Optical SignaltoNoise Ratio Monitoring Using Beat Noise

1
Optical Signal-to-Noise Ratio Monitoring Using
Beat Noise

• Vivian, Chen Wei
• Supervisors Rod Tucker and Jamie Evans

2
Outline

• Overview of optical performance monitoring
• Introduction to optical noise and optical
  signal-to-noise ratio (OSNR) monitoring
• Theory of beat noise
• OSNR monitoring by measuring low frequency noise
  (Shin00)
• OSNR monitoring by analyzing the
  spontaneous-spontaneous beat noise
• OSNR monitoring by analyzing the uncorrelated
  signal-spontaneous beat noise
• Conclusions and research plan

3
Optical Impairments

• Noise amplifier noise, receiver noise, laser
  noise
• Distortion dispersion, filtering, non-linear
  effects

fibre
optical amplifier
Degrade the performance of optical link.
4
Optical Network Management
Optical Network Management
Fault Management --- localizing, diagnosing
Maintenance
Protection and Restoration
Require Optical Performance Monitoring
5
Implement OPM in Optical Link
Optical Cross-Connect
Fibre Link
Coupler
Optical Tuneable Filter
Optical Performance Monitor

• Perform non-disruptive monitoring.

performance cost
6
Different Aspects of OPM
Channel ID monitoring
Power monitoring
Optical performance monitoring
Dispersion monitoring
Polarization monitoring
Bit error rate (BER) monitoring
Optical Signal-to-Noise Ratio (OSNR) monitoring
7
What is optical noise?
Data
Transmitter
Fibre
Data
Receiver
Optical Amplifier
Optical Link
8
Generation of ASE noise in EDFA
spontaneous emitted photons
Fibre core
Er3
Cladding
9
Definition of OSNR
optical spectrum for one channel
optical power
Psignal
PASE
ASE noise
?
wavelength
??
OSNRPsignal / PASE
OSNR dBPsignaldBm-PASEdBm
?? reference optical bandwidth (usually 0.1nm)
10
Current OSNR monitoring Techniques
out-of-band noise measurement
optical carrier
optical power
modulated signal
ASE noise
?
wavelength
modulation bandwidth

• Optical process
• Traditional optical spectrum measurement
• Linearly interpolation method Suzuki99

11
Current OSNR monitoring Techniques (Cont.)
in-band noise measurement
optical carrier
optical power
modulated signal
ASE noise
?
wavelength
modulation bandwidth

• Optical process
• Polarization nulling method Wiech98Lee01
• Electrical process
• Orthogonal delayed-homodyne method Youn02
• Half clock frequency constellation method
  Stuart03
• Low frequency beat noise method (Shin00)

12
Theory of Beat Noise
2
signalASE
R signalASE2
optical Filter (Bo)
photodiode
R signalASE2R signal2 2signalASE
ASE2
photodiode responsivity
13
Beat Noise Power Density (Olsson 1989)
2
amplified signal (polarized) ASE (unpolarized)
R signalASE2
optical Filter (Bo)
photodiode
Psig optical signal power NASE ASE noise
power density in one polarization
direction Nsig-sp sig-sp beat noise power
density Nsp-sp sp-sp beat noise power
density R photodiode responsivity Bo optical
filter bandwidth RBW reference bandwidth for
OSNR
electrical spectrum
optical spectrum
Nsig-sp4R2PsigNASE , 0, Bo/2
Psig
2NASE
Nsp-sp(f)4(RNASE)2(Bo-f), 0, Bo
0
fc
fcBo/2
fc-Bo/2
Bo/2
Bo
f
frequency
frequency
OSNRPsig/(2NASERBW) Nsig-sp2R2P2sig/(RBWOSNR)
, 0, Bo/2 Nsp-sp(f)R2P2sig(Bo-f)/(RBWOSNR)2,
0, Bo PtotalPsigPASEPsig1Bo/(RBWOSNR)
14
Low Frequency Beat Noise Method (Shin00)-Concept
We can observe noise in the low frequency range
of the electrical spectrum, if there is no
signal components.
15
Low Frequency Beat Noise Method (Shin00)-PRBS
spectrum

• PRBS electrical spectrum

PRBS Pseudo-Random Binary Sequence fb bit
rate Pattern length 2N-1
16
Low Frequency Beat Noise Method (Shin00)-Schematic
modulator
laser
attenuator
11 Coupler
ASE source (AWGN)
LPF, A/D, DSP
optical filter
attenuator
photodiode
VPI simulation schematic of the low frequency
beat noise method
17
Low Frequency Beat Noise Method
(Shin00)-Simulation
OSNR Estimation Error vs. True OSNR
Difference of Estimated OSNR and True OSNR (dB)
True OSNR (dB)
18
Low Frequency Beat Noise Method
(Shin00)-Discussion
Discussion May not be applicable for very
random data or data modulated by long pattern
length sequence. (for example 10Gb/s modulation
by 231-1 PRBS fc 5Hz)
19
sp-sp Beat Noise Method - Concept
electrical power spectrum density (polarized
signal and unpolarized ASE)
signal-spontaneous
spontaneous-spontaneous
2R2P2sig/(RBWOSNR)
R2P2sigBo/(RBWOSNR)2
Bo/2
Bo
Bo optical bandwidth RBW reference bandwidth
for OSNR
Sig-sp beat noise power density will not change
with Bo Whereas, sp-sp beat noise power density
varies with the change of Bo.
Nbeat2-Nbeat1R2P2sig(Bo2-Bo1)/(RBWOSNR)2
20
sp-sp Beat Noise Method - Schematic
responsivity1A/W turn off shot noise thermal
noise
Bo140GHz fo1fc
bit rate10Gb/s Pattern length215-1
LPF, A/D, DSP (subtraction the data from
two photodiodes)
optical filter
modulator
photodiode
laser
attenuator
11 Coupler
ASE source (AWGN)
optical filter
attenuator
photodiode
Bo280GHz fo2fc
measurement frequency 50kHz
500kHz resolution bandwidth 19kHz
VPI simulation schematic of sp-sp beat noise
method
21
sp-sp Beat Noise Method Simulation Results
OSNR Estimation Error vs. True OSNR
Difference of Estimated OSNR and True OSNR (dB)
True OSNR (dB)
22
sp-sp Beat Noise Method Analysis
zoomed electrical spectrum (OSNR20dB)
Power spectral density dBm/19kHz
Power spectral density dBm/19kHz
-33
-33
(Bo140GHz)
(Bo280GHz)

-80

-80
-120
-120








0 100 200 300
400 500 600
0 100 200 300
400 500 600
Frequency (kHz)
Frequency (kHz)
23
sp-sp Beat Noise Method Discussion
How to remove the signal components? Maybe it is
better to use sig-sp beat noise?
24
Uncorrelated sig-sp Beat Noise Method Concept
correlated
two optical filters position in optical spectrum
25
Uncorrelated sig-sp Beat Noise Method Schematic
Bo1Bo fo1fcfb
RF Down Converter, A/D, DSP (subtraction the
data from two photodiodes)
optical filter
modulator
photodiode
laser
attenuator
11 Coupler
ASE source (AWGN)
optical filter
attenuator
photodiode
Bo2Bo fo2fc-fb
VPI simulation schematic of uncorrelated sig-sp
beat noise method
26
Uncorrelated sig-sp Beat Noise Method RF
Spectrum
sig-sp beat noise electrical spectrum
27
Uncorrelated sig-sp Beat Noise Method
Simulation Setting
responsivity1A/W turn off shot noise thermal
noise
Bo140GHz fo1fc10GHz
bit rate10Gb/s Pattern length230-1
RF Down Converter, A/D, DSP (subtraction the
data from two photodiodes)
optical filter
modulator
photodiode
laser
attenuator
11 Coupler
ASE source (AWGN)
optical filter
attenuator
photodiode
Bo240GHz fo2fc-10GHz
VPI simulation schematic of uncorrelated sig-sp
beat noise method
28
Uncorrelated sig-sp Beat Noise Method OSA
optical spectrum after two optical filters
(OSNR20dB)
Power spectral density dBm/1.22MHz
Power spectral density dBm/1.22MHz
-11 -20 -40 -60 -80 -87
-11 -20 -40 -60 -80 -87
-80 -40-30-20-10 0 10 20 30 40
80
-80 -40-30-20-10 0 10 20 30 40
80
optical frequency relative to 193.1THz GHz
optical frequency relative to 193.1THz GHz
29
Uncorrelated sig-sp Beat Noise Method RFSA
electrical spectrum for the subtracted data
(OSNR20dB)
Power spectral density dBm/1.22MHz
-80 -100 -120 -140 -160 -
167
0 10 20 30
40
80
RF frequency GHz
30
Uncorrelated sig-sp Beat Noise Method Simulation
OSNR Estimation Error vs. True OSNR
(measurement frequency 15GHz to 15.24GHz )
Difference of Estimated OSNR and True OSNR (dB)
True OSNR (dB)
31
Uncorrelated sig-sp Beat Noise Method Discussion

• Independent on pattern length.
• Simulation results show that OSNR estimation is
  accurate.
• How to shift the observation frequency lower in
  electrical domain?

32
Uncorrelated sig-sp Beat Noise Method Cont.
optical spectrum
electrical spectrum
Electrical spectrum
BPF2
BPF1
correlated
sig-sp beat noise power density
fb
3fb
Frequency
Decrease the overlap (correlated part) in optical
spectrum Lower the start frequency of
uncorrelated sig-sp beat noise in electrical
spectrum.
33
Uncorrelated sig-sp Beat Noise Method Cont.
Power spectral density dBm/1.22MHz
-5 -20 -40 -60 -80 -92
Power spectral density dBm/1.22MHz
-80 -100 -120 -140
-80 -40-30-20-10 0 10 20 30 40
80
-5 -20 -40 -60 -80 -92
0 10 20 30 40
80
RF frequency GHz
electrical spectrum
-80 -40-30-20-10 0 10 20 30 40
80
optical frequency relative to 193.1THz GHz
optical spectrum
34
Conclusions

• Optical performance monitoring is essential for
  managing high capacity WDM optical networks.
• Optical Signal-to-Noise Ratio is one important
  quantity to monitor.
• Several techniques for OSNR monitoring have been
  developed by optical measurement or electronic
  measurement.
• The uncorrelated signal-spontaneous beat noise
  method we proposed effectively removes the
  signal components in electrical field and
  calculates OSNR by beat noise.

35
Research Plan

• Continue working on OSNR monitoring by analyzing
  beat noise properties.
• Experimentally verify the methods we proposed.
• Investigate more techniques for OSNR monitoring.
• Design low cost OSNR monitors.
• Discover ways to improve the accuracy and dynamic
  range of OSNR monitors.
• Develop multi-functional monitors, which can
  simultaneously monitor OSNR, dispersion,
  wavelength, power, etc.

36
Thank you!