (𝑡 ) with the CW
pump, 𝑃 1
, with an appropriate relative complex coefficient adjusted in a
programmable filter. The final optical output can be denoted as 𝑆 𝑜𝑢𝑡 (𝑡 ) = 𝐿𝑂 +
𝑤𝑅 (𝑡 ), where 𝐿𝑂 ∝ 𝑃 1
2
. By sending the output to photo-diodes and setting 𝑤 to ±1 or
±j, similar to a 90
0
optical hybrid, both in-phase and quadrature components of the
noise mitigated input signal can be obtained.
Figure 3.1 The concept of the homodyne phase noise mitigation scheme. Low frequency components
of the phase noise (e.g., laser phase noise) of an incoming signal is filtered by multiplying the signal to
its delayed conjugate. Simultaneously, the noise mitigated signal is automatically locked to a “local”
laser, P
1
, and sent to a PD for detection.
3.3 Experimental Setup
The experimental setup is depicted in Fig. 3.2 (a). A nested Mach–Zehnder modulator
is used to modulate 20-Gbaud QPSK data (PRBS 2
31
-1) at ~1538 nm. The signal is
phase modulated with an ASE source to emulate phase noise with different powers
and bandwidths. The noisy signal is coupled with a CW pump around 1541 nm and
sent to a periodically-poled-lithium-niobate (PPLN-1) waveguide. The signals are then
sent to a ~25 m DCF and a SLM filter for adjusting the phases. The resulting signals are
amplified and coupled with another CW pump around 1530 nm and sent to PPLN-2
Phase noise mitigated output signal
(Phase-locked multiplexed with pump-1)
Phase Conjugate
Generation
(Nonlinear Element)
One-Symbol Delay
(Dispersive Element)
Phase/Freq. Locked
Multiplexing
+
Phase Noise
Filtering
(Nonlinear Element)
Photo Diode (PD)
-Φ
s
(t-T) Φ
s
(t)
P
1
ω
P1
P
2
SFG, SHG Mixing
DFG Mixing
ω
out
ω
P2
0 2000 4000 6000 8000 10000 12000 -100
-80
-60
-40
-20
0
20
40
60
80
100
0 2000 4000 6000 8000 10000 12000 -100
-80
-60
-40
-20
0
20
40
60
80
100
…
-π/4 π/4 -3π/4 π/4 π/4 -3π/4
…
…
π/4 -π/4 3π/4 -π/4 -π/4 3π/4
…
0 2000 4000 6000 8000 10000 12000 -100
-80
-60
-40
-20
0
20
40
60
80
100
…
π/2 -π π 0 -π
…
High-Speed
Input
Slow-Speed
Phase Noise, φ(t)
time
time
Phase-Conjugate, -φ(t-τ)
time
S
Out
(t)
φ
Out
(t) ~φ (t) -φ(t-τ)
Encoded Output
T
Reduced
Phase
Noise
Input Signal
Phase-Conjugate
& Delayed Copy Output
P
1
P
2
Signal on a laser with large line-width
(High phase-noise)
ω
P1
SHG+ DFG Mixing
P
1
-Φ
s
(t) Φ
s
(t)
Signal
conjugate,
S*(t)
Noisy
received
signal,
S(t)
S(t)
S*(t-τ)
S
Out
(t)
LO
44
waveguide to 1) coherently mix the signal and its conjugate copy, and, 2) to phase-
locked multiplex the first pump in the output. The multiplexed signal is then filtered,
amplified and sent to a photodiode.
Figure 3.2 (a) Experimental setup. DCF: Dispersion compensating fiber, SLM: Spatial light modulator
(b) Optical spectra of PPLN-1, PPLN-2.
3.4 Results
The performance of the system is assessed by implementing the proposed homodyne
phase noise mitigation scheme on 20-Gbaud QPSK signals which are degraded by
inducing phase noise at different powers and bandwidths. Both in-phase and
quadrature-phase components of the data signal are detected. Fig. 3.3(a) shows the I/Q
eye diagrams of the detected noise-mitigated signal for phase noise with 300MHz, 750
MHz and 3300 MHz spectral density bandwidths and for three different power levels.
As it can be seen, the phase noise reduction is significant especially for the phase noise
with lower bandwidth. Here, the reduction factor is defined as the ratio of input phase
noise standard deviation over the output phase noise standard deviation. In Fig. 3.3(b)
the performance of the system is assessed by measuring this reduction factor in the
three levels of phase noise and for different noise-bandwidths.
200 400 600 800 1000 1200 1400
-70
-60
-50
-40
-30
-20
-10
0
90
2 nm 1 nm
VOA
Broadband
ASE Source
λ
p1
1 nm
RF Amp. PD
20 dBm
23 dBm
PM
9 nm
20 dBm
1 nm
SLM
Filter
PPLN
Waveguide
20/32 Gbaud PRBS (2
31
-1)
Phase Conjugate
Generation
a)
λ
p2
20 dBm
Noise Mitigation+
Phase Locked Muxing
λ
SIG
b)
1538 1541 1544 nm
PD
10 dB/D
PPLN-1
Waveguide
DCF
LPF
250 300 350 400 450 500 550 600 650 700 750 800
-70
-60
-50
-40
-30
-20
-10
0
PPLN-1
P
1
-Φ
s
Φ
s
10 dB/D
P
2
Φ
s
-Φ
s
P
1
Output
QPM
QPM
1529 1538 1541 1544 1553 nm
PPLN-2
45
Figure 3.3 (a) The I and Q eye diagrams of the detected 20-Gbaud QPSK signals with three different
power levels of induced phase noise and for three different noise-bandwidths. (b) Phase noise reduction
factor for three levels of induced phase noise with different noise-bandwidths.
3.5 Demonstration of Tolerance to VCSEL-Wavelength-Drift and
DFB-High-Phase-Noise in an All-Optical Homodyne Receiver
It might be beneficial if the automatically–locked homodyne detection approach can
also provide added tolerance to other issues that are normally problematic for
homodyne detection of phase encoded data signals. For example, the wavelength drift
of an uncooled VCSEL or the relatively high phase noise of a DFB laser could pose
performance issues for these systems.
Hence, the tolerance to wavelength drift and high phase noise in an all-optical
homodyne receiver that uses nonlinear wave mixing are demonstrated. An uncooled
VCSEL with random wavelength drifts is used as a laser source and modulated with
20 and 32 Gbaud QPSK signals. We demonstrate that the conjugate copy of the signal
which has the same amount of drifts but in the opposite directions of the signal’s drifts
can compensate the wavelength drifts in the automatically locked homodyne detection
method. In addition, a DFB laser with high phase noise is used as another laser source
for 20 and 32 Gbaud QPSK signals. We demonstrate that the produced delayed signal
conjugate and nonlinear wave-mixing mitigate the phase noise of the laser in the
homodyne detection scheme. By multiplying the signal to its delayed conjugate copy
the phase noise coming from the laser can be filtered. Open eyes are obtained for both
in-phase and quadrature components of the signal and both mentioned laser sources.
Noise
B.W.
300 MHz
750 MHz
Noise
Level-2:
Noise
Level-3:
3300 MHz
Noise
Level-1:
EVM:
33.8%
EVM:
25.9%
EVM:
17.3%
No
Noise:
EVM:
10.5%
Input
Noise Bandwidth (MHz)
250 100 500 1000 2500 5000
3
2.5
2
1.5
1
Noise Reduction Factor
a) b)
100 250 500 750 1000 2500 5000
1
1.5
2
2.5
3
Noise level-1
Noise level-2
Noise level-3
LO+R LO+jR LO+R LO+jR
LO+R LO+jR
LO+R LO+jR LO+R LO+jR LO+R LO+jR
LO+R LO+jR LO+R LO+jR
LO+R LO+jR
LO+R LO+jR
LO+R LO+jR
LO+R LO+jR
46
The conceptual block diagram of the all-optical homodyne system is shown in
Fig. 3.4. A QPSK signal which is modulated on an unstable laser source at frequency
𝜔 𝑠 with large linewidth and random frequency drift 𝛿𝜔 (𝑡 ) is received. The signal,
along with a CW pump 𝑃 1
at 𝜔 𝑃 1
are injected into a nonlinear wave mixer (Nonlinear
element-1) to generate the conjugate copy of the signal at 𝜔 𝑐 = [(2𝜔 𝑃 1
− 𝜔 𝑠 ) −
𝛿𝜔 (𝑡 )]. The signal, the conjugate copy, and the pump are sent to the programmable
optical filter in which appropriate amplitudes and phases (complex weights) are
applied. In addition, one symbol delay is induced between the signal and the copy in
the filter. The signal, the conjugate copy and the pump-1, are coupled with a second
pump laser and injected into a 𝜒 (2)
nonlinear element in which three tasks will be done
simultaneously: 1) Compensation of random frequency drifts from the signal laser
source: In the 𝜒 (2)
element the signal and its conjugate copy interact through a
nonlinear sum frequency generation (SFG) process and generate a new signal at
[𝜔 𝑠 + 𝛿𝜔 (𝑡 )] + [(2𝜔 𝑃 1
− 𝜔 𝑠 )− 𝛿𝜔 (𝑡 )]. Since the amounts of frequency drifts in the
signal and the conjugate copy are in opposite directions, in the SFG process the
frequency drifts are canceled. Finally, in a difference frequency generation (DFG)
process with the second pump laser, a new signal is generated at 2𝜔 𝑃 1
− 𝜔 𝑃 2
. 2) Phase
noise mitigation: In the cascaded SFG-DFG processes the signal and its delayed
conjugate copy are mixed. We denote the generated signal by 𝑅 (𝑡 ) ∝ 𝑒𝑥𝑝 (𝑗 𝛷 𝑠 (𝑡 )−
𝛷 𝑠 (𝑡 − 𝑇 )). Since the phase of the incoming signal consists of data and phase noise,
then the phase of 𝑅 (𝑡 ) becomes 𝛷 𝑅 (𝑡 ) = [𝛷 𝐷 (𝑡 )− 𝛷 𝐷 (𝑡 − 𝑇 )] + [𝜑 𝑁 (𝑡 ) − 𝜑 𝑁 (𝑡 −
𝑇 )]. In which, the first bracket is only a decoding of the original data and the second
bracket can be viewed as a filter on the phase noise, which is able to filter out the laser
linewidth phase noise. The second simultaneous nonlinear process leads to 3) Phase-
locked multiplexing by the “local” laser 𝑃 1
to generate an automatically locked LO in
phase and frequency for homodyne detection: This can be done by multiplexing the
phase noise mitigated signal 𝑅 (𝑡 ) with the CW pump, 𝑃 1
, with an appropriate relative
complex coefficient being adjusted in the programmable filter. The final optical output
47
can be denoted as 𝑆 𝑜𝑢𝑡 (𝑡 ) ∝ 𝐿𝑂 + 𝑤𝑅 (𝑡 ), where 𝐿𝑂 ∝ 𝑃 1
2
. By sending the output to
photo-detector and setting 𝑤 to ±1 or ±j, both in-phase and quadrature components of
the noise mitigated input signal can be obtained.
Figure 3.4 The Conceptual block diagram of the all-optical homodyne system.
The experimental setup for the all-optical homodyne system is depicted in Fig.
3.5 (a). A DFB laser source at ~1559 nm and an uncooled VCSEL at ~ 1559 nm are
used as laser sources for modulating the QPSK signal. A nested Mach–Zehnder
modulator is used to generate 20 and 32-Gbaud QPSK data (PRBS 2
15
-1). The
modulated signal is amplified and coupled with an amplified CW pump around 1551.5
nm and sent to a periodically-poled-lithium-niobate (PPLN-1) waveguide. The PPLN-
1 output are sent to a programmable filter based on liquid crystal on silicon (LCoS)
technology for adjusting the amplitudes, phases, and delays. The output signals are
coupled with another CW pump at ~ 1560 nm and sent to PPLN-2 waveguide to 1)
coherently mix the signal and its conjugate copy for wavelength drift compensation
and phase noise mitigation and, 2) to phase-locked multiplex the first pump in the
output as an LO for homodyne detection. The multiplexed signal is then filtered,
amplified and sent to a photodiode. In the Fig. 3.5 (b) the output spectra of the PPLNs
are shown for 20-Gbaud QPSK data. As can be seen, the multiplexed signal is always
frequency/phase locked with respect to the generated carrier.
ω
P1
P
1
Φ
s
-Φ
s
Phase Conjugate
Generation
P
1
ω
s
Adjusting
Delays,
Coefficients
Nonlinear Element-2
P
2
1- Wavelength Drift Compensation
2-Phase Noise Mitigation
3-Phase/Frequency-Locking with LO
Programmable
Optical Filter
Nonlinear Element-1
P
1
P
2
DFG Mixing
ω
out
ω
P2
SFG Mixing
ω
P1
SHG+DFG
Mixing
Incoming Signal
w/ Frequency Drift
and Phase Noise
Homodyne Detection
(w=1,j)
δω(t)
ω
s
δω(t)
ω
c
δω(t)
Φ
s
(t) -Φ
s
(t-T)
ω
s
δω(t)
ω
c
δω(t)
48
Figure 3.5 (a) Experimental setup. DCF: Dispersion compensating fiber, SLM: Spatial light modulator
(b) Optical spectra of PPLN-1, PPLN-2.
An uncooled VCSEL at ~ 1559 nm with bias current at ~15 mA and <500 µA
rms ripple is used as a laser source to modulate the signal. This source is used to show
the tolerant of the homodyne detection scheme to the random wavelength drifts of the
laser source. In addition, we changed this source with a DFB laser with high phase
noise to show the inherent phase noise mitigation property of the automatically locked
homodyne detection scheme. Figure 3.6 shows the spectrum of the VCSEL source at
different times with a few seconds difference. In order to investigate the constellation
diagram using a coherent receiver, we filter out the pump-1 in the LCoS filter (Fig.
3.5(b)-the second spectrum) and remove the LO from final optical output in PPLN-2.
Figure 3.6 Wavelength drifts of the VCSEL source. The random drifts of <0.05 nm are occurred at
different times with a few seconds difference.
90
1 nm
λ
P1
1 nm
9 nm
1 nm
LCoS
Filter
PPLN-2
Waveguide
20 Gbaud PRBS (2
15
-1)
Phase Conjugate
Generation
a)
λ
P2
Frequency Drift
Cancellation+
Noise Mitigation+
Phase Locked Muxing
λ
SIG
b)
PD
10 dB/D
PPLN-1
Waveguide
10 dB/D
1542 1550 1560 nm
1 nm
VCSEL/
DFB
1542 1544 1546 1548 1550 1552 1554 1556 1558
-80
-70
-60
-50
-40
-30
-20
PPLN-1
1542 1544 1546 1548 1550 1552 1554 1556 1558 1560
-80
-70
-60
-50
-40
-30
-20
-10
Φ
s
-Φ
s
-Φ
s
Φ
s
PPLN-2
out
P2
P1
1542 1544 1546 1548 1550 1552 1554 1556 1558 1560
-80
-60
-40
-20
0
PPLN-2
P2
Φ
s -Φ
s
P1
P1
Blocked
10 dB/D
1542 1550 1560 nm
QPM
1543 1550 1558 nm
Adjusting Delays, and
Complex Weights
LO
1558.6 1558.7 1558.8 1558.9 1559
-70
-60
-50
-40
10 dB/div
49
Figure 3.7, first row, shows the constellation diagram of the 20-Gbaud QPSK signal
modulated on the uncooled VCSEL source. The constellation diagrams are shown for
three different shots obtained from the coherent receiver with a few-seconds time
differences. As can be seen, due to the random wavelength drifts of the VCSEL, the
detected QPSK signal is not stable. The second row of Fig. 15 shows the constellation
diagrams of generated output in PPLN-2. The stability of the constellation is from the
wavelength drift compensation in PPLN-2.
Figure 3.7 The constellation diagrams of 20-Gbaud QPSK signals from the back to back VCSEL source
without wavelength drifts compensation on the top, and the all-optical wavelength drifts compensated
output from PPLN-2 on the bottom.
Figure 3.8 shows the constellation diagrams when 20 and 32 Gbaud QPSK is
modulated on a DFB laser source. As can be seen, due to a phase noise mitigation in
PPLN-2, the amount of EVM is reduced by ~25% and ~40% for 20 and 32 Gbaud
QPSK signals respectively.
Evm=30.9% Evm=45.6% Evm=15.2%
Evm=11.5%
Evm=11.8%
Evm=12%
Without
Wavelength Drifts
compensation
With
All-optical
Wavelength Drift
Compensation
@ @ @
50
Figure 3.8 constellation diagrams of 20 and 32-Gbaud QPSK signals from the back to back DFB source
without phase noise mitigation on the left , and the all-optical phase noise mitigated output from PPLN-
2 on the right.
When pump-1 is not filtered by LCoS filter before PPLN-2, it generates an LO through
a nonlinear SHG-DFG process, which is automatically locked with the output signal
of PPLN-2. By utilizing a photo-diode, a homodyne detection is possible in order to
capture the eyes. Figures 3.9 shows the resultant eye-diagrams for 20 and 32 Gbaud
signals modulated on the DFB laser source and the uncooled VCSEL. Both in-phase
and quadrature-phase components of the data signal are detected. Open eyes are
achieved due to the wavelength drift and phase noise mitigation in PPLN-2.
Figure 3.9 The I and Q eye diagrams for 20 and 32 Gbaud QPSK signals modulated on DFB laser source
and an uncooled VCSEL.
Evm=19% Evm=14.2
%
Without
Phase Noise Mitigation
With
All-optical phase Noise Mitigation
Evm=27.5%
Evm=16%
20 Gbaud 32Gbaud
LO+R LO+jR
LO+R LO+jR
LO+R LO+jR
LO+R LO+jR
DFB
Homodyne Detection
Uncooled VCSEL
20 Gbaud 32 Gbaud
51
Chapter 4 Tunable Nonlinear Phase-Noise
Mitigation and Automatic Frequency/Phase
Locking for a Homodyne Receiver using Optical
Mixing of Nonlinearly Generated Higher
Harmonics
4.1 Introduction
Optical homodyne systems are known to provide superior sensitivity and performance as
compared to heterodyne system [48–50]. Similar to coherent systems in general, homodyne
systems also enable the recovery and digital signal processing of phase-encoded data signals
(e.g., phase-shift-keying, PSK). However, homodyne systems require the local oscillator (LO)
in the receiver to have the same frequency and phase as the incoming data signal, i.e., the data
signal and the LO must be “locked” to each other. In addition, nonlinear phase noise
originating from interaction of ASE noise and Kerr nonlinearity can pose a key limitation in
such a phase sensitive systems and there is a need for phase noise mitigation in an efficient
homodyne system [37–41]. In terms of phase-noise mitigation in coherent systems, several
methods have included different variations of a phase-sensitive-amplifier-based approach to
achieve a level of “phase squeezing” [15,41–45]. However, these methods tend to require
coherency between a data signal and a pump, thereby typically necessitating phase-based
feedback loops and injection locking [15,41–45]. A laudable goal might be to attempt to meet
simultaneously both these challenges of automatic frequency/phase locking and nonlinear
phase-noise mitigation in a homodyne system.
Here, we experimentally demonstrate tunable phase-noise mitigation and automatic
frequency/phase locking for a 20-32 Gbaud QPSK homodyne receiver using optical mixing
of nonlinearly generated higher harmonics.
52
In our scheme, a “local” pump laser is used to generate the signal conjugate and the 3rd order
signal harmonics. By utilizing a second pump laser, the signal and the generated harmonics
are multiplexed and the signal phase noise is mitigated. Simultaneously, in another nonlinear
process, in the same nonlinear element, the “local” first pump laser is automatically locked
and multiplexed to the noise mitigated signal. Open eyes are obtained for the both in-phase
and quadrature components of the signal after noise mitigation and ~2dB OSNR gain is
achieved at BER 10
-3
.
4.2 Concept
The conceptual block diagram of the homodyne phase noise mitigation scheme is
shown in Fig. 4.1. An incoming QPSK signal contaminated with phase noise is
coupled with a CW pump laser 𝑃 1
and injected into a nonlinear wave mixer to generate
a conjugate copy of the signal. In more details, the signal 𝑆 (𝑡 ) and the 𝑃 1
interact
through the four-wave mixing nonlinear process through a highly nonlinear fiber
(HNLF) as a nonlinear medium and generate a conjugate copy of the signal with
electric field 𝑃 1
2
𝑆 ∗
(𝑡 ).
The signal, the 𝑃 1
, and the conjugate copy are sent into a nonlinear medium with
the third-order nonlinear susceptibility,𝜒 (3)
, to generate the third-order harmonics of
the signal and the conjugate copy. The signal and its conjugate are mixed through a
FWM in a second HNLF and generate the third-order harmonics with the electric fields
proportional to 𝑃 1
∗
2
𝑆 3
(𝑡 ) and 𝑃 1
4
𝑆 ∗
3
(𝑡 ). In a liquid crystal on silicon (LCoS)
programmable filter, one symbol delay (𝑇 ) is applied between: (i) the signal and its
conjugate copy, and (ii) third-order harmonic signals generated in the second HNLF.
The adjusted signals along with a new CW pump 𝑃 are injected into a last nonlinear
stage with the second-order nonlinear susceptibility, 𝜒 (2)
in which two tasks will be
done simultaneously. Task (1), phase quantization: a staircase phase transfer function
is realized based on the signal, its conjugate, and the generated third-order harmonics
to quantize the phase of the input signal. The signal and its one symbol delayed
53
conjugate copy are mixed through a cascaded process of sum-frequency-generation
(SFG) and difference-frequency-generation (DFG) nonlinear processes and generate a
new signal 𝐸 1
(𝑡 ). Similarly, in parallel processes, the conjugate third-order harmonic
signal and one symbol delayed third-order harmonic signal are mixed through a
cascaded SFG-DFG processes and generate a signal 𝐸 2
(𝑡 ). The coherent summation
of the two fields 𝐸 1
(𝑡 ) and 𝐸 2
(𝑡 ) results in phase quantization of the input signal. The
phase quantized signal 𝑆 𝑄 (𝑡 )can be written as
𝑆 𝑄 (𝑡 ) ∝ 𝑃 2
∗
𝑆 (𝑡 )𝑃 1
2
𝑆 ∗
(𝑡 − 𝑇 )+ 𝛼 𝑃 2
∗
𝑃 1
∗
2
𝑆 3
(𝑡 − 𝑇 )𝑃 1
4
𝑆 ∗
3
(𝑡 )
= 𝑃 2
∗
𝑃 1
2
(𝑆 (𝑡 )𝑆 ∗
(𝑡 − 𝑇 )+ 𝛼 [𝑆 (𝑡 )𝑆 ∗
(𝑡 − 𝑇 )]
∗
3
)
= 𝑒 𝑗 (Φ
𝑠 (𝑡 )−Φ
𝑠 (𝑡 −𝑇 ))
+ 𝛼 𝑒 −𝑗 3(Φ
𝑠 (𝑡 )−Φ
𝑠 (𝑡 −𝑇 ))
= 𝑒 𝑗 ∆Φ
𝑠 (𝑡 )
+ 𝛼 𝑒 −𝑗 3∆Φ
𝑠 (𝑡 )
(4.1)
where the appropriate value of 𝛼 can be adjusted in the LCoS filter by changing the
relative phase and amplitude values of the signal and the harmonics. Moreover, Δ is
defined as a difference operator over one symbol interval 𝑇 , i.e., ∆Φ
𝑠 (𝑡 ) = Φ
𝑠 (𝑡 )−
Φ
𝑠 (𝑡 − 𝑇 ). The last term in Eq. 4.1 represents the phase quantization function which
enables squeezing the phase noise of the input signal.
The second simultaneous task inside PPLN is automatically phase and frequency locking of
the local pump laser 𝑃 1
with the phase noise mitigated signal. In details, the pump 𝑃 1
generates a signal 𝐿𝑂 ∝ 𝑃 2
∗
𝑃 1
2
through the SFG-DFG processes. Therefore 𝑃 1
has the same
phase reference as the noise mitigated signal written in the second line of Eq. 4.1. This
indicates that although the input signal 𝑆 (𝑡 ) is generated from a laser which is not phase-
locked to the pump laser 𝑃 2
, the noise mitigated signal and the generated 𝐿𝑂 are automatically
phase and frequency locked at the frequency output of 2𝜔 𝑃 1
− 𝜔 𝑃 2
. The noise mitigated
signal and the generated 𝐿𝑂 are coherently mixed with an appropriate relative complex
coefficient adjusted in the programmable filter and the final optical output can be denoted as
𝑆 𝑜𝑢𝑡 (𝑡 ) ∝ 𝐿𝑂 + 𝑤 𝑆 𝑄 (𝑡 ). By sending the output to photo-detectors and setting 𝑤 to ±1 or
±j, similar to a 90
0
optical hybrid, both in-phase and quadrature components of the noise
mitigated input signal can be obtained.
54
Figure 4.1 The concept of the homodyne nonlinear phase noise mitigation scheme.
4.3 Experimental Setup
The experimental setup for the homodyne phase noise mitigation scheme is depicted
in Fig. 4.2 (a). A nested Mach–Zehnder modulator is used to generate 20/32-Gbaud
QPSK data (PRBS 2
31
-1) at ~1550nm. The signal is amplified and phase modulated
with an ASE source to emulate phase noise. The noisy signal is amplified and coupled
with an amplified CW pump around 1551.5 nm and sent to a ~500 m HNLF-1 with
~1551 nm zero dispersion wavelength (ZDW). The HNLF-1 output are sent to a SLM
filter for adjusting the signals amplitudes. The output is sent to a ~300 m HNLF-2 with
ZDW ~1551 nm to generate the third-order harmonics. The HNLF-2 output is sent to
a SLM filter to apply appropriate delays, phases, and amplitudes on the signals. The
resulting signals are amplified and coupled with another CW pump around 1558 nm
and sent to the PPLN waveguide to 1) coherently mix the product of the signal and its
conjugate copy and the product of the third harmonic and its conjugate through the
cascaded processes of sum frequency generation (SFG) followed by difference
frequency generation (DFG), and, 2) to multiplex the first pump in the output through
the cascaded processes of second harmonic generation followed by DFG. The
multiplexed signal is then filtered, amplified and sent to a photodiode to capture the
Nonlinear Element
(PPLN/HNLF)
Nonlinear Element
(HNLF)
Amplitude/
Phase
Filter
Nonlinear Element
(PPLN)
Nonlinear Element
(PPLN)
Phase Conjugate
Generation
Adjusting
Delays,
Coefficients
-order
Harmonics
Generation
Nonlinear Phase Noise
Mitigation+
Phase-Locked
Multiplexing
Noisy QPSK Signal
ω
P1
-Φ
s
Φ
s
P
1
ω
P1
3Φ
s
-3Φ
s
-Φ
s
(t-T) Φ
s
(t)
P
1
ω
P1
3Φ
s
(t-T) -3Φ
s
(t)
P
2
SFG, SHG Mixing
DFG Mixing
SHG+ DFG Mixing
Four Wave Mixing
ω
out
ω
P2
P
1
-Φ
s
Φ
s
Higher Harmonics Generation
Q
Q
55
eye diagrams and perform the BER measurement. In the Fig. 4.2 (b,c) the output
spectra of the HNLFs, and the PPLN are shown for 20/32-Gbaud QPSK data. As can
be seen, the multiplexed signal is always in frequency/phase locked with respect to the
generated carrier.
Figure 4.2 (a) Experimental setup. PM: Phase modulator, HNLF: Highly nonlinear fiber, PPLN:
Periodically poled lithium niobate, SLM: Spatial light modulator, PD: Photo detector, VOA: Variable
optical attenuator. (b) 20 Gbaud optical spectra (c) 32 Gbaud optical spectra
4.4 Experimental Results
The performance of the system is assessed by implementing the proposed homodyne
phase noise mitigation scheme on 20-32 Gbaud QPSK signals. In order to show the
performance of the system, the incoming data signal are contaminated with two
different levels of phase noise. Both in-phase and quadrature-phase components of the
data signal are detected. The eyes are detected based on 1) the homodyne detection
scheme without phase noise mitigation and 2) the proposed homodyne phase noise
mitigation method. Figure 4.3 show the resultant eye-diagrams for 20, and 32 Gbaud
signals, respectively. Open eyes can be achieved in both methods when the incoming
data signal is not noisy. As can be seen in case of noisy input signal for the two
different noise levels, the homodyne noise mitigation scheme reduces the amount of
noise in detected in-phase (I) and quadrature (Q) eyes compared to the homodyne
90
2 nm 1 nm
VOA
Broadband
ASE Source
λ
p1
1 nm
RF Amp. PD
20 dBm
23 dBm
PM
4 nm
SLM
Filter
9 nm
20 dBm
1 nm
SLM
Filter
PPLN
Waveguide
10 dB/D
20/32 Gbaud PRBS (2
31
-1)
20 dBm
HNLF-2
Higher Harmonics Generation
a)
λ
p2
20 dBm
Noise Mitigation+
Phase Locked Muxing
λ
SIG
b)
600 700 800 900 1000
-70
-60
-50
-40
-30
-20
-10
0
P
1
1550 1551.5 1553 nm
-Φ
s
Φ
s
600 700 800 900 1000
-70
-60
-50
-40
-30
-20
-10
0
Φ
s
-Φ
s
P
1
200 400 600 800 1000 1200
-60
-50
-40
-30
-20
-10
200 400 600 800 1000 1200
-50
-40
-30
-20
-10
0
400 600 800 1000 1200 1400 1600
-70
-60
-50
-40
-30
-20
-10
3Φ
s
Φ
s
-Φ
s
-3Φ
s
P
1
P
2
1550 1551.5 1553 nm
1547 1551.5 1556 nm 1547 1551.5 1556 nm
1546 1551.5 1556 nm 1546 1551.5 1556 nm
1544 1546 1548 1550 1552 1554 1556
-70
-60
-50
-40
-30
-20
-10
QPM QPM
Out 3Φ
s
Φ
s
-Φ
s
-3Φ
s
P
1
P
2
Out
3Φ
s
Φ
s
-Φ
s
-3Φ
s
3Φ
s
Φ
s -Φ
s
-3Φ
s
HNLF-1
HNLF-2
HNLF-1
HNLF-2
PPLN PPLN
c)
HNLF-1
PD
10 dB/D 10 dB/D
10 dB/D
10 dB/D
10 dB/D
20 Gbaud 32 Gbaud
56
receiver without phase noise cancellation. Since the local oscillator is automatically
“locked” in frequency and phase to the incoming data signal, the proposed homodyne
receiver does not require phase and frequency recovery.
Figure 4.3 The I and Q eye diagrams of 20 and 32 Gbaud QPSK signals without noise, with noise level-
1, and with noise level-2. The results are shown for the homodyne detection scheme without phase noise
mitigation, and, the proposed homodyne phase noise mitigation method.
Figure 4.4 shows the BER performance of the homodyne phase noise
mitigation for two different noise levels. This scheme results in ~2 dB OSNR gain at
a BER of 10-3. All results are captured without phase and frequency tracking as
opposed to conventional intradyne detection.
a) 20 Gbaud
Homodyne
Detection
Homodyne
Phase noise-
Mitigation
Noise Level-1:
Noise Level-2:
Without Phase Noise:
Homodyne
Detection
Homodyne
Phase noise-
Mitigation
Noise Level-1:
Noise Level-2:
Without Phase Noise:
b) 32 Gbaud
LO+R
LO+jR LO+R
LO+jR
LO+R LO+jR
LO+R
LO+jR LO+R
LO+jR
LO+R LO+jR
LO+R
LO+jR LO+R
LO+jR
LO+R LO+jR
LO+R
LO+jR LO+R
LO+jR
LO+R LO+jR
57
Figure 4.4 BER versus for homodyne detection scheme with/without phase noise mitigation and for two
different noise levels. (a) 20Gbaud (b) 32 Gbaud.
2
3
4
5
2
3
4
5
a) b)
1
10 12 14 16 18 20
-log
10
(BER)
OSNR (dB)
Without noise mitigation_Noise Level 2
with noise mitigation_Noise Level 2
without noise mitigation_Noise Level 1
with noise mitigation_Noise Level 1
20 Gbaud
1
10 12 14 16 18 20 22
-log
10
(BER)
OSNR (dB)
without noise mitigation_Noise Level 2
with noise mitigation_Noise Level 2
without noise mitigation_Noise Level 1
with noise mitigation_Noise Level 1
32 Gbaud
58
Chapter 5 Multiplexing and Transmission of
QPSK-to-16QAM Channels using Wave Mixing for
Aggregation and Noise Mitigation
5.1 Introduction
Transmission systems have traditionally been composed of many different wavelength
channels that originate at many transmitters and terminate without much change at the
receivers [57]. However, networks are envisioned to become more dynamic and
heterogeneous in the future [57,58]. This becomes ever more important as: (i) optical
spectrum becomes scarce at certain points in the network, (ii) certain parts of the
network have higher capacity or more available spectrum than other parts, and (iii)
higher channel spectral efficiency is globally desired [57,59,60].
One possibility of alleviating the scarce spectrum at certain points in the network
is to aggregate multiple data channels onto a single higher-spectral-efficiency channel
for subsequent transmission. For example, two incoming quadrature-phase-shift-
keyed (QPSK) data channels can be optically multiplexed into a single 16-quadrature-
amplitude-modulated (QAM) channel that has twice the spectral efficiency [9,61].
Importantly, these two data channels might contain phase noise that one would
want to mitigate. In the past, there have been reports of techniques to optically
aggregate lower-order modulation format channels into a single higher one [59,61]
and other reports to optically mitigate phase noise [62]. However, there has been
limited reported of any approach that aggregates and mitigates phase noise at the same
time, as well as any of the above approaches in a transmission system.
Here, we demonstrate the optical multiplexing and transmission of QPSK-to-
16QAM channels over 100 km using wave mixing for aggregation and noise
mitigation. In our scheme, the phase noise of the incoming QPSK signals are filtered
in an optical nonlinear element using optical wave mixing and tunable optical delays.
59
Simultaneously, in the same nonlinear device, the noise mitigated QPSK signals are
coherently multiplexed and generate a 16QAM signal. The system performance is
verified by constellation analysis, the error-vector-magnitude (EVM), and the bit error
rate (BER) measurements before and after transmission of 100 Km. For aggregated 10
and 20 Gbaud 16QAM signals, the EVM reduction of more than 30% is achieved for
high phase noise of ~50 degree and 300 MHz phase noise bandwidth. For the input
signals of with phase noise of ~50 and ~35 degree and 300 MHz bandwidth the OSNR
penalty of ~1dB is achieved at BER 10
-3
for 20 Gbaud 16QAM signal compared to the
case of aggregation from signals with negligible or no amount of phase noise phase
noise.
5.2 Concept
The conceptual block diagram of the simultaneous optical aggregation and phase noise
mitigation system is shown in Fig. 5.1. The two incoming QPSK signals contaminated
with phase noise, e.g. phase noise from lasers with large linewidths, are received. The
noisy signals, along with a CW pump P
1
, are injected into a nonlinear wave mixer
(Nonlinear element-1) to generate the conjugate copies of the signals. The signals and
the conjugate copies are then sent into a delay stage (dispersive medium) to apply one
symbol delay between them. The signals, the copies and the pump are sent to the
programmable optical filter in which the pump is filtered out and appropriate
amplitudes and phases (complex weights) are applied on the signals and the conjugate
copies. The signals and the conjugates with a second pump are injected into a second
nonlinear element in which two tasks will be done simultaneously: 1) Phase noise
mitigation by coherently mixing each signal and its corresponding delayed conjugate.
We denote the generated noise mitigated signals by X
i
(t) ∝ exp (jΦ
s
i
(t)− Φ
s
i
(t−
T)). Since the phase of each incoming signal consists of data and phase noise, then the
phase of X
i
(t) becomes Φ
X
i
(t) = [Φ
D
(t)− Φ
D
(t− T)] + [φ
N
(t) − φ
N
(t− T)]. In
which, the first bracket is only a decoding of the original data and the second bracket
can be viewed as a filter on the phase noise, which is able to filter out the low frequency
60
components of the phase noise (e.g., laser phase noise). The second simultaneous
nonlinear process leads to 2) All-optical aggregation: this will be done by phase-locked
multiplexing of the noise mitigated signals X
1
(t) and X
2
(t) with appropriate relative
complex coefficients adjusted in the programmable optical filter. The final optical
output can be denoted as S
out
(t) = X
1
(t) + wX
2
(t) where w is the relative complex
coefficient. The output aggregated signal is a 16-QAM signal with the same baud rate
which can be amplified and sent through the optical link.
Figure 5.1 The concept of the simultaneous optical aggregation and phase noise mitigation system. The
incoming two QPSK signals contaminated with phase noise, e.g. phase noise of lasers with large
linewidths, are received. The conjugate copies of signals with one symbol delay are generated using
PPLN-1 waveguide and a dispersive medium. In PPLN-2, each signal multiplied with its conjugate
copy to filter out the phase noise, and in a simultaneous process, they coherently multiplexed to generate
the output 16 QAM signal.
5.5 Experimental Setup
The experimental setup of simultaneous all-optical aggregation with phase noise
mitigation is depicted in Fig. 5.2 (a). A nested Mach–Zehnder modulator is used to
generate 10-20 Gbaud QPSK data (PRBS 2
15
-1) at ~1537.5nm and 1539nm. The
signals are amplified and phase modulated with an ASE source followed by a variable
optical attenuator, PD, and RF filter to emulate phase noise with different powers and
bandwidths. The noisy signals are sent to transmission links of length of 100 km,
Tx Link
Tx Link
Tx Link
Programmable
LCoS
Filter
Phase Conjugate
Generation
Adjusting
Complex
Coefficients
One Symbol
Delay
Aggregation
+
Phase Noise
Mitigstion
-Φ
s
Φ
s
ω
P1
Φ
s
-Φ
s
SHG+ DFG Mixing
-Φ
s
(t-T
s
) Φ
s
(t)
ω
P1
Φ
s
(t-T
s
) -Φ
s
(t)
P
2
SFG Mixing
DFG Mixing
ω
out
ω
P2
Conjugates Generation with Appropriate
Delays, Amplitudes and Phases
PPLN-1: PPLN-2:
Output
QPM
P
1
Nonlinear Element
(PPLN-1)
Incoming Signals
Dispersive Element
(DCF)
P
2
Nonlinear Element
(PPLN-1)
Aggregated + Phase Noise
Mitigated Signal
λ SIG1
λ SIG2
61
which consists of two erbium-doped fiber amplifiers, 80 km SMF-28, and 20 km DCF.
The outputs are amplified and coupled with an amplified CW pump around 1541.6 nm
and injected to periodically-poled-lithium-niobate (PPLN-1) waveguide. The signals
and the copies are then sent to a ~150 m dispersion compensated fiber (DCF) to apply
one symbol delay between them. The signals and copies are sent to a spatial light
modulator (SLM) phase and amplitude programmable filter based on liquid crystal on
silicon (LCoS) technology for adjusting the amplitudes and phases. The resulting
signals are amplified and coupled with another CW pump around 1532.5 nm and sent
to PPLN-2 waveguide to 1) coherently mix each signal with its conjugate copy (optical
phase noise filter) through the nonlinear process of sum frequency generation (SFG),
and, 2) to phase-locked multiplex the noise mitigated signals (aggregation) through
the cascaded processes of SFG followed by difference frequency generation (DFG).
The aggregated signal is then filtered and sent to a transmission link of length of 100
km and finally is detected by a coherent receiver to capture the constellation diagrams
and perform the BER measurement.
Figure 5.2 (a) Experimental setup. PM: Phase modulator, PPLN: Periodically poled lithium niobate,
DCF: Dispersion compensated fiber, SLM: Spatial light modulator, PD: Photo detector, VOA: Variable
optical attenuator. (b) 20 Gbaud optical spectra after nonlinear stages (PPLN-1, 2).
90
5 nm
VOA
Broadband
ASE Source
λ
p1
1 nm
23 dBm
PM
9 nm 1 nm
LCoS
Filter
PPLN
Waveguide
10/20 Gbaud PRBS (2
15
-1)
Phase Conjugate
Generation
a)
λ
p2
Noise Mitigation+
Phase Locked Muxing
PPLN-1
Waveguide
DCF
LPF
λ
SIG1
λ
SIG2
ΔƬ
~ 80 km
2 nm 2 nm
SMF-28 DCF
~ 20 km
Tx Link
Coherent
Receiver
Phase Noise
Emulator
Phase Noise
Emulator
Tx Link
Tx Link
400 450 500 550 600 650 700 750 800 850
-60
-50
-40
-30
-20
-10
200 400 600 800 1000 1200
-70
-60
-50
-40
-30
-20
-10
1541 1538 1544
nm 1537 1546 1532 1550
nm
PPLN-1 P
1
-Φ
S1
Φ
S1
QPM
QPM
PPLN-2
Φ
S2
-Φ
S2
P
2
-Φ
S1
Φ
S1
Φ
S2
-Φ
S2
Out
b)
10 dB/div 10 dB/div
62
5.3 Experimental Results
The performance of the system is assessed by implementing the proposed simultaneous phase
noise mitigation and optical aggregation scheme on 10-20 Gbaud QPSK-to-16QAM channels.
Figure 5.3(a) shows the system input/output constellation diagrams when only one of the
incoming QPSK signal with phase noise of more than 50 degree and the bandwidth of 300
MHz is sent to the system and the other QPSK channel is blocked. As it can be seen, the 300
MHz phase noise is successfully filtered out in the output of the system. The EVM is reduced
by more than 50% for both 10 and 20 Gbaud QPSK signals in the output of the system. Figure
5.3(b) shows the constellation diagrams of the aggregated 16QAM signal of incoming noisy
QPSK signals. The phase noise levels of the two channels are around 50 and 35 degree and
the phase noise bandwidth of the channel with higher phase noise varies from 300 MHz to
5000 MHz and the phase noise bandwidth of the other channel is 300 MHz. It can be seen that
the phase noise reduction for the cases with lower bandwidths phase noise (Δν~300 MHz,
1000 MHz) compared to the cases with higher bandwidths (3300 MHz, 5000 MHz) is
significant. Figure 5.4(a) shows the EVM reduction of the 20 Gbaud aggregated 16QAM
signal for different bandwidths of phase noise compared to the case of aggregation with phase
noise of 5000 MHz bandwidth. The reduction values are obtained, when one of the QPSK
channel has phase noise with different levels of ~50, ~45, and ~35 degree (noise levels 1 to 3)
and different bandwidths of 300 MHz to 5000 MHz, and the other channel has no or negligible
amount of phase noise.
63
Figure 5.3 The constellations diagrams for 10-20 Gbaud signals (a) The input and output constellation
diagrams when only a noisy QPSK channel is sent to the system. (b) The aggregated 16 QAM signals
from QPSK channels with phase noise levels of around 50 and 35 degree. The phase noise bandwidth
of the channel with higher phase noise varies from 300 to 5000 MHz and the phase noise bandwidth of
the other channel is 300 MHz.
Figure 5.4(b) shows the BER performance of the aggregated 16QAM signal before and
after 100 Km transmission of 80 km SMF-28 and 20 km DCF. The BER curves are obtained
for QPSK channels with phase noise levels of around 50 and 35 degree. The phase noise
bandwidth of the channel with higher phase noise is changed to 300, and 1000 MHz and the
phase noise bandwidth of the other channel is 300 MHz. For the input signals with phase noise
of ~50 and ~35 degree and 300 MHz noise bandwidth the OSNR penalty of ~1dB is achieved
at BER 10
-3
for 20 Gbaud 16QAM signal compared to the case of aggregation from signals
with negligible or no amount of phase noise.
300 MHz 1000 MHz 3300 MHz 5000 MHz
System In/Out
Only One Channel ON
(Δν=300 MHz)
EVM=10.1 % EVM=10.6 % EVM=12.4 % EVM=14.9 % EVM=27.5 %
Phase Noise
B.W. (Δν):
EVM=9.9 % EVM=10.4 % EVM=12.1 % EVM=14.5 % EVM=26.1 %
10 Gbaud
Input Output
EVM=12.7 %
EVM=13.3%
a) b)
Input Output
20 Gbaud
Aggregated Output Signal of Noisy QPSK Signals
(Both Channels ON)
10 Gbaud 20 Gbaud
64
Figure 5.4 (a) The EVM reduction of 20 Gbaud aggregated 16QAM signal with different phase noise
bandwidths and for different levels of phase noise, ~50, 45, and 35 degree (Noise levels 1 to 3);
compared to the case of 5000 MHz noise bandwidth (b) The BER performance of the 16QAM
aggregated signal for input signals with phase noise of ~50 and ~35 degree and different phase noise
bandwidths, before and after 100 Km transmission.
0
5
10
15
20
25
30
35
40
300 3000
Noise Level-1
Noise Level-2
Noise Level-3
0
10
20
30
EVM Reduction %
300 600 1200 2400 5000
Phase Noise Bandwidth (MHz)
1
17 19 21 23 25 27 29 31
w/ No Phase Noise
w/ No Phase Noise_Transmitted
w/ 300MHz Phase Noise
w/ 300MHz Phase Noise_Transmitted
w/ 1000MHz Phase Noise
w/ 1000MHz Phase Noise_Transmitted
1
2
3
4
-log
10
(BER)
OSNR(dB)
a) b)
65
Chapter 6 Optical inter-channel interference
mitigation for spectrally overlapped 16QAM data
channels using nonlinear wave mixing
6.1 Introduction
A key figure-of-merit in optical communication systems is the spectral efficiency of
the available wavelength range [63,64]. One approach for increasing the spectral
efficiency is to reduce the guard band between adjacent data channels [64]. However,
a more aggressive approach to increasing spectral efficiency is to partially overlap
different optical channels. This typically produces increased ICI and requires benefits
from compensation techniques to recover data [65,66].
There have been different studies which attempt to reduce ICI in spectrally
overlapped channels using digital signal processing (DSP) [67–70]. These techniques
typically require receiving all the overlapping channels that create ICI for the desired
channel [49,67–74]. Multiple receivers are typically used to receive both the main
channel and the channels inducing ICI to use common digital multi-channel ICI
compensation algorithms. As a result, there might be an advantage to mitigate ICI
before detection and in the optical domain to enable use of only a single optical
receiver for the desired channel.
Recently, an optical approach for ICI mitigation of quadrature-phase-shift-keyed
(QPSK) data channels has been reported [75]. In that approach, the conjugate copies
of the QPSK data channels are generated by utilizing optical nonlinearities, and the
overlapping part of the spectrum is filtered and multiplied by a single complex
coefficient. This signal is conjugated again and mixed with the desired channel to
mitigate ICI. In that technique, ICI compensation is achieved only for QPSK signals
since a single tunable complex coefficient is used to compensate for ICI crosstalk.
66
Here, we demonstrate an optical ICI mitigation method for spectrally overlapped
optical 16QAM data channels without needing multi-channel detection and channel
spacing estimation. Our method enhances the previous approach of [75] in which only
a single copy of signal is generated, adjusted, and multiplexed with the signal. In this
method, two copies of the overlapping signals are generated in 𝜒 2
optical nonlinear
elements. The signals are multiplied by adjusted complex coefficient taps, and the
signal of interest is coherently multiplexed with the overlapping signal and its delayed
variant to mitigate ICI [76]. Here, we experimentally demonstrate optical ICI
mitigation for spectrally overlapped two-sub-carrier signals with the same
16QAM/16QAM and different modulation formats of 16QAM/QPSK. The
performance of the system is measured for 20 and 25 Gbaud signals and the results are
evaluated for different channel spacings and for different numbers of complex taps.
6.2 Concept
Figure 6.1 shows the conceptual block diagram of the proposed all optical ICI
compensation method. There are three main stages: (i) generation of optical copies,
(ii) adjustment of amplitudes, phases, and delays for selected signals, and (iii) coherent
multiplexing.
The incoming signals 𝑆 1
and 𝑆 2
with channel spacing of ∆𝑓 are partially
overlapped. Signal 𝑌 represents the combination of these two signals and is defined as
follows
𝑌 (𝑓 ) = 𝑆 1
(𝑓 + ∆𝑓 /2 )+ 𝑆 2
(𝑓 − ∆𝑓 /2) (6.1)
The signal 𝑌 along with three dummy pump lasers (𝑃 𝑑 𝑖 ,𝑖 = 1,2,3) and a CW pump (𝑃 1
) is
sent to a periodically poled lithium niobate (PPLN) waveguide to generate three signal copies
𝑌 1
, 𝑌 2
, and 𝑌 3
through the cascaded nonlinear processes of sum frequency generation (SFG)
and difference frequency generation (DFG). The signal 𝑌 𝑖 (i = 1,2,3) can be written as
67
𝑌 𝑖 (𝑓 ) ∝ 𝑆 1
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 𝑖 +
∆𝑓 2
) + 𝑆 2
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 𝑖 −
∆𝑓 2
) (6.2)
where 𝑓 𝑄𝑃𝑀 denotes the quasi phase matching (QPM) frequency of the PPLN waveguide and
𝑓 𝑑 𝑖 represents the frequency of i
th
pump laser 𝑃 𝑑 𝑖 . In the next stage, the signal copies with their
corresponding pumps are sent to an optical programmable filter to select signals and adjust
their amplitudes, phases, and delays. In more detail, at port 1 in Fig. 6.1, signal 𝑆 1
is filtered
from the first generated copy at 𝑓 𝑐 1
, including unavoidable interference from signal 𝑆 2
. In
addition, signal 𝑆 2
and its delayed copy are filtered respectively at 𝑓 𝑐 2
and 𝑓 𝑐 3
including
similarly unavoidable overlapped parts of signal 𝑆 1
, and they are multiplied by complex
coefficients 𝑐 1
and 𝑐 2
. These adjusted signals are denoted by 𝑋 𝑖 and can be written as follows
𝑋 1
∝ 𝑆 1
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 1
+
∆𝑓 2
) + 𝛼 1
(𝑓 )𝑆 2
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 1
−
∆𝑓 2
) (6.3)
𝑋 2
∝ 𝑐 1
𝛼 2
(𝑓 )𝑆 1
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 2
+
∆𝑓 2
) + 𝑐 1
𝑆 2
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 2
−
∆𝑓 2
) (6.4)
𝑋 3
∝ 𝑐 2
𝑒 𝑗 2𝜋 ∆𝑓𝜏
𝛼 3
(𝑓 )𝑆 1
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 3
+
∆𝑓 2
) + 𝑐 2
𝑒 𝑗 2𝜋 ∆𝑓𝜏
𝑆 2
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 3
−
∆𝑓 2
)
(6.5)
In equations (3-5), 𝛼 𝑖 (𝑓 ) for each 𝑋 i
denotes the filtering effect of the optical programmable
filter on the overlapped signal. 𝜏 represents the amount of delay and it approximately equals
to one symbol interval. In stage three, the signals 𝑋 i
with the corresponding dummy pumps
and a new CW pump laser 𝑃 2
are sent to another PPLN waveguide in which they are
coherently multiplexed through cascaded nonlinear processes of SFG-DFG. We denote this
new signal by 𝑆 ̃
1
:
𝑆 ̃
1
∝ (1 + 𝑐 1
𝛼 2
(𝑓 )+ 𝑐 2
𝑒 𝑗 2𝜋 ∆𝑓𝜏
𝛼 3
(𝑓 ))𝑆 1
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 3
+
∆𝑓 2
) + (𝛼 1
(𝑓 )+ 𝑐 1
+
𝑐 2
𝑒 𝑗 2𝜋 ∆𝑓𝜏
)𝑆 2
(2𝑓 𝑄𝑃𝑀 − 𝑓 𝑑 3
−
∆𝑓 2
) (6.6)
The interference caused by channel 𝑆 2
can be mitigated by tuning 𝑐 1
and 𝑐 2
complex coefficients. In this method, there is no need to estimate the amount of
68
channel spacing ∆𝑓 because the pump lasers are preserved from previous stages and
the two terms of the signals 𝑋 i
in Eqs. (6.3-6.5) are added coherently in stage 3 with
the exact channel spacing ∆𝑓 . The ICI mitigated signal 𝑆 ̃
1
, is filtered (denoted by
signal 𝑆 ̂
1
after filtering) and is sent to coherent detector. The similar process at port 2
of the programmable filter in Fig. 6.1 mitigates the ICI of channel 𝑆 2
. This process can
be followed by exchanging the signals 𝑆 1
and 𝑆 2
in Eqs. (6.3-6.6).
Figure 6.1 Conceptual block diagram of the proposed optical inter-channel interference (ICI) mitigation
method. In a PPLN waveguide, three copies of the signals are generated. In a programmable optical
filter, the signals are selected, delayed, and multiplied with complex coefficients by adjusting their
amplitudes and phases. In PPLN waveguides at ports 1 and 2, the signals are coherently multiplexed to
mitigate the ICI.
6.3 Experimental Setup
Figure 6.2 shows the experimental setup of the proposed all optical ICI mitigation method.
Two data channels, either 16QAM/16QAM signals or 16QAM/QPSK signals are generated.
To generate 16QAM/QPSK data signals, the channels are sent to separate QPSK and 16QAM
nested Mach-Zehnder modulators (MZM) and are modulated by independent data streams.
To create 16QAM/16QAM data channels, two lasers are coupled and sent to a 16QAM
S
1 S
2
𝒇
𝒇
2
1
S
1
(t) S
2
(t) S
2
(t-τ)
1
P
c
2
c
1
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
P
d1
P
d2
P
d3
DFG
S
1
(t) S
2
(t) S
2
(t-τ)
S
1 S
2
P
1
c
2
c
1
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
𝒇
P
d1 P
d2
P
d3
𝒇
SFG
DFG
Programmable
optical amplitude/
phase filter
(LCoS filter)
ICI Compensation for S
1
(PPLN)
ICI Compensation for S
2
(PPLN)
Copy generation
(PPLN)
Port 1
Port 2
Signals at port-1
of LCoS filter:
ICI mitigated
signal
ICI mitigated
signal
Overlapped
signals with ICI
69
MZM. To decorrelate the data, they become separated and independently delayed. Each signal
is sent to a pre-amplifier to maintain the same optical power for the channels and then
overlapped by using an optical 50/50 coupler. We placed a polarization controller in each
signal path and before 50/50 coupler to maximize the amount of interference. The overlapped
channels are amplified through an erbium-doped fiber amplifier (EDFA), and are sent to a
PPLN waveguide along with three coupled amplified pump lasers at ~1551, ~1552, and
~1553 nm and another CW pump laser at ~ 1562 nm. The QPM wavelength is temperature
tuned at approximately 1550 nm to achieve maximum conversion efficiency. Through the
cascaded nonlinear SFG-DFG processes, three signal copies are generated in PPLN and along
with the pump lasers are sent to a spatial light modulator (SLM) filter based on liquid crystal
on silicon (LCoS) technology to select desired channels from each copy, induce relative
delays, and adjust amplitudes and phases for the selected channels. The adjusted signals and
the pump lasers are coupled with another CW pump laser at ~ 1560 nm and sent to a new
PPLN waveguide where they are coherently multiplexed through the cascaded nonlinear
SFG-DFG processes to generate an ICI mitigated signal. The ICI mitigated signal is filtered
and sent to a coherent receiver to record the constellation and measure BER of the signal.
Figure 6.2 (a) Experimental setup. PC: polarization controller, PPLN: periodically poled lithium niobate, LCoS:
liquid crystal on silicon, (b) optical spectra after PPLN-1 and PPLN-2 waveguides.
Ch2
1 nm
2 nm 2 nm
Data1
Mod.
BPF
PC
1 nm
1 nm
Mod.
Data2
Ch1
Coherent
Receiver
11 nm
PC
EDFA
5 nm
𝑑 1
PC
𝑑 2
𝑑 3
𝑝 1
𝑝 2
1 nm
1 nm
a)
b)
Dummy pumps
Overlapped
Signals
PPLN-1
1538 1550 1562 (nm)
Generated copies
ICI
Mitigated
Signal
Adjusted Signals
PPLN-2
1538 1550 1562 (nm)
PPLN-1
(copy generation)
PPLN-2
(coherent multiplexing)
Programmable
optical filter
(LCoS filter )
70
6.4 Results
The proposed optical ICI mitigation method can be applied on channels with different
modulation formats. The proposed optical ICI mitigation method is also baud rate tunable.
The performance of the scheme is assessed for optical channels modulated with
16QAM/16QAM data signals and for optical channels with QPSK/16QAM data signals. In
each case, two different baud rates are considered, 20Gbaud signals and 25Gbaud signals. In
addition, the amount of spectral spacing between channels is varied to measure the system
performance for different amount of interference. For channels with 20 Gbaud signals,
channel spacings of 17.5GHz, 20GHz, and 25GHz; and for 25Gbaud signals, channels
spacings of 22.5GHz and 25GHz are considered. In all scenarios, there is no spectrum shaping
or filtering in the transmitter to emphasis the ICI effect.
Figure 6.3 shows the constellation diagrams of the two overlapped 16QAM/QPSK
channels. The constellations are recorded for the following three cases: (i) without optical ICI
mitigation (back to back), (ii) with one tap optical ICI mitigation (𝑐 2
is set to zero in Eq. 6.6
by blocking its corresponding pump laser), and (iii) with two taps optical ICI mitigation (both
taps 𝑐 1
and 𝑐 2
in Eq. (6.4-6.6) are tuned to mitigate the crosstalk). The ICI mitigation method
provides negligible benefit when the channel spacing is larger than the baud rate of the signals,
because the ICI is already insignificant. This case is shown for ∆𝑓 =25GHz in Fig 6. 3. When
the channel spacing is equal to or less than the signal baud rate, the ICI is significant and the
impact of ICI mitigation by the proposed method becomes noticeable. These cases are shown
for ∆𝑓 =20GHz and 17.5GHz in Fig. 6.3. The ICI mitigation is however not perfect and for
channels with more severe ICI ( ∆𝑓 =17.5GHz) the output signal quality is worse than the
channels with lower ICI (∆𝑓 =20GHz).
71
Figure 6.3 Experimentally recorded signal constellation diagrams with one tap and two taps optical ICI mitigation
and without (w/o) optical ICI compensation. The diagrams are obtained for overlapped channels of 20Gbaud
16QAM signal and QPSK signal at different channel spacings Δf.
Figure 6.4 illustrates the BER results of the overlapped QSPK and 16QAM
signals with and without optical ICI mitigation. The BER curves are measured for
20Gbaud signals with 17.5GHz and 22.5GHz channel spacings in Fig. 6.4(a,b), and
for 25 Gbaud signals with 22.5GHz and 25GHz channel spacings in Fig. 6.4(c,d). An
OSNR gain of ~4dB at a BER of 10
-3
is observed for QPSK channel with severe ICI
(∆𝑓 =17.5GHz) after optical ICI mitigation. Additionally, the scheme results in a
noticeable OSNR benefit at a BER of 10
-3
for the 16QAM data channel after optical
ICI mitigation.
2-taps
ICI Comp.
w/o ICI
Comp.
1-tap
ICI Comp.
2-taps
ICI Comp.
w/o ICI
Comp.
1-tap
ICI Comp.
2-taps
ICI Comp.
Channel spacing=17.5 GHz
Channel spacing= 20 GHz Channel spacing= 25 GHz
w/o ICI
Comp.
1-tap
ICI Comp.
16QAM Ch. QPSK Ch.
72
Figure 6.4 BER measurements with and without optical ICI compensation method for QPSK/16QAM
overlapped channels and for different channel spacings ∆𝒇 (a,b) 20Gbaud QSPK and 16QAM channels (c,d)
25Gbaud QPSK and 16QAM channels.
Figure 6.5 summarizes the obtained constellation diagrams for 20Gbaud 16QAM/16QAM
data channels after optical ICI mitigation with one tap and two taps and without any optical
ICI compensation (back to back). ICI mitigation is noticeable when ICI is large
(∆𝑓 =17.5GHz). Further, the results indicate that two taps ICI compensation allows for better
results.
-log
10
(BER)
OSNR (dB)
-log
10
(BER)
OSNR (dB)
-log
10
(BER)
OSNR (dB)
-log
10
(BER)
OSNR (dB)
2
2.5
3
3.5
4
4.5
5
10 15 20 25
Δf =20GHz with Comp.
Δf =17.5GHz with Comp.
Δf =20GHz without Comp.
Δf =17.5GHz without Comp.
2
2.5
3
3.5
4
4.5
12 17 22 27
Δf =25GHz with Comp.
Δf =22.5GHz with Comp.
Δf =25GHz without Comp.
Δf =22.5GHz without Comp.
2
2.5
3
3.5
18 20 22 24 26
Δf =20GHz with Comp.
Δf =17.5GHz without Comp.
Δf =20GHz without Comp.
Δf =17.5GHz without Comp.
20 Gbaud QPSK channel
25 Gbaud QPSK channel
20 Gbaud 16QAM channel
25 Gbaud 16QAM channel
2
2.5
3
3.5
20 22 24 26
Δf =25GHz with ICI Comp.
Δf =22.5GHz with ICI Comp.
Δf =25GHz without ICI Comp.
Δf =22.5GHz without ICI Comp.
a) b)
c)
d)
73
Figure 6.5 Experimentally measured signal constellation diagrams with one tap and two taps optical ICI
mitigation and without(w/o) optical ICI compensation for overlapped channels of 20Gbaud 16QAM signals at
different channel spacings Δf.
Figure 6.6 illustrates the BER results for the 16QAM channels for 20Gbaud and
25Gbaud signals and for different channel spacings. The BER are obtained for
20Gbaud overlapped 16QAM channels with channel spacings of 17.5GHz and 20GHz,
and for 25Gbaud overlapped 16QAM channels with channel spacings of 22.5GHz and
25GHz.
Figure 6.6 BER measurements with and without optical ICI compensation method for 16QAM/16QAM
overlapped channels and for different channel spacings ∆𝒇 (a,b) 20Gbaud 16QAM channel-1 and
channel-2 (c,d) 25Gbaud 16QAM channel-1 and channel-2.
2-taps
ICI Comp.
w/o ICI
Comp.
1-tap
ICI Comp.
2-taps
ICI Comp.
w/o ICI
Comp.
1-tap
ICI Comp.
2-taps
ICI Comp.
Channel spacing=17.5 GHz
Channel spacing= 20 GHz Channel spacing= 25 GHz
w/o ICI
Comp.
1-tap
ICI Comp.
16QAM Ch.1 16QAM Ch.2
-log 10 (BER)
OSNR (dB)
-log 10 (BER)
OSNR (dB)
-log 10 (BER)
OSNR (dB)
-log 10 (BER)
OSNR (dB)
2
2.5
3
3.5
18 20 22 24 26
Δf =20GHz with Comp.
Δf =17.5GHz without Comp.
Δf =20GHz without Comp.
Δf =17.5GHz without Comp.
20 Gbaud 16QAM channel-1
25 Gbaud 16QAM channel-1
20 Gbaud 16QAM channel-2
25 Gbaud 16QAM channel-2
2
2.5
3
3.5
20 22 24 26
Δf =25GHz with ICI Comp.
Δf =22.5GHz with ICI Comp.
Δf =25GHz without ICI Comp.
Δf =22.5GHz without ICI Comp.
a) b)
c) d)
2
2.5
3
3.5
18 20 22 24 26
Δf =20GHz with Comp.
Δf =17.5GHz without Comp.
Δf =20GHz without Comp.
Δf =17.5GHz without Comp.
2
2.5
3
3.5
20 22 24 26
Δf =25GHz with ICI Comp.
Δf =22.5GHz with ICI Comp.
Δf =25GHz without ICI Comp.
Δf =22.5GHz without ICI Comp.
74
Chapter 7 Optical Mitigation of Inter-Channel Crosstalk
for Multiple Spectrally Overlapped WDM Channels using
Nonlinear Wave Mixing
7.1 Introduction
It is considered fairly important to maximize the efficient utilization of the available
spectrum [77,78]. One approach to increasing the spectral efficiency in terms of
bits/sec/Hz is to spectrally overlap the data channels, which produces increased inter-
channel crosstalk [66,79].
This crosstalk can be mitigated by electronic means. In a multi-channel
wavelength-division-multiplexed (WDM) system, each wavelength channel is
individually recovered, and the information about a data channel’s spectrally close
adjacent channels is used by a digital signal processing (DSP) algorithm to reduce the
crosstalk [49,68,69,80].
Alternatively, inter-channel crosstalk can be mitigated using optical techniques.
Previously, optical multicasting, complex tailoring, and multiplexing was used to
reduce crosstalk for each individual channel for data recovery [75,76]. It might be
desirable to change this approach such that multiple, spectrally overlapped WDM
channels can be recovered simultaneously with reduced crosstalk in concurrent optical
nonlinear processes.
Here, we demonstrate optical mitigation of inter-channel crosstalk for multiple
spectrally overlapped 40-Gbit/s quadrature-phase-shift-keyed (QPSK) WDM
channels using nonlinear wave mixing without multi-channel detection and channel
spacing estimation. The optical ICI mitigation is performed by mixing the signals with
the conjugate copies of neighboring channels and their delay variants. In this method,
the conjugate signal copies are separated in two sets of even and odd channels, and the
amplitude, phase, and delays of signals in each set are adjusted. The signals are mixed
75
with the corresponding neighbors in either even or odd sets to mitigate the ICIs. The
system performance is experimentally evaluated for 20Gbaud QPSK overlapped data
channels under different channel spacing conditions. The improved signal
constellations and bit error rates demonstrate the effectiveness of this approach. Near
4dB OSNR gain is achieved for QPSK data channels at a BER of 10
-3
.
7.2 Concept
Figure 7.1 shows the conceptual block diagram of the proposed scheme for
optical ICI mitigation of seven overlapped data channels. The incoming overlapped
channels along with a CW pump laser are injected into a periodically poled lithium
niobate (PPLN) waveguide. Inside the PPLN, the conjugate copies of the signals are
generated through the cascaded nonlinear processes of second harmonic generation
(SHG) and difference frequency generation (DFG). Next, the signals are sent through
an optical programmable filter based on liquid crystal on silicon (LCoS) technology
and are separated at the ports 1 and 2 of the LCoS filter, respectively (Fig. 7.1(a)). The
signals at each port is then directed to separate optical ICI compensation modules. The
output of each module is filtered out to render the ICI mitigated desired channels.
Figure 7.1 (b) shows the conceptual block diagram of the embedded optical ICI
compensation module of Fig. 7.1(a) for ICI mitigation of even channels as an example.
This module is composed of two PPLNs and a programmable LCoS amplitude and
phase filter. To mitigate the ICI of even (odd) channels, the odd (even) channels from
the conjugate copies are filtered and the amplitude and phase of each conjugate copy
is adjusted in the first LCoS filter. The signals and the selected conjugate copies are
sent through the second PPLN waveguide where each target channel is mixed with
neighboring cross talk channels. By mixing the target channels with amplitude/phase-
adjusted cross talk channels, a moderate level of ICI mitigation can be obtained. The
signals and their conjugates are then sent to another LCoS filter where new complex
coefficients and delays are properly applied to the conjugate copies of the crosstalk
76
signals. The target signals and their corresponding cross talk signals are mixed
coherently in another PPLN waveguide where the ICI of the target channels are further
mitigated through the cascaded nonlinear processes of SHG-DFG. As an example,
suppose signal S2 is a target signal, this signal is mixed with its neighboring channels
and their delay variants to reduce the crosstalk of the channel, i.e., the signal is
coherently added to c1.S1, c3.S3, c1
'
.S1(t-τ1) and c3
'
.S3(t-τ3) where c1, c3 ,c1
'
, and c3
'
are
the appropriate complex coefficients adjusted in LCoS filter. Note that since the pump
laser is preserved through the nonlinear processes, the desired signal and the adjacent
interfering signals are added with the same channel spacing as of the overlapped
channels. Therefore, accurate channel spacing estimation is unnecessary in this
method.
Figure 7.1 (a) Conceptual diagram of the proposed optical inter-channel interference (ICI) mitigation
method. The signal conjugate copies are generated in the first PPLN waveguide. In an optical
programmable filter, the conjugate copies of data channels are separated in two sets of even and odd
channels and are adjusted with desired complex taps. The original signals with the even and odd sets of
conjugate copies are directed to optical ICI compensation modules. (b) Conceptual diagram of the
optical ICI compensation module for even channels. This module is composed of two PPLNs and a
LCoS filter for amplitude, phase and delay adjustments. The target signals are mixed with neighboring
channels and their delayed variants to mitigate the ICI.
ω 2
ω 2
ω
2
Optical ICI
Compensation
(Even channels) Programmable
optical
amplitude and
phase filter
(LCoS filter)
Conjugate copy
generation
(PPLN)
Optical ICI
Compensation
(Odd channels)
Port-1
Port-2
Optical
filter
Optical
filter
Reduced ICI
Incoming optical
overlapped channels
a)
b)
λ
ω
2
ω
2
.𝑆
∗
+
.𝑆
∗
( − )
.𝑆
∗
+
.𝑆
∗
( − )
.𝑆
∗
+
.𝑆
∗
( − )
.𝑆
∗
+
.𝑆
∗
( − )
WDM
λ-conversion
(PPLN)
WDM
λ-conversion
(PPLN)
LCoS filter
(complex taps,
delays)
Reduced
ICI
Optical ICI compensation at port 1 (even channels):
λ
ω
2
ω
2
𝑆
∗
𝑆
∗
𝑆
∗
𝑆
∗
𝑆
∗
𝑆
∗
𝑆
∗
SHG+ DFG Mixing
Conjugate copy generation via SFG-DFG
nonlinear processes in the first PPLN:
77
7.3 Experimental Setup
Figure 7.2 shows the experimental setup of the optical ICI mitigation method for seven
overlapped channels with channel spacing Δf. Odd and even channels are modulated
with independent QPSK data in separate Mach Zehnder modulators. All channels
along with a CW pump laser at 1540 nm are amplified and then sent into a PPLN
waveguide to generate the conjugate copies.
The quasi-phase matching (QPM) wavelength of PPLN is temperature-tuned to
the CW pump wavelength. The signals, including the pump, are sent into a spatial light
modulator (SLM) filter based on LCoS technology for channel selection and
amplitude/phase adjustment. In the LCoS, the conjugate copies of the neighboring
channels of each target channel are selected. The selected and adjusted signals along
with the pump are sent to a second PPLN with the same QPM as the first PPLN.
Through the second PPLN, each target channel is mixed with properly amplitude and
phase adjusted cross talk signals. The signals are sent to another LCoS filter where
complex coefficients and delays are applied on the signals. The signals and the pump
are injected into another PPLN waveguide where the target channels and the crosstalk
channels with complex coefficients and delays are mixed to mitigate the ICI of the
target channels. The channels with reduced ICI are filtered and sent to coherent
receiver to detect the constellation diagrams and measure the BER of the signals. In
this method, there is no need to estimate the amount of channel spacing ∆f since the
lasers are preserved from previous stages and the signals are mixed coherently with
the exact channel spacing.
78
Figure 7.2 (a) Experimental setup. PC: polarization Controller, PPLN: periodically poled lithium
niobate, LCoS: liquid crystal on silicon, (b) optical spectra of conjugate copies of 20 Gbaud overlapped
QPSK signals (A) and ICI mitigated channels 1,3,5 and 7 (odd channels) after the last nonlinear stage
(B).
7.3 Experimental Results
The performance of the system is assessed for overlapped channels of seven overlapped 20
Gbaud QPSK signals in different channel spacing of 17.5, 20, and 25 GHz. Figure 7.3 shows
the constellation diagrams of channels 1, 3, and 6 with and without the ICI mitigation method.
The ICI mitigation for the smaller channel spacing of 17.5 GHz is more significant than the
channel spacing of 20 GHz. The ICI mitigation is insignificant when the channel spacing is
larger than the baud rate of the signals, i.e., 25 GHz.
Figure 7.3 Experimentally measured signal constellation diagrams of channels 1, 3, and 6 with(w.) and
without(w/o) optical ICI mitigation method for 20 Gbaud overlapped QPSK signals and at channel
spacings, Δf of 17.5 GHz, 20 GHz and 25 GHz.
Ch5
Ch2
5 nm 3 nm 5 nm
Data1
Mod.
BPF
PC
CW Pump
(
𝑝 1540 nm)
1 nm
5 nm
Mod.
Data2
PPLN
Ch3
Ch7
Ch4
Coherent
Receiver
LCoS
Filter
13 nm
PPLN
LCoS
Filter
PC
EDFA
13 nm
PPLN
Ch1
Ch6
A
B
0 100 200 300 400 500 600 700 800 900 1000
-70
-60
-50
-40
-30
-20
-10
0
0 100 200 300 400 500 600 700 800 900 1000
-80
-70
-60
-50
-40
-30
-20
-10
Pump
Conjugate copies
of 7 channels
7 overlapped
Channels
ICI mitigated
channels 1,3,5,7
1540 1536 1544 nm
1540 1536 1544 nm
Conjugate
Coopies
Pump
A
B
Channel-6 Channel-1 Channel-3 Channel-6
Channel spacing=17.5 GHz
Channel spacing= 20 GHz
Channel-1 Channel-3
w/o optical ICI
mitigation
w. optical ICI
mitigation
Channel spacing= 25 GHz
Channel-6 Channel-1 Channel-3
79
Figures 7.4(a,b) shows the BER results for 20-Gbaud QPSK signals for channel
spacings of 17.5 GHz and 20 GHz respectively. As it can be seen, the ICI mitigation results
in lower BER at same OSNR values for both channel spacings of 17.5 GHz and 20 GHz. This
scheme provides near 4dB OSNR gain for QPSK data channels at a BER of 10
-3
.
Figure 7.4 BER measurements with(w.) and without(w/o) optical ICI compensation method for QPSK
overlapped channels (1,3 and 6) and for different channel spacing conditions. (a) ∆f =17.5 GHz (b) ∆f
= 20 GHz.
2
2.5
3
3.5
4
4.5
10 15 20 25
ch1 w/o ICI Comp.
ch1 w. ICI Comp.
ch3 w/o ICI Comp.
ch3 w. ICI Comp.
ch6 w/o ICI Comp.
ch6 w. ICI Comp.
2
2.5
3
3.5
4
4.5
5
10 15 20 25
ch1 w/o ICI Comp.
ch1 w. ICI Comp.
ch3 w/o ICI Comp.
ch3 w. ICI Comp.
ch6 w/o ICI Comp.
ch6 w. ICI Comp.
20 Gbaud16QAM, ∆f=17.5 Hz 20 Gbaud16QAM, ∆f =20 GHz
-log
10
(BER)
OSNR (dB)
-log
10
(BER)
OSNR (dB)
a) b)
80
References
1. A. E. Willner, S. Khaleghi, M. R. Chitgarha, and O. F. Yilmaz, "All-optical
signal processing," J. Light. Technol. 32, 660–680 (2014).
2. G. P. (Govind P. . Agrawal, Nonlinear Fiber Optics (Elsevier Science, 2013).
3. G. Li, "Recent advances in coherent optical communication," Adv. Opt.
Photonics 1, 279 (2009).
4. P. Winzer, "Beyond 100G ethernet," IEEE Commun. Mag. 48, 26–30 (2010).
5. E. Ip, A. P. T. Lau, D. J. F. Barros, and J. M. Kahn, "Coherent detection in
optical fiber systems," Opt. Express 16, 753 (2008).
6. Y. Koizumi, K. Toyoda, M. Yoshida, and M. Nakazawa, "1024 QAM (60
Gbit/s) single-carrier coherent optical transmission over 150 km," Opt. Express
20, 12508 (2012).
7. A. H. Gnauck, P. J. Winzer, A. Konczykowska, F. Jorge, J.-Y. Dupuy, M. Riet,
G. Charlet, B. Zhu, and D. W. Peckham, "Generation and Transmission of 21.4-
Gbaud PDM 64-QAM Using a Novel High-Power DAC Driving a Single I/Q
Modulator," J. Light. Technol. 30, 532–536 (2012).
8. R. Schmogrow, D. Hillerkuss, M. Dreschmann, M. Huebner, M. Winter, J.
Meyer, B. Nebendahl, C. Koos, J. Becker, W. Freude, and J. Leuthold, "Real-
Time Software-Defined Multiformat Transmitter Generating 64QAM at 28
GBd," IEEE Photonics Technol. Lett. 22, 1601–1603 (2010).
9. M. R. Chitgarha, S. Khaleghi, Z. Bakhtiari, M. Ziyadi, O. Gerstel, L. Paraschis,
C. Langrock, M. M. Fejer, and A. E. Willner, "Demonstration of reconfigurable
optical generation of higher-order modulation formats up to 64 QAM using
optical nonlinearity," Opt. Lett. 38, 3350 (2013).
10. S. Khaleghi, M. R. Chitgarha, O. F. Yilmaz, M. Tur, M. W. Haney, C.
Langrock, M. M. Fejer, and A. E. Willner, "Reconfigurable optical quadrature
amplitude modulation converter/encoder using a tunable complex coefficient
optical tapped delay line.," Opt. Lett. 38, 1600–2 (2013).
11. J. Leuthold, C. Koos, and W. Freude, "Nonlinear silicon photonics," Nat.
Photonics 4, 535–544 (2010).
12. G. Steinmeyer, D. H. Sutter, L. Gallmann, N. Matuschek, and U. Keller,
"Frontiers in Ultrashort Pulse Generation: Pushing the Limits in Linear and
Nonlinear Optics.," Science 286, 1507–1512 (1999).
13. S. Khaleghi, O. F. Yilmaz, M. R. Chitgarha, M. Tur, N. Ahmed, S. R. Nuccio,
I. M. Fazal, Xiaoxia Wu, M. W. Haney, C. Langrock, M. M. Fejer, and A. E.
81
Willner, "High-Speed Correlation and Equalization Using a Continuously
Tunable All-Optical Tapped Delay Line," IEEE Photonics J. 4, 1220–1235
(2012).
14. S. J. B. Yoo, "Wavelength conversion technologies for WDM network
applications," J. Light. Technol. 14, 955–966 (1996).
15. J. Kakande, R. Slavík, F. Parmigiani, A. Bogris, D. Syvridis, L. Grüner-Nielsen,
R. Phelan, P. Petropoulos, and D. J. Richardson, "Multilevel quantization of
optical phase in a novel coherent parametric mixer architecture," Nat. Photonics
5, 748–752 (2011).
16. H. Hu, E. Palushani, M. Galili, H. C. H. Mulvad, A. Clausen, L. K. Oxenløwe,
and P. Jeppesen, "640 Gbit/s and 128 Tbit/s polarisation insensitive all optical
wavelength conversion," Opt. Express 18, 9961 (2010).
17. A. Bogoni, X. Wu, S. R. Nuccio, J. Wang, Z. Bakhtiari, and A. E. Willner,
"Photonic 640-Gb/s Reconfigurable OTDM Add–Drop Multiplexer Based on
Pump Depletion in a Single PPLN Waveguide," IEEE J. Sel. Top. Quantum
Electron. 18, 709–716 (2012).
18. S. Radic, "Parametric signal processing," IEEE J. Sel. Top. Quantum Electron.
18, 670–680 (2012).
19. G. Agrawal, "Nonlinear Fiber Optics," New York 467 (2001).
20. M. Adams, "Optical waves in crystals," IEEE J. Quantum Electron. 20, 1294
(1984).
21. A. E. Willner, C. Langrock, J. E. McGeehan, M. M. Fejer, and S. Kumar, "All-
Optical Signal Processing Using χ
(2)
Nonlinearities in Guided-Wave Devices,"
J. Light. Technol. Vol. 24, Issue 7, pp. 2579- 24, 2579 (2006).
22. B. J. Eggleton, B. Luther-Davies, and K. Richardson, "Chalcogenide
photonics," Nat. Photonics 5, 141–148 (2011).
23. I. Liberal, A. M. Mahmoud, Y. Li, B. Edwards, and N. Engheta, "Photonic
doping of epsilon-near-zero media," Science (80-. ). 355, 1058–1062 (2017).
24. S. Takasaka, Y. Taniguchi, M. Takahashi, J. Hiroishi, M. Tadakuma, and R.
Sugizaki, "Wideband parametric processing with 1-dB bandwidth of 40nm
using dispersion stable PM-HNLF," in European Conference on Optical
Communication, ECOC (2014).
25. P. J. Winzer and R. J. Essiambre, "Advanced modulation formats for high-
capacity optical transport networks," J. Light. Technol. 24, 4711–4728 (2006).
26. S. J. Savory, "Digital filters for coherent optical receivers," Opt. Express 16,
804 (2008).
82
27. N. R. Newbury, "Searching for applications with a fine-tooth comb," Nat.
Photonics 5, 186–188 (2011).
28. G. Contestabile, M. Presi, and E. Ciaramella, "Multiple wavelength conversion
for WDM multicasting by FWM in an SOA," IEEE Photonics Technol. Lett.
16, 1775–1777 (2004).
29. A. Bogoni, A. Malacarne, G. Berrettini, G. Meloni, L. Potì, and N. Sambo,
"Optical Multicasting of 16QAM Signals in Periodically-Poled Lithium
Niobate Waveguide," J. Light. Technol. Vol. 31, Issue 11, pp. 1797-1803
31, 1797–1803 (2013).
30. A. E. Willner, O. F. Yilmaz, J. Wang, X. Wu, A. Bogoni, L. Zhang, and S. R.
Nuccio, "Optically efficient nonlinear signal processing," IEEE J. Sel. Top.
Quantum Electron. 17, 320–322 (2011).
31. B. Olsson and D. J. Blumenthal, "WDM to OTDM multiplexing using an
ultrafast all-optical wavelength converter," IEEE Photonics Technol. Lett. 13,
1005–1007 (2001).
32. E. J. M. Verdurmen, G. D. Khoe, A. M. J. Koonen, and H. De Waardt, "All-
optical data format conversion from WDM to OTDM based on FWM," Microw.
Opt. Technol. Lett. 48, 992–994 (2006).
33. M. R. Chitgarha, S. Khaleghi, M. Ziyadi, A. Almaiman, A. Mohajerin-Ariaei,
O. Gerstel, L. Paraschis, C. Langrock, M. M. Fejer, J. Touch, and A. E. Willner,
"Demonstration of tunable optical generation of higher-order modulation
formats using nonlinearities and coherent frequency comb," Opt. Lett. 39,
4915–4918 (2014).
34. J. Kakande, A. Bogris, R. Slavik, F. Parmigiani, D. Syvridis, P. Petropoulos, D.
Richardson, M. Westlund, and M. Sköld, "QPSK Phase and Amplitude
Regeneration at 56 Gbaud in a Novel Idler-Free Non-Degenerate Phase
Sensitive Amplifier," in Optical Fiber Communication Conference/National
Fiber Optic Engineers Conference 2011, OSA Technical Digest (CD) (Optical
Society of America, 2011), pp. OMT4-OMT4.
35. M. Ziyadi, A. Mohajerin-Ariaei, A. Almaiman, Y. Cao, M. R. Chitgarha, L.
Paraschis, M. Tur, C. Langrock, M. M. Fejer, J. D. Touch, and A. E. Willner,
"Optical channel de-aggregation of quadrature-phase-shift-keying and eight-
phase-shift-keying data using mapping onto constellation axes," Opt. Lett. 40,
4899 (2015).
36. G.-W. Lu, T. Miyazaki, K. Mishina, S. M. Nissanka, A. Maruta, S. Mitani, K.
Ishida, K. Shimizu, T. Hatta, and K. Kitayama, "Optical phase erasure based on
FWM in HNLF enabling format conversion from 320-Gb/s RZ- DQPSK to 160-
Gb/s RZ-DPSK," (n.d.).
83
37. J. P. Gordon and L. F. Mollenauer, "Phase noise in photonic communications
systems using linear amplifiers," Opt. Lett. 15, 1351 (1990).
38. A. Demir, "Nonlinear phase noise in optical-fiber-communication systems," J.
Light. Technol. 25, 2002–2032 (2007).
39. F. Parmigiani, K. R. Bottrill, G. Hesketh, P. Horak, P. Petropoulos, and D. J.
Richardson, "Signal Regeneration Techniques for Advanced Modulation
Formats," in CLEO: 2014 (OSA, 2014), p. STu2J.1.
40. X. Li, X. Chen, G. Goldfarb, E. Mateo, I. Kim, F. Yaman, and G. Li, "Electronic
post-compensation of WDM transmission impairments using coherent
detection and digital signal processing," Opt. Express 16, 880 (2008).
41. R. Slavík, F. Parmigiani, J. Kakande, C. Lundström, M. Sjödin, P. A.
Andrekson, R. Weerasuriya, S. Sygletos, A. D. Ellis, L. Grüner-Nielsen, D.
Jakobsen, S. Herstrøm, R. Phelan, J. O’Gorman, A. Bogris, D. Syvridis, S.
Dasgupta, P. Petropoulos, and D. J. Richardson, "All-optical phase and
amplitude regenerator for next-generation telecommunications systems," Nat.
Photonics 4, 690–695 (2010).
42. Z. Zheng, L. An, Z. Li, X. Zhao, and X. Liu, "All-optical regeneration of
DQPSK/QPSK signals based on phase-sensitive amplification," Opt. Commun.
281, 2755–2759 (2008).
43. T. Umeki, M. Asobe, H. Takara, Y. Miyamoto, and H. Takenouchi, "Multi-span
transmission using phase and amplitude regeneration in PPLN-based PSA,"
Opt. Express 21, 18170 (2013).
44. K. Solis-Trapala, J. Kurumida, M. Gao, T. Inoue, and S. Namiki, "Tunable
Optical Parametric Regenerator Assessment in a 43 Gb/s RZ-DPSK Signal
Transmission Link," IEEE Photonics Technol. Lett. 26, 629–632 (2014).
45. M. Asobe, T. Umeki, H. Takenouchi, and Y. Miyamoto, "In-line phase-
sensitive amplification of QPSK signal using multiple quasi-phase matched
LiNbO 3 waveguide," Opt. Express 22, 26642 (2014).
46. M. R. Chitgarha, S. Khaleghi, M. Ziyadi, A. Mohajerin-Ariaei, A. Almaiman,
W. Daab, D. Rogawski, M. Tur, J. D. Touch, C. Langrock, M. M. Fejer, and A.
E. Willner, "Demonstration of all-optical phase noise suppression scheme using
optical nonlinearity and conversion/dispersion delay," Opt. Lett. 39, 2928
(2014).
47. J. Touch, A. Mohajerin-Ariaei, M. Chitgarha, M. Ziyadi, S. Khaleghi, Y.
Akasaka, J. -y. Yang, and M. Sekiya, "The Impact of Errors on Differential
Optical Processing," USC/ISI Tech. Rep. ISI-TR-690, (2014).
48. L. Kazovsky, "Balanced phase-locked loops for optical homodyne receivers:
84
Performance analysis, design considerations, and laser linewidth requirements,"
J. Light. Technol. 4, 182–195 (1986).
49. S. Shinada, M. Nakamura, Y. Kamio, and N. Wada, "16-QAM optical packet
switching and real-time self-homodyne detection using polarization-
multiplexed pilot-carrier," Opt. Express 20, B535 (2012).
50. T. Miyazaki and F. Kubota, "PSK self-homodyne detection using a pilot carrier
for multibit/symbol transmission with inverse-RZ signal," IEEE Photonics
Technol. Lett. 17, 1334–1336 (2005).
51. M. Nakamura, Y. Kamio, and T. Miyazaki, "Linewidth-tolerant 10-Gbit/s 16-
QAM transmission using a pilot-carrier based phase-noise cancelling
technique," Opt. Express 16, 10611 (2008).
52. K. Balakier, M. J. Fice, L. Ponnampalam, A. J. Seeds, and C. C. Renaud,
"Monolithically Integrated Optical Phase Lock Loop for Microwave
Photonics," J. Light. Technol. 32, 3893–3900 (2014).
53. M. R. Chitgarha, A. Mohajerin-Ariaei, Y. Cao, M. Ziyadi, S. Khaleghi, A.
Almaiman, J. D. Touch, C. Langrock, M. M. Fejer, and A. E. Willner, "Tunable
Homodyne Detection of an Incoming QPSK Data Signal Using Two Fixed
Pump Lasers," J. Light. Technol. 33, 1344–1350 (2015).
54. C. Henry, "Theory of the linewidth of semiconductor lasers," IEEE J. Quantum
Electron. 18, 259–264 (1982).
55. A. Silva, M. Drummond, R. Ribeiro, and P. Monteiro, "Impact and
compensation techniques of laser phase noise in ultra-dense coherent access
networks," in 2013 15th International Conference on Transparent Optical
Networks (ICTON) (IEEE, 2013), pp. 1–4.
56. X. Shao, P.-Y. Kam, and C. Yu, "Maximum likelihood sequence detection in
laser phase noise-impaired coherent optical systems," Opt. Express 19, 22600
(2011).
57. J. M. Simmons, Optical Network Design and Planning, Optical Networks
(Springer International Publishing, 2014).
58. J. Berthold, A. A. M. Saleh, L. Blair, and J. M. Simmons, "Optical Networking:
Past, Present, and Future," J. Light. Technol. 26, 1104–1118 (2008).
59. A. A. M. Saleh and J. M. Simmons, "All-Optical Networking—Evolution,
Benefits, Challenges, and Future Vision," Proc. IEEE 100, 1105–1117 (2012).
60. B. Mukherjee, J. Zhang, J. Zhang, M. Song, X. Yu, Y. Zhao, and Y. Ji,
"Dynamic Traffic Grooming in Sliceable Bandwidth-Variable Transponder-
Enabled Elastic Optical Networks," J. Light. Technol. Vol. 33, Issue 1, pp. 183-
191 33, 183–191 (2015).
85
61. M. Irfan, "All-Optical Traffic Grooming in Elastic Optical Network," in Optical
Fiber Communication Conference/National Fiber Optic Engineers Conference
2013 (OSA, 2013), p. OW3A.3.
62. K. R. Bottrill, G. D. Hesketh, F. Parmigiani, P. Horak, D. J. Richardson, and P.
Petropoulos, "An Optical Phase Quantiser Exhibiting Suppressed Phase
Dependent Gain Variation," in Optical Fiber Communication Conference
(OSA, 2014), p. W3F.7.
63. N. Sambo, P. Castoldi, A. D’Errico, E. Riccardi, A. Pagano, M. S. Moreolo, J.
M. Fabrega, D. Rafique, A. Napoli, S. Frigerio, E. H. Salas, G. Zervas, M.
Nolle, J. K. Fischer, A. Lord, and J. P. F.-P. Gimenez, "Next generation
sliceable bandwidth variable transponders," IEEE Commun. Mag. 53, 163–171
(2015).
64. P. Sayyad Khodashenas, J. M. Rivas-Moscoso, B. Shariati, D. M. Marom, D.
Klonidis, and I. Tomkos, "Investigation of Spectrum Granularity for
Performance Optimization of Flexible Nyquist-WDM-Based Optical
Networks," J. Light. Technol. 33, 4767–4774 (2015).
65. C. Liu, J. Pan, T. Detwiler, A. Stark, Y.-T. Hsueh, G.-K. Chang, and S. E.
Ralph, "Joint digital signal processing for superchannel coherent optical
communication systems," Opt. Express 21, 8342 (2013).
66. M. Tadakuma, K. Watanabe, T. Saito, T. Tsuritani, K. Maeda, K. Igarashi, M.
Suzuki, Y. Tsuchida, R. Sugizaki, I. Morita, and K. Imamura, "1.03-
Exabit/s.km Super-Nyquist-WDM Transmission over 7,326-km Seven-Core
Fiber," in 39th European Conference and Exhibition on Optical
Communication (ECOC 2013) (Institution of Engineering and Technology,
2013), pp. 1251–1253.
67. M. Xiang, S. Fu, M. Tang, H. Tang, P. Shum, and D. Liu, "Nyquist WDM
superchannel using offset-16QAM and receiver-side digital spectral shaping,"
Opt. Express 22, 17448 (2014).
68. J. Pan, C. Liu, T. Detwiler, A. J. Stark, Y.-T. Hsueh, and S. E. Ralph, "Inter-
Channel Crosstalk Cancellation for Nyquist-WDM Superchannel
Applications," J. Light. Technol. 30, 3993–3999 (2012).
69. S. Chandrasekhar and X. Liu, "Experimental investigation on the performance
of closely spaced multi-carrier PDM-QPSK with digital coherent detection,"
Opt. Express 17, 21350 (2009).
70. B. C. Thomsen, D. Lavery, K. Shi, M. Sato, P. Bayvel, and R. Maher,
"Frequency Diversity MIMO Detection for DP- QAM Transmission," J. Light.
Technol. Vol. 33, Issue 7, pp. 1388-1394 33, 1388–1394 (2015).
71. E. Al-Rawachy, R. P. Giddings, and J. M. Tang, "Experimental demonstration
86
of a DSP-based cross-channel interference cancellation technique for
application in digital filter multiple access PONs," Opt. Express 25, 3850
(2017).
72. T. Zeng, "Superchannel transmission system based on multi-channel
equalization," Opt. Express 21, 14799 (2013).
73. S. Yamamoto, K. Saito, F. Hamaoka, T. Matsuda, A. Naka, and H. Maeda,
"Characteristics Investigation of High-Speed Multi-Carrier Transmission Using
MIMO-Based Crosstalk Compensation in Homodyne Detection Scheme," J.
Light. Technol. 34, 2824–2832 (2016).
74. F. Lehmann, P. Ramantanis, and Y. Frignac, "Joint Channel Estimation,
Interference Mitigation, and Decoding for WDM Coherent Optical
Communications," J. Opt. Commun. Netw. 6, 315 (2014).
75. Y. Cao, M. Ziyadi, A. Mohajerin-Ariaei, A. Almaiman, P. Liao, C. Bao, F.
Alishahi, A. Falahpour, B. Shamee, J. Yang, Y. Akasaka, M. Sekiya, M. Tur,
C. Langrock, M. Fejer, J. Touch, and A. E. Willner, "Reconfigurable optical
inter-channel interference mitigation for spectrally overlapped QPSK signals
using nonlinear wave mixing in cascaded PPLN waveguides," Opt. Lett. 41,
3233 (2016).
76. A. Mohajerin-Ariaei, M. Ziyadi, Y. Cao, A. Almaiman, F. Alishahi, A.
Fallahpour, C. Bao, P. Liao, B. Shamee, J. Touch, M. Tur, C. Langrock, M. M.
Fejer, and A. E. Willner, "Demonstration of Tunable Mitigation of Interchannel
Interference of Spectrally Overlapped 16-QAM/QPSK Data Channels using
Wave Mixing of Delayed Copies," in Optical Fiber Communication
Conference (OSA, 2017), p. Th3J.5.
77. P. J. Winzer, "High-Spectral-Efficiency Optical Modulation Formats," J. Light.
Technol. Vol. 30, Issue 24, pp. 3824-3835 30, 3824–3835 (2012).
78. I. B. Martins, I. Aldaya, G. Perez-Sanchez, and P. Gallion, "Optimization of
spectral band utilization in gridless WDM optical network," in A. K. Srivastava,
ed. (2014), p. 90100K.
79. S. Yamamoto, K. Saito, A. Naka, and H. Maeda, "Compatibility between
nonlinear compensation and crosstalk compensation using MIMO processing
in super-high-density multi-carrier transmission system," in 2015 European
Conference on Optical Communication (ECOC) (IEEE, 2015), pp. 1–3.
80. K. Shibahara, A. Masuda, S. Kawai, and M. Fukutoku, "Multi-stage successive
interference cancellation for spectrally-efficient super-Nyquist transmission,"
in 2015 European Conference on Optical Communication (ECOC) (IEEE,
2015), pp. 1–3.
81. F. Parmigiani, R. Slavík, J. Kakande, P. Petropoulos, and D. Richardson,
87
"Optical Regeneration," in All-Optical Signal Processing, S. Wabnitz and B.
Eggleton, eds. (Springer International Publishing, 2015), pp. 129–155.
82. K. R. H. Bottrill, R. Kakarla, F. Parmigiani, D. Venkitesh, and P. Petropoulos,
"Phase regeneration of QPSK signal in SOA using single-stage, wavelength
converting PSA," IEEE Photonics Technol. Lett. 28, 205–208 (2016).
83. A. Mohajerin-Ariaei, M. Ziyadi, M. R. Chitgarha, A. Almaiman, Y. Cao, B.
Shamee, J. Yang, Y. Akasaka, M. Sekiya, S. Takasaka, R. Sugizaki, J. D.
Touch, M. Tur, C. Langrock, M. M. Fejer, and A. E. Willner, "Phase noise
mitigation of QPSK signal utilizing phase-locked multiplexing of signal
harmonics and amplitude saturation," Opt. Lett. 40, 3328–3331 (2015).
84. N.-K. Kjøller, F. Da Ros, K. M. Røge, M. Galili, and L. K. Oxenløwe, "QPSK
Regeneration without Active Phase-Locking," in Conference on Lasers and
Electro-Optics, OSA Technical Digest (2016) (Optical Society of America,
2016), p. JTh2A.119-JTh2A.119.
85. A. Almaiman, Y. Cao, M. Ziyadi, A. Mohajerin-Ariaei, P. Liao, C. Bao, F.
Alishahi, A. Fallahpour, B. Shamee, N. Ahmed, A. J. Willner, Y. Akasaka, T.
Ikeuchi, S. Takasaka, R. Sugizaki, S. Wilkinson, J. D. Touch, M. Tur, and A.
E. Willner, "Experimental demonstration of phase-sensitive regeneration of a
binary phase-shift keying channel without a phase-locked loop using Brillouin
amplification," Opt. Lett. 41, 5434–5437 (2016).
86. A. Almainman, Y. Cao, M. Ziyadi, A. Mohajerin-Ariaei, P. Liao, C. Bao, F.
Alishahi, A. Fallahpour, B. Shamee, J. Touch, Y. Akasaka, T. Ikeuchi, S.
Wilkinson, M. Tur, and A. E. Willner, "Experimental Demonstration of Phase-
Sensitive Regeneration of a 20-40 Gb/s QPSK Channel without Phase-Locked
Loop using Brillouin Amplification," in ECOC 2016; 42nd European
Conference on Optical Communication (2016), pp. 1–3.
87. W. Wei, L. Yi, Y. Jaouën, and W. Hu, "Bandwidth-tunable narrowband
rectangular optical filter based on stimulated Brillouin scattering in optical
fiber," Opt. Express 22, 23249 (2014).
88. B. J. Eggleton, C. G. Poulton, and R. Pant, "Inducing and harnessing stimulated
Brillouin scattering in photonic integrated circuits," Adv. Opt. Photonics 5, 536
(2013).
89. A. Annoni and F. Morichetti, "Enhancing the Sensitivity of Interferometer
Based In-Band OSNR Monitoring by Narrow Band Filtering," J. Light.
Technol. Vol. 31, Issue 9, pp. 1447-1453 31, 1447–1453 (2013).
Abstract (if available)
Abstract
The exponential trend of bandwidth demanding applications such as cloud computing, photos and video sharing, data storage systems and recent technological advances in high-speed data networks have created a demand for higher speeds in data processing and transmission. Over the past few years, the increase in system capacity has been realized by a combination of coherent technologies and advanced modulation formats. Coherent detection and advanced optical modulations with the use of high-speed digital signal processing (DSP), encode information on the four optical domains of wavelength, amplitude, phase, and polarization. In particular, coherent transceivers enable systems that can support spectrally efficient modulation formats to transmit and receive many bits of information in one symbol time. ❧ High speed optical signal processing has been one of the main research areas in photonics for more than a decade. Optical signal processing has been of great interest due to its inherent ultrafast THz bandwidth and its potentially phase-preserving nature. One of the main interest for using optical signal processing techniques is the optical methods such as the ones based on nonlinear processes which do not need to switch every bit as electronic transistors do. Optical amplifiers, can amplify very high-speed signals (Tb/s) without processing the signal at the bit level. Considering wavelength conversion technique as an another example, by using a pump laser and a nonlinear element, ultra-high-speed optical data can be transferred from one wavelength to another as optical signals pass through the optical elements. More importantly, advances in photonic integrated circuits (PICs), nonlinear materials and devices with higher efficiencies are the important factors for any practical realization of optical signal processing systems in the future. ❧ The current dissertation explores the potential of ultra-high speed optical systems to assist DSP in processing and noise mitigation of huge amounts of data. In general, phase or amplitude noise can cause a degradation in the received data signal-to-noise ratio, resulting in a system power penalty. In specific, phase noise originating from the interaction of ASE noise and Kerr nonlinearity can pose a key limitation in such systems. This dissertation demonstrates optical systems to mitigate the noise in the optical domain which can provide several advantages such as avoiding the impact of optical-to-electronic conversion and supporting in-line signal processing for high baud rate signals. By utilizing various forms of photonic nonlinear interactions, different functionalities are demonstrated which includes
Linked assets
University of Southern California Dissertations and Theses
Conceptually similar
PDF
Reconfigurable and flexible high-speed optical signal processing and spectrally shared optical subsystems
PDF
Reconfigurable high‐speed optical signal processing and high‐capacity optical transmitter
PDF
High-speed and reconfigurable all-optical signal processing for phase and amplitude modulated signals
PDF
Reconfigurable optical signal processing for efficient spectrum utilization in high-speed optical communication systems
PDF
All-optical signal processing toward reconfigurable optical networks
PDF
Microresonator-based Kerr frequency comb for high-speed optical communications and signal processing
PDF
Optical wave mixing for tunable delays and high‐speed signal processing
PDF
Reconfigurable high speed optical signal processing for optical communication and modulation format manipulation
PDF
Optical signal processing for high-speed, reconfigurable fiber optic networks
PDF
Optical signal processing for enabling high-speed, highly spectrally efficient and high capacity optical systems
PDF
Nonlinear optical signal processing for high-speed, spectrally efficient fiber optic systems and networks
PDF
Integrated silicon waveguides and specialty optical fibers for optical communications system applications
PDF
optical communication systems and space-time wave packet generation using frequency combs and spatial modes
PDF
Applications of all optical signal processing for advanced optical modulation formats
PDF
Silicon photonics integrated circuits for analog and digital optical signal processing
PDF
Detection and optical signal processing using phase based optical modulation formats
PDF
Orbital angular momentum based spatially multiplexed optical and millimeter-wave communications
PDF
Applications in optical communications: quantum communication systems and optical nonlinear device
PDF
Nonuniform sampling and digital signal processing for analog-to-digital conversion
PDF
Optical communications, optical ranging, and optical signal processing using tunable and reconfigurable technologies
Asset Metadata
Creator
Mohajerin Ariaei, Amirhossein
(author)
Core Title
Reconfigurable high-speed processing and noise mitigation of optical data
School
Viterbi School of Engineering
Degree
Doctor of Philosophy
Degree Program
Electrical Engineering
Publication Date
07/09/2018
Defense Date
05/04/2018
Publisher
University of Southern California
(original),
University of Southern California. Libraries
(digital)
Tag
data processing,noise mitigation,OAI-PMH Harvest,optical systems,signal processing
Format
application/pdf
(imt)
Language
English
Contributor
Electronically uploaded by the author
(provenance)
Advisor
Willner, Alan (
committee chair
), Haas, Stephan (
committee member
), Wu, Wei (
committee member
)
Creator Email
a.mohajerin.a@gmail.com,mohajera@usc.edu
Permanent Link (DOI)
https://doi.org/10.25549/usctheses-c40-509785
Unique identifier
UC11267255
Identifier
etd-MohajerinA-6377.pdf (filename),usctheses-c40-509785 (legacy record id)
Legacy Identifier
etd-MohajerinA-6377.pdf
Dmrecord
509785
Document Type
Dissertation
Format
application/pdf (imt)
Rights
Mohajerin Ariaei, Amirhossein
Type
texts
Source
University of Southern California
(contributing entity),
University of Southern California Dissertations and Theses
(collection)
Access Conditions
The author retains rights to his/her dissertation, thesis or other graduate work according to U.S. copyright law. Electronic access is being provided by the USC Libraries in agreement with the a...
Repository Name
University of Southern California Digital Library
Repository Location
USC Digital Library, University of Southern California, University Park Campus MC 2810, 3434 South Grand Avenue, 2nd Floor, Los Angeles, California 90089-2810, USA
Tags
data processing
noise mitigation
optical systems
signal processing