Introduction to Optical Transmission Systems Seminar

Transcription

Introduction to Optical Transmission Systems Seminar
Maximizing Network Capacity, Reach and Value
Over land, under sea, worldwide
Xtera Communications, Inc.
Introduction To Optical Transmission Systems
Seminar
19 September 2013
Bertrand Clesca – Head of Global marketing
[email protected]
© 2013 Xtera Communications, Inc. Proprietary & Confidential
1
Xtera and Optical Transmission Systems
•
Xtera is a US-based optical infrastructures company focused on the
delivery of equipment and turnkey solutions for high-capacity, long-haul
optical networks, both terrestrial and subsea.
•
Examples of offering (optical transport perspective):
–
–
–
–
•
15-Tbit/s (150 x 100G) line capacity with ultra-long reach
Unregenerated links on 3,000+ km in terrestrial networks
Unrepeatered links on 500+ km
Ultra efficient and highly-integrated Raman amplification solution over land
and under water
Examples of achievements:
–
–
–
–
3.4 Tbit/s (34 x 100G) unrepeatered on 430 km
22,000-km 100G backbone network (in Mexico)
2,500-km 100G all-optical link (no intermediate regeneration site)
The industry’s first 100G repeatered subsea cable system in commercial
service
© 2013 Xtera Communications, Inc. Proprietary & Confidential
2
Content
•
Optical Transport Basics
•
Wavelength Division Multiplexing
Basics
•
Optical Amplification Basics
•
C and L Transmission Bands and
Optical Amplifier Bandwidth
•
Coherent Detection
•
Beyond 100G Channel Rate
•
Soft-Decision Forward Error
Correction
•
Optical Amplification and Noise
•
Span Length, Fiber Loss and
Optical Noise Accumulation
•
Practical Illustration of How
Extended Power Budget and Low
Loss Delivers More Cost Efficient
Network
•
Differences Between Submarine
and Terrestrial Long-Haul Optical
Networks
•
Summary
© 2013 Xtera Communications, Inc. Proprietary & Confidential
3
Optical Transport
Basics
© 2013 Xtera Communications, Inc. Proprietary & Confidential
4
Optical Fiber Transport
•
Optical fibers have outstanding features enabling long-haul,
high-capacity transport:
– Low attenuation
– Huge bandwidth
– Allow Wavelength Division
Multiplexing (WDM)
•
However:
– Attenuation is not equal to zero due to absorption and scattering.
– Optical pulses are degraded/distorted by:
• Chromatic dispersion
• Polarization Mode Dispersion (PMD)
• Non-linear effects:
– Self-Phase Modulation (SPM)
– Cross-Phase Modulation (XPM)
– Four-Wave Mixing (FWM)
– Stimulated Rayleigh Scattering (SRS)
– Stimulated Brillouin Scattering (SBS)
– Most of these effects are wavelength dependent.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
5
Standard fiber
(ITU-T G.652)
Standard fiber
(ITU-T G.654)
1.0
20
0.8
Typical NZDSF fiber
(ITU-T G.655)
10
0.4
0
Chromatic dispersion (ps/nm/km)
Fiber attenuation (dB/km)
Some Fiber Wavelength-Dependent Effects
0.2
Usual spectral band (i.e. EDFA
C band) of operation for longhaul optical communications
1.2
1.3
1.4
1.5
1.6
1.7
Optical wavelength (µm)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
6
Optical Transmitter and Transport
Optical
transmitter
•
Power
Data IN
Space
Mark
(symbol "0") (symbol "1")
Tbit
Time
Optical
fiber
Optical communications were based till 2010 only on digital binary
signals:
– Series of On/Off power-modulated pulses generated by an optical
transmitter and launched into the optical fiber
– Most common modulation format: NRZ (Non Return to Zero) On/Off keying
•
The optical signal leaves the optical transmitter at full strength but as it
travels many kilometers down the fiber, it falls victim to the absorption,
dispersion, distortion and scattering effects.
 Regeneration or amplification is needed.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
7
Regeneration
•
Regeneration required when the signal has been degraded to the extent
that it is difficult to determine accurately whether any bit is a one or a
zero
•
Regeneration process based on Optical-Electrical-Optical (OEO)
conversion:
– Conversion of the incoming optical signal to an electrical signal (OE
conversion)
– Regeneration of the electrical signal with high-speed electronics
– Conversion back to an optical signal for transmission of a cleaned-up version
of the original optical signal along the next span (EO conversion)
OE
conversion
Electrical
regeneration
EO
conversion
– Per wavelength process: one regenerator per wavelength
© 2013 Xtera Communications, Inc. Proprietary & Confidential
8
Optical Amplification
•
Optical amplification required when the signal is simply an attenuated
version of the original signal
•
Optical amplification process based on the way atoms and molecules in
the fiber react when excited with a high-power optical pump
•
Two basic options:
– Optical amplification occurs in a
specific coil of fiber placed in
intermediate optical repeating sites
(which is the case in Erbium-Doped
Fiber Amplifiers – EDFAs)
Signal in
Signal out
EDF
Transmission
fiber
Pump
Signal in
– Optical amplification occurs
in the transmission fiber itself
(such as in Raman amplifiers)
•
Transmission
fiber
Signal out
Pump
Optical amplifiers can handle multiple wavelengths.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
9
Wavelength Division Multiplexing (WDM)
Basics
© 2013 Xtera Communications, Inc. Proprietary & Confidential
10
Wavelength Division Multiplexing
•
Wavelength Division Multiplexing (WDM): Combining multiple optical
carriers into the same optical fiber for high line capacity
•
Optical multiplexing device with single-wavelength input ports and one
multi-wavelength output port
Terminal equipment to
combine multiple
optical carriers into the
same optical fiber
Wavelength
spacing
Signals to be
transmitted
•
Each wavelength can carry a data signal at different channel rate,
modulation, protocol…
•
The wavelength spacing is typically constant across the spectrum and
set at 0.4 nm / 50 GHz (for reference, the system optical bandwidth
ranges from 35 to 100 nm).
© 2013 Xtera Communications, Inc. Proprietary & Confidential
11
Combination of High Channel Rate and Wavelength
Division Multiplexing for Fiber Capacity Explosion
100,000
Xtera WDM solutions
150 x 100 G
15 Tb/s
10,000
8.8 Tb/s
88 x 100G
240 x 10G
2.4 Tb/s
Fiber Capacity (Gbit/s)
(44 x 400G)
80 x 40G
1000
80 x 10G
800 Gb/s
40 x 10G
100
80 x 2.5G
16 x 10G
100 Gb/s
160 x 2.5G
40 x 2.5G
16 x 2.5G
10
8 x 2.5G
10 Gb/s
40G
10G
Single-wavelength
system
2.5G
1
0.1
1984
565 Mb/s
140 Mb/s
1986
1988 1990
1992
1994
1996
1998
2000
2002
2004
2006
2008
2010
2012
2014
Year
© 2013 Xtera Communications, Inc. Proprietary & Confidential
12
Wavelength Division Multiplexing and Optical
Amplification
•
Wavelength Division Multiplexing (WDM): Combining multiple optical
carriers into the same optical fiber for high line capacity
•
Optical amplification: Amplifying simultaneously all the wavelengths in
the fiber for compensating the fiber attenuation along the optical routes
Terminal equipment to
combine multiple
optical carriers into the
same optical fiber
Optical amplifier to
compensate the fiber
attenuation
Signals to be
transmitted
•
Main values of optical amplifiers:
– Common line equipment shared by all the wavelengths
– Independent on channel rate, modulation, protocol…
© 2013 Xtera Communications, Inc. Proprietary & Confidential
13
Optical Amplification
•
•
Conventional EDFA-based optical amplification technology
High-performance optical amplification technology
Larger
and flatter
bandwidth
Longer reach
© 2013 Xtera Communications, Inc. Proprietary & Confidential
14
C and L Transmission Bands and
Optical Amplifier Bandwidth
© 2013 Xtera Communications, Inc. Proprietary & Confidential
15
Fiber attenuation (dB/km)
C Band
1.0
0.8
0.4
0.2
Usual spectral band (i.e. EDFA
C band) of operation for longhaul optical communications
1.2
1.3
1.4
1.6
1.5
1.7
Optical wavelength (µm)
•
The C-band is the “center” DWDM transmission band, occupying the
1530- to 1562-nm wavelength range.
•
Close to the region where fiber loss is minimal
•
Traditional Erbium-Doped Fiber Amplifiers (EDFAs) operate in C band,
with advertised channel counts varying from 16 to 96 channels.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
16
Fiber attenuation (dB/km)
L Band
1.0
0.8
0.4
0.2
L band
1.2
1.3
1.4
1.5
1.6
1.7
Optical wavelength (µm)
•
The L-band is the “long” DWDM transmission band, occupying the 1570to 1610-nm wavelength range.
•
The wavelength at which fiber loss is minimal usually lies in the L band.
•
EDFA technology can operate in the L-band window but with low
efficiency so this is not a popular option.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
17
EDFA Optical Amplification Bandwidth
•
The intrinsic EDFA gain profile is not uniform across C band
EDFA gain (linear scale)
 Additional gain flattening filters are mandatory (not power efficient as
the extra optical gain is shaved off, and not redistributed in other spectral
regions).
C band
Increasing
pump power
Amplification
0
Attenuation
1450
1500
1550
1600
Wavelength (nm)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
18
Raman Optical Amplification Bandwidth
Optical pump
launched in the
optical fiber
6
Resulting gain in the optical fiber
5
About
100 nm
higher
4
3
2
1
0
0
6
12
18
24
30
36
42
48
Frequency shift (THz)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
19
Raman Optical Amplification Bandwidth
Composite signal optical
bandwidth
 100-nm continuous
optical bandwidth
6
5
4
3
2
1
0
0
6
12
18
24
30
36
42
48
Frequency shift (THz)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
20
Amplifier Technology, Optical Bandwidth and
Channel Count
Band # 4
1519 nm
Band # 3
1543 nm
Band # 2
1568 nm
Standard EDFA C band (36 nm)
Band # 1
1593 nm
1618 nm
Note: Conventional EDFA-based systems
are limited to this C band
Raman amplifier bandwidth (63 nm)
Raman+ amplifier bandwidth (100 nm)
•
Assuming 0.4-nm / 50-GHz channel spacing:
– EDFA amplifier
 36-nm bandwidth  88 channels
– Raman amplifier  63-nm bandwidth  150 channels
– Raman+ amplifier  100-nm bandwidth  240 channels
© 2013 Xtera Communications, Inc. Proprietary & Confidential
21
Coherent Detection
© 2013 Xtera Communications, Inc. Proprietary & Confidential
22
100G
PM-QPSK: How It Works
•
Dual polarization with quadrature phase shift keying modulation
– Spectrally efficient, improved signal-to-noise ratio, CMOS electronics
•
Trade speed for parallelism (at the expense of added complexity)
Polarization Multiplexing
(PM)
Quadrature Phase Shift Keying
(QPSK)
I
01
200-GHz filter
bandwidth
100-GHz filter
bandwidth
1 x 100G
2 x 50G
Two independent polarizations
• Same optical frequency
• Halves the data rate
• Halves the spectral width
• Doubles components
11
Q
00
10
50-GHz filter
bandwidth
4 x 25G
Data encoded into 4-phase states
• Phase symbol – 2 bits of data
• Halves the symbol rate
• Halves the spectral width
• Doubles components
© 2013 Xtera Communications, Inc. Proprietary & Confidential
23
100G
Digital Coherent PM-QPSK Receiver
H-pol: I
Polarization
beam
splitter
Optical
coupler H-pol: Q
network
V-pol: I
and
photodetectors V-pol: Q
Local
oscillator
(Laser)
•
Analog
to
digital
converter
Adaptive
equalizer
Error signal
+
Decision
circuit
Soft-decision
FEC
Adaptive
algorithm
Constellation
plots provided
by EXFO
Optical frequency control
Powerful Digital Signal Processor (DSP) correcting linear impairments:
– Chromatic dispersion:
• Today: up to ±70,000 ps/nm (about 3,500 km of G.652 fiber)
• Future: up to ±250,000 ps/nm (OK for transoceanic cable systems)
– Polarization Mode Dispersion (PMD): up to 30 ps
– Higher performance/robustness at 100G with coherent detection than at 10G
with direct detection!
© 2013 Xtera Communications, Inc. Proprietary & Confidential
24
100G
A Few Words on Chromatic Dispersion Compensation
•
100G digital coherent receiver performs the compensation of chromatic
dispersion with excellent performance.
•
However, for not too demanding links with 10G channel rate and direct
detection, the spread of the light pulse caused by chromatic dispersion
imposes the use of Dispersion Compensating Module (DCM).
Amplitude
Note: Typical tolerance to chromatic dispersion for a 10G NRZ signal: ±800 ps/nm (≈20 km of G.652 fiber)
lsignal  1550 nm
G.652 line
fiber
Amplitude
Anomalous dispersion:
vblue > vred
Blue
0
Red
50
100
150
200
Time (ps)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
Signal spectrum
bandwidth depends on:
• Bit rate
• Coding
• Modulation
25
100G
A Few Notes on Interoperability and Standards
•
Interoperability:
– There is no “absolute” 100G standard at line level: WDM node from
Vendor A is not expected to interoperate with WDM node from Vendor B.
– The interface cards (that deliver 100G channel rate to the line fiber) shall
interoperate with client equipment via standard client interface (e.g. STM-64
or 10GE).
WDM node
WDM node
from Vendor A
Client
interfaces
•
from Vendor B
Client
interfaces
An implementation agreement (OIF) exists, however, defining a single
set of technologies for building 100G line interfaces:
– Avoids the multiplication of 100G modulation formats (not repeating 40G
mistakes)
– Facilitates multi-source supply
– Leads to higher volume manufacturing
© 2013 Xtera Communications, Inc. Proprietary & Confidential
26
Beyond 100G Channel Rate
© 2013 Xtera Communications, Inc. Proprietary & Confidential
27
Options for Increasing Line Capacity
1. Increasing spectral efficiency (bit/s/Hz) at the line interface card level
I
I
Bits per symbol
(Increase
constellation size)
Soft-decision FEC
(Increase
coding gain)
01
11
00
10
Q
1101
1001
0001
0101
1100
1000
0000
0001
1110
1010
0010
0110
1111
1011
0011
0111
Q
QPSK
16-QAM
Symbols per
second
(Increase
symbol rate)
Super channel
(Group of denselypacked waves)
Higher cost and/or
complexity on a per
wavelength basis
© 2013 Xtera Communications, Inc. Proprietary & Confidential
28
Options for Increasing Line Capacity
1. Increasing spectral efficiency (bit/s/Hz) at the line interface card level
I
I
Bits per symbol
(Increase
constellation size)
Soft-decision FEC
(Increase
coding gain)
01
11
00
10
Q
1101
1001
0001
0101
1100
1000
0000
0001
1110
1010
0010
0110
1111
1011
0011
0111
Q
QPSK
16-QAM
Symbols per
second
(Increase
symbol rate)
Super channel
(Group of denselypacked waves)
Moderate increase in
the capacity or strong
reach limitation
© 2013 Xtera Communications, Inc. Proprietary & Confidential
29
Options for Increasing Line Capacity
2. Improved common equipment (line fiber, optical amplifier…)
I
I
Bits per symbol
(Increase
constellation size)
Soft-decision FEC
(Increase
coding gain)
01
11
00
10
Q
1101
1001
0001
0101
1100
1000
0000
0001
1110
1010
0010
0110
1111
1011
0011
0111
Q
QPSK
16-QAM
Symbols per
second
(Increase
symbol rate)
Super channel
(Group of denselypacked waves)
Common line equipment
• Broader optical amplifier bandwidth
• Distributed amplification in the line fiber
to lower the amount of nonlinear effects
• New fiber types
Extra cost shared
by all the wavelengths
© 2013 Xtera Communications, Inc. Proprietary & Confidential
30
Today’s Status
•
100G signal rate:
– On the client side: typically 10 x 10GE, 10 x STM-64 or 1 x 100GE
– On the line side: 100G signal mapped into an OTU4 frame
(Optical Transport Network – OTN – ITU-T G.709 standard) with PM-QPSK
modulation format
Client signals
Client signals
(N x 10G,
1 x 100G)
100G channel
rate
Optical network
Client signals
© 2013 Xtera Communications, Inc. Proprietary & Confidential
31
Next Step(s)
•
On the client side: 400GE is currently being standardized.
•
On the line side:
– The standardization of the OTN frame of the 400G line will wait for the
finalization of the client 400GE standard.
– The technical implementation (e.g. which modulation format) is not yet
defined.
Client signals
400G client
signals
Channel
rate: could
be 400G
Optical network
Client signals
© 2013 Xtera Communications, Inc. Proprietary & Confidential
32
Brief Status Report on Beyond 100G Project
•
IEEE formed a working committee on 400GE interface in March 2013
and had the kickoff meeting in Victoria, BC in May 2013.
•
The committee met in York, UK in September 2013 to start work on
defining the 400GE interface. It is expected that it will take many more
meetings prior to the completion of work.
•
The ITU-T SG15/Q11 is also working on the Beyond 100G standards on
the OTN side.
•
The timing of making a decision on standardization of B100G was also
discussed. It was generally agreed that waiting on the IEEE 400GE
design to define a client side OTN interface for interoperability will be
the right thing to do.
•
In general, it is estimated that the standards for B100G will be done in
the 2015 time frame or later.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
33
ITU-T Standards
ITU-T
Study Group 15 (SG15)
Optical channel
grid
G.694.1
G.694.2
Safety
G.644
Single channel
G.957
G.691
G.693
G.959.1
G.698.2
GFP
G.7041
Fiber
G.651
G.652
G.653
G.655
© 2013 Xtera Communications, Inc. Proprietary & Confidential
SDH
G.707
OTN
G.709
34
Soft-Decision Forward Error Correction
© 2013 Xtera Communications, Inc. Proprietary & Confidential
35
Forward Error Correction (FEC)
Definition: Technique used for controlling errors in data transmission
over unreliable or noisy communication channels based on the sender
encoding the message in a redundant way by using an Error-Correcting
Code (ECC).
•
The redundancy allows the receiver to detect a limited number of errors
that may occur anywhere in the message, and to correct these errors
without retransmission.
Output BER
•
10-1
Uncorrected
10-3
10-5
10-7
Corrected
10-9
10-11
10-13
10-5
10-4
10-3
10-2
10-1
Input Bit Error Rate (BER)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
36
Coding Gain and Soft-Decision FEC
•
The advantage of using FEC is that the probability of an error remaining
in the decoded data is lower than the probability of an error if an FEC
algorithm, such as Reed-Solomon, is not used. This is coding gain in
essence.
•
Coding Gain is difference in Input SNR for a given Output BER.
•
Shortly after the advent of digital coherent receivers, a new type of
forward error correction method was introduced: Soft-Decision Forward
Error Correction (SD-FEC).
Net coding gain (dB)
– This was enabled by higher-speed analog-to-digital converters and digital
signal processors.
13
12
– Using appropriate coding and decoding,
11
it is possible in theory to produce coding
10
9
gains 3 dB better than the “standard”
8
hard-decision FEC approach.
7
Theoretical limit using 6 bit softdecision decoding
Hard-decision
(binary)
theoretical limit
6
5
5
10
15
20
25
30
FEC overhead (%)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
37
Optical Amplification and Noise
© 2013 Xtera Communications, Inc. Proprietary & Confidential
38
Generation of Optical Noise Within
An Erbium-Doped Fiber
Optical signal
Optical pump
Erbium ions
Amplified Spontaneous Emission (ASE)
Unguided spontaneously-emitted photon
Guided
spontaneous
emission
Optical gain experienced by the signal
Amplified
Spontaneous
Emission (ASE)
•
Optical amplification is achieved at the expense of optical noise generation.
•
Raman amplifiers also generate optical noise but in a lower amount.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
39
Generation of Optical Noise From
An Erbium-Doped Fiber
Single wavelength amplified by an old-generation EDFA (unflatenned)
Power (dBm/0.25 nm)
•
20
0
Amplified
signal
-20
-40
-60
1520
Optical Signal-to-Noise
Ratio (OSNR)
Amplified
Spontaneous
Emission (ASE)
1540
1560
1580
Wavelength (nm)
•
ASE is wideband, non-coherent, un-polarized light which will be detected
as noise by the receive interface.
•
Optical Signal-to-Noise Ratio (OSNR, in dB / 0.1 nm) – ratio between the
signal power and the ASE noise power in a given optical bandwidth – is
an important parameter for engineering optical links as any combination
of modulation format and receiver technology imposes a minimal OSNR
figure for properly detecting the data.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
40
Generation of Optical Noise From
A Raman Amplifiers String
•
Multiple wavelengths at the input and output of a Raman amplifiers
string
Input spectrum
Output spectrum
Input Spectrum
-20
Output Spectrum
-20
Spectrum- Input Into Line
Spectrum- Ouput of Line
-25
-25
Power (dBm)
Power (dBm)
-30
-35
-40
-45
-30
-35
-50
-55
-40
-60
1541
•
1546
1551
1556
Wavelength (nm)
1561
1541
1546
1551
1556
Wavelength (nm)
1561
At the receive end the link, the system optimization parameter was the
uniformization of the OSNR characteristic for all the wavelengths.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
41
Span Length, Fiber Loss and
Optical Noise Accumulation
© 2013 Xtera Communications, Inc. Proprietary & Confidential
42
Optical Noise Accumulation
Optical noise
Optical noise
Optical noise
•
Longer transmission distances and longer spans decrease the
Optical Signal-to-Noise Ratio (OSNR) figure at the output end.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
43
Per channel power profile
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Non-linear limitation
Too large per channel power:
• Nonlinear effects within the line fiber
• Overload of the receiver circuitry
Amplifier
site
Amplifier
site
Noise limitation
Distance
Too small per channel power:
• Below the minimum OSNR
requirement
• Below the receiver sensitivity
© 2013 Xtera Communications, Inc. Proprietary & Confidential
44
Digital Pulse Patterns
And Eye Diagrams (For NRZ Signals)
•
Eye diagram: oscilloscope display in which a digital data signal from a
receiver is repetitively sampled and applied to the vertical input, while
the data rate (clock signal) is used to trigger the horizontal sweep.
Synchronization by
the 10-Gbit/s clock signal:
 Eye diagram
Amplitude (a.u.)
Amplitude (a.u.)
Synchronization by
the 10-Gbit/s data frame:
 Pulse pattern
Tbit
0
200
800
1200
1600
2000
Eye
opening
Time (ps)
0
100
Tbit
200
300
Time (ps)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
45
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Input eye diagram
(NRZ signal)

Non-linear limitation
Amplifier
power
boost
Output eye diagram
Fiber
attenuation
Optical
amplifier noise
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
46
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Input eye diagram
(NRZ signal)

Non-linear limitation
Amplifier
power
boost
Output eye diagram
Fiber
attenuation
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
Pulse
compression
(or distorsion)
47
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Input eye diagram
(NRZ signal)

Non-linear limitation
Amplifier
power
boost
Output eye diagram
Fiber
attenuation
Optical
amplifier noise
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
Pulse
compression
(or distorsion)
48
Per channel power profile
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management

Non-linear limitation
Amplifier
power
boost
Fiber
attenuation
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
49
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Long span

Non-linear
Non-linear
limitation
limitation
Nonlinear issue!
Noise
Noise limitation
limitation
OSNR issue!
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
50
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
•
Options to overcome the long span:
– Build an intermediate site?
Issues with
1.
2.
3.
CapEx and OpEx
Permitting
Lead time
– Terminate the long span at either end with back-to-back terminals?
Issues with
1.
2.
–
CapEx and OpEx
Incremental cost
Use lower-loss line fiber
© 2013 Xtera Communications, Inc. Proprietary & Confidential
51
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Long span

Non-linear
Non-linear
limitation
limitation
Lower-loss fiber
Noise
Noise limitation
limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
52
Optical Noise Accumulation
•
Going from 0.22- to 0.18-dB/km loss on the long span:
80 km
0.22 dB/km
22 dB
90 km
0.22 dB/km
24 dB
150 km
0.22 dB/km
37 dB
70 km
0.22 dB/km
19 dB
80 km
0.22 dB/km
22 dB
Optical noise
80 km
0.22 dB/km
22 dB
90 km
0.22 dB/km
24 dB
150 km
0.18 dB/km
31 dB
70 km
0.22 dB/km
19 dB
80 km
0.22 dB/km
22 dB
Optical noise
 Immediate increase in the output Optical Signal-to-Noise Ratio (OSNR)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
53
Optical Noise Accumulation
•
Decreasing the fiber loss in another span will also increase the output
OSNR:
80 km
0.22 dB/km
22 dB
90 km
0.22 dB/km
24 dB
150 km
0.18 dB/km
31 dB
70 km
0.22 dB/km
19 dB
80 km
0.22 dB/km
22 dB
Optical noise
80 km
0.22 dB/km
22 dB
90 km
0.18 dB/km
20 dB
150 km
0.18 dB/km
31 dB
70 km
0.22 dB/km
19 dB
80 km
0.22 dB/km
22 dB
Optical noise
 Along an all-optical path, the loss of all the spans matters for
the output OSNR.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
54
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
•
Options to overcome the long span:
– Build an intermediate site?
Issues with
1.
2.
3.
CapEx and OpEx
Permitting
Lead time
– Terminate the long span at either end with back-to-back terminals?
Issues with
1.
2.
–
–
CapEx and OpEx
Incremental cost
Use lower-loss line fiber
Create optical gain within the line fiber to avoid nonlinear/OSNR limitations
and extend span performance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
55
Per channel power profile
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management

Non-linear limitation
Optical gain created
within the line fiber by
distributed Raman
amplification via
backward pumping
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
56
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Longer span

Nonlinear
issue!
Non-linear limitation
OSNR issue!
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
57
Optical Signal-to-Noise Ratio (OSNR)
And Per Channel Power Management
Per channel power profile
Longer span

Non-linear limitation
Forward Raman pumping
Backward
Raman
pumping
Noise limitation
Distance
© 2013 Xtera Communications, Inc. Proprietary & Confidential
58
Practical Illustration of How Extended
Power Budget and Low Loss Delivers
More Cost Efficient Network
© 2013 Xtera Communications, Inc. Proprietary & Confidential
59
All-Optical 1,350-km Link with Long Spans
45 km
76 km
ROADM
ILA
135 km
ROADM
250 km
ROADM
Core
Amplifier
ROADM
65 km
ROADM
170 km
ROADM
37 km
ILA
147 km
55 km
ILA
100 km
ILA
Backward span
Extension module
ROADM
ILA
137 km
97 km
31 km
ROADM
Forward span
Extension module
© 2013 Xtera Communications, Inc. Proprietary & Confidential
ROADM
Line
fiber
60
In the Absence of Low-Loss Fiber…
45 km
76 km
ROADM
135 km
ILA
ROADM
ROADM
125
km
ROADM
37 km
ILA
125 ROADM
km
97 km
ROADM
Cut the long span in two
250 km
ROADM
ROADM
170 km
ROADM
Core
Amplifier
65 km
ILA
Backward span
Extension module
55 km
ILA
Terminate ILA
the signals at each
end
of the long span (with back-toback terminal equipment)
137 km
Regeneration site
147 km
100 km
ILA
250 kmROADM
Forward span
Extension module
© 2013 Xtera Communications, Inc. Proprietary & Confidential
31 km
Regeneration site
ROADM
Line
fiber
61
Summary
© 2013 Xtera Communications, Inc. Proprietary & Confidential
62
How to Cope With Fiber Impairments?
•
It may be a digital signal, but it’s analog transmission!
Multiplexing
site
Multiplexing
site
Attenuation
Dispersion
Nonlinearity
Distortion
Transmitted data waveform
© 2013 Xtera Communications, Inc. Proprietary & Confidential
Waveform after 1,000 km
63
Numerous Optical Impairments
Optical
impairments
experienced by
DWDM signals
Linear effects
Nonlinear effects
Kerr
effects
Attenuation
(Insertion loss)
Self-Phase
Modulation
(SPM)
Chromatic
dispersion
Polarization
Mode
Dispersion
(PMD)
Cross-Phase
Modulation
(XPM)
Four-Wave
Mixing
(FWM)
© 2013 Xtera Communications, Inc. Proprietary & Confidential
Scattering
effects
Stimulated
Raman
Scattering
(SRS)
Stimulated
Brillouin
Scattering
(SBS)
64
The Art of Span Engineering
Span
engineering
Linear impairments
management
Forward Error
Correction
(FEC)
Loss
management
Chromatic
Dispersion
management
Nonlinear impairments
management
Per channel
power
management
PMD
management
Stimulated
Brillouin
Scattering
(SBS)
suppression
System
transmission
quality
© 2013 Xtera Communications, Inc. Proprietary & Confidential
65
Span Engineer’s Best Friend
•
The best friend of an optical span engineer / optical network designer is
a high-performance line fiber with:
– Low loss
– High dispersion
– Large effective area
•
Such characteristics are very likely to be future proof with respect to
new modulation formats and channel rates.
© 2013 Xtera Communications, Inc. Proprietary & Confidential
66
Maximizing Network Capacity, Reach and Value
Over land, under sea, worldwide
© 2013 Xtera Communications, Inc. Proprietary & Confidential
67