OPTICAL NETWORKING BEYOND WDM[Nakazawa et al., OFC’10] PDM 64-QAM 21 GBaud (256 Gb/s) [Gnauck et...

COPYRIGHT © 2011 ALCATEL-LUCENT. ALL RIGHTS RESERVED.


OPTICAL NETWORKING BEYOND WDM SPATIAL MULTIPLEXING AND MIMO FOR THE PETABIT ERA Peter J. Winzer Optical Transmission Systems and Networks Research Bell Labs, Alcatel-Lucent, USA


ACKNOWLEDGMENT

Bell Labs S.Chandrasekhar A.R.Chraplyvy C.R.Doerr R.-J.Essiambre G.J.Foschini A.H.Gnauck H.Kogelnik G.Kramer X.Liu S.Randel G.Raybon R.Ryf R.W.Tkach

Optical Fiber Solutions D.DiGiovanni J.Fini R.Lingle D.Peckham T.F.Taunay B.Zhu

Sumitomo Electric M.Hirano Y.Yamamoto T.Sasaki

AND MANY OTHERS


1 Traffic evolution in data networks


HUMAN-DRIVEN TRAFFIC GROWTH

HiDef Video Communication Panasonic’s LifeWall

Human desire to communicate in a multi-media manner Huge transport capacities (especially for non-cacheable real time apps)

1995 2000 2005 2010

2 dB / year (58%/year)

10 Gb/s

100 Gb/s

1 Tb/s

10 Tb/s

US data network traffic

100 Tb/s

60% 10 log10(1.6) dB 2 dB

[R.W.Tkach, BLTJ, 2010]


MACHINE-DRIVEN TRAFFIC GROWTH Amdahl’s rule of thumb 1 Floating point operation (Flop) triggers ~1 Byte/s of transport Cloud services couple network traffic to exponential growth of processor power

100 TFlops

10 TFlops

1 TFlops

100 GFlops

10 GFlops

1 GFlops

1995 2000 2005 2010

2.7 dB / year (86%/year)

2 dB / year (58%/year)

10 Gb/s

100 Gb/s

1 Tb/s

10 Tb/s

US data network traffic

Top 500 Supercomputers

100 Tb/s

60% 10 log10(1.6) dB 2 dB

http://www.circuitboards1.com


2 The role of optical transmission technologies


OPTICAL NETWORKS: WORKHORSE OF THE INTERNET

Line interface (e.g., 100 Gbit/s)

Router

Optical network

WDM: Wavelength-division multiplexing

Reconfigurable optical add/ drop multiplexer (ROADM)

Client interface (e.g., 4 x 25 Gbit/s)

WDM system (e.g., 80 x 100 Gbit/s)

Objectives: Increase per-wavelength interface rates (client and line side) Increase per-fiber aggregate capacities (line side)


HIGH-SPEED OPTICAL INTERFACE EVOLUTION

1986 1990 1994 1998 2002 2006

10

100

1

10

100 Se

rial I

nter

face

Rat

es a

nd W

DM

Cap

aciti

es

Gb/

s Tb

/s

2010 2014 2018 1

2022

• From direct (envelope) detection to coherent (field) detection • Polarization multiplexed 16-QAM at 448 Gb/s demonstrated

Direct detection

9.3 ps

107 Gbaud On/Off

Coherent detection

80 Gbaud QPSK

56 Gbaud 16-QAM


• From direct (envelope) detection to coherent (field) detection • Polarization multiplexed 16-QAM at 448 Gb/s demonstrated • 112-Gbit/s coherent interfaces commercially available since June 2010

HIGH-SPEED OPTICAL INTERFACES – PRODUCTS

1986 1990 1994 1998 2002 2006

10

100

1

10

100 Se

rial I

nter

face

Rat

es a

nd W

DM

Cap

aciti

es

Gb/

s Tb

/s

2010 2014 2018 1

2022

ASIC: 70M+ gates

Direct detection

Coherent detection


SCALING INTERFACES TO TERABIT ETHERNET

PDM 512-QAM 3 GBaud (54 Gb/s) [Okamoto et al., ECOC’10]

PDM 256-QAM 4 GBaud (64 Gb/s) [Nakazawa et al., OFC’10]

PDM 64-QAM 21 GBaud (256 Gb/s) [Gnauck et al., OFC’11]

PDM 16-QAM 56 GBaud (448 Gb/s) [Winzer et al., ECOC’10]

PDM 32-QAM 9 GBaud (90 Gb/s) [Zhou et al., OFC’11]

60 GHz

448 Gb/s (10 subcarriers) 16-QAM 5 bit/s/Hz 2000 km transm. [Liu et al., OFC’10]

65 GHz

606 Gb/s (10 subcarriers) 32-QAM 7 bit/s/Hz 2000 km transm. [Liu et al., ECOC’10]

1.2 Tb/s (24 subcarriers) QPSK 3 bit/s/Hz 7200 km transm. [Chandrasekhar et al., ECOC’09]

300 GHz

Bandw idth, A/ D resolution Optical parallelism

A/ D resolution Bandw idth


THE SCALING OF WAVELENGTH-DIVISION MULTIPLEXING

1986 1990 1994 1998 2002 2006

10

100

1

10

100

Seri

al Inte

rfac

e R

ates

and W

DM

Cap

acit

ies

Gb/s

Tb/s

2010 2014 2018

1

2022

•2.5 dB/yr

•0.8 dB/yr

•0.5 dB/yr

~10 Terabit/s WDM systems are now commercially available ~100 Terabit/s WDM systems have been demonstrated in research Growth of WDM system capacities has noticeably slowed down since ~2000

[P.J.Winzer, IEEE Comm. Mag., June 2010]


3 Fiber nonlinearities and the “nonlinear Shannon limit”


WHY IS AN OPTICAL FIBER NONLINEAR ?

• Core diameter ~8 mm • Cladding diameter ~125 mm • Light is kept within the core (total internal reflection)

Cladding (n1)

Core (n2 > n1)

Megawatt / cm2 optical intensities n2 = n2,0+n2,2 Popt + … (Kerr effect) Leads to nonlinear distortions over hundreds of kilometers


PHYSICAL PHENOMENA AT PLAY

… Optical field propagating in the fiber’s transverse mode

(Nonlinear Schrödinger Equation) + N

ROADM … Reconfigurable optical add/drop multiplexer

Fiber Nonlinearity

Filtering Effects

Chromatic Dispersion Optical Filtering (ROADMs)

Spontaneous emission from in-line optical amplifiers

Noise


THE NONLINEAR SHANNON LIMIT Increasing the signal power (i.e. the SNR) creates signal distortions from fiber

nonlinearity, eventually limiting system performance

SNR [dB]

Cap

acity

C [b

its/s

] Maximum capacity

Signal launch power [dBm]

Tx Rx

Distributed Noise

C = B log2 (1 + SNR)

Nonlinear distortions

Quantum mechanics dictates a lower bound on amplifier noise [R.-J. Essiambre et al., Phys. Rev. Lett. (2008) or J. Lightwave Technol. (2010)]


AN LOWER BOUND ESTIMATE FOR THE SHANNON LIMIT • Assume ring constellations • Deterministic signal back-propagation to remove (most of the) channel memory • Numerical solution of nonlinear Schrödinger equation Numerical statistics

SNR [dB]

Cap

acity

C [b

its/s

] Maximum capacity

Signal launch power [dBm]

Tx Rx

Distributed Noise

[R.-J. Essiambre et al., Phys. Rev. Lett. (2008) or J. Lightwave Technol. (2010)]

0 5 10 15 20 25 30 35 40 SNR (dB)

Capa

city

per

uni

t ba

ndw

idth

(b

its/s

/Hz)

0

1

2

3

4

5

6

8

7 1 ring 2 rings 4 rings 8 rings

16 rings

Numerical statistics


0 5 10 15 20 25 30 35 40 SNR (dB)

Capa

city

per

uni

t ba

ndw

idth

(b

its/s

/Hz)

0

1

2

3

4

5

6

8

7 1 ring 2 rings 4 rings 8 rings 16 rings -0.2 -0.1 0 0.1 0.2

-0.2

-0.1

0

0.1

0.2

Imag

par

t of

fie

ld [

mW

1/2 ]

Real part of field [ mW 1/2 ]

-1 -0.5 0 0.5 1

-1

-0.5

0

0.5

1


Imag

par

t of

fie

ld [

mW

1/2 ]

-2 -1 0 1 2

-2

-1

0

1

2


Imag

par

t of

fie

ld [

mW

1/2 ]

SOME EXAMPLE RESULTS

R.-J. Essiambre et al., Phys. Rev. Lett. (2008) or J. Lightwave Technol. (2010)

Note: Capacity maximum occurs at fairly high SNRs


REALITY CHECK: WHERE ARE WE EXPERIMENTALLY ?

5

10

15

2 100 1,000 10,000

Transmission distance [km]

Spec

tral

eff

icie

ncy

[b/s

/Hz]

~2x

Current WDM products

Commercial WDM needs, ca. 2016

~2 dB / year

WDM: Wavelength division multiplexing NL: Nonlinear


4 Spatial multiplexing: The next frontier


DIMENSIONS FOR MODULATION AND MULTIPLEXING

http://www.occfiber.com/

Space

but space

Polarization Frequency

Time Quadrature

Current WDM products use all dimensions

Physical dimensions

WDM: Wavelength division multiplexing


TWO OPTIONS TO REACH OUR 2016 CAPACITY GOAL


COMPARISON OF REQUIRED TRANSPONDER HARDWARE

10

20

100 1,000 10,000 Transmission distance [km]

Spec

tral

eff

icie

ncy

[b/s

/Hz]

N1= 1

N1: Number of TX/RX in high-SE, multiply regenerated system

N1=3

N1=30 N1=10

N1=300 N1=100

N1=1,000 N1=3,000 N1=10,000

3x

[P. J. Winzer, PTL 2011]


COMPARISON OF REQUIRED TRANSPONDER HARDWARE

10

20


Spec

tral

eff

icie

ncy

[b/s

/Hz]

N2 = 1

N2: Number of TX/RX in low-SE, parallel system

3x

N2=3 N2=4 N2=5 N2=6 N2=2



COMPARISON OF REQUIRED TRANSPONDER HARDWARE N2=2 N2=3 N2=4 N2=5 N2=6

10

20


Spec

tral

eff

icie

ncy

[b/s

/Hz]

N1= N2 = 1

N1: Number of TX/RX in high-SE, multiply regenerated system

N2: Number of TX/RX in low-SE, parallel system

N1=3

N1=30 N1=10

N1=300 N1=100

N1=1,000 N1=3,000 N1=10,000



5 Spatial multiplexing: The integration challenge


NEED INTEGRATION FOR ECONOMIC SUSTAINABILITY

TX RX

TX RX

TX RX

Integrated transponders, Integrated amplifiers, Multi-mode or Multi-core fiber

Schematic view

Deploying M spatial paths is better than using multiple regenerators But: M systems cost M times as much & consume M times the energy Cost/bit (or energy/bit) remains constant

Integration is key to scale space-division multiplexed systems


INTEGRATED 7-CORE MULTI-MODE INTERCONNECT

[B. G. Lee et al., IEEE Photonics Society Summer Topicals, 2010]

Fiber coupled to VCSEL array for 100-m interface at 120 Gb/s


INTEGRATED 7-CORE SINGLE-MODE RECEIVER

4.0 mm

TE TM

MZI demux Grating coupler

Photodiode

Fiber

1 2 3 4 5 6 7

[C.R.Doerr et al., PTL 2011]


7-CORE FIBER: LOW LOSS AND LOW CROSSTALK

[B.Zhu et al., ECOC 2011 and OptEx, 2011]

[K.Imamura et al., ECOC 2011]

[T. Hayashi et al., ECOC 2011]


SPATIAL MULTIPLEXING SETS FIRST CAPACITY RECORD

1986 1990 1994 1998 2002 2006 10

100

1

10

100

Syst

em c

apac

ity

Gb/

s Tb

/s

2010 2014 2018

WD

M c

hann

els

1

10

Pb/s

Spat

ial

mod

es


LONG DISTANCES AND HIGH SPECTRAL EFFICIENCIES

1,000 10,000 Transmission distance [km]

100

10

Agg

rega

te s

pec

tral

eff

icie

ncy

[b/s

/Hz]

20

30

40

50

60


ULTIMATELY, INTEGRATION WILL LEAD TO CROSSTALK

TX RX

TX RX

TX RX

Integrated transponders, Integrated amplifiers, Multi-mode or Multi-core fiber

Schematic view

How much crosstalk is tolerable?

Multiple-input multiple-output (MIMO) is a very successfully crosstalk cancellation technique... (Caveat: MIMO in optical has different boundary conditions from wireless)

[Winzer et al., ECOC 2011]


6 SPATIAL MULTIPLEXING USING MIMO


MIMO channel

MIMO-SDM IN FIBER – DIFFERENCES TO WIRELESS • Potential addressability of all propagation modes (complete set) • “Perturbed unitary” channel (mode-dependent loss) • Fiber nonlinearity (likely to set per-mode peak-power constraints) • High reliability requirements (99.999%), low outage probabilities (10-5) • Distributed noise • RX TX feedback almost always impossible • Nonlinear MIMO signal processing ?

MIM

O

Dec

oder

+

Channel matrix H

MT N0

N0

+ MR

MIM

O

Enco

der

M x M


Segment

MIM

O

Dec

oder

N1

n1 H1

M

n2 H2

+

nK HK

N2 NK

N1 N2 NK

MIM

O

Enco

der

M

+

+

+

+

+

DISTRIBUTED NOISE LOADING

In optics, noise loading is usually distributed over propagation (e.g., on a “per-span” or “per-segment” basis)

Spatial noise correlation

If segment matrices are unitary: [Winzer and Foschini, Optics Express, 2011]


MODE-DEPENDENT LOSS

Outage probabilities for M = 16 modes, K = 64 segments Mode-dependent loss MDLi per segment

1

10-2

Out

age

prob

abilit

y

0.4 0.6 0.8 1

10-4

Capacity (C/M)

MDLi [dB] = 5 2

1 0.5

Noise loading at receiver Distributed noise loading

Noise loading at receiver is worse than distributed noise loading A per-segment MDL of 1 dB only reduces capacity by


THE FIRST 6x6 OPTICAL MIMO EXPERIMENT

SDM

M

UX

3-mode fiber

• All guided modes are selectively launched and coherently detected (otherwise: System outage)

• Modes are strongly coupled during propagation in the fiber • Digital signal processing decouples received signals

Orthogonal mode

coupling

SDM

D

EMU

X

PD-Coh RX0

Out1 Out2 Out3 Out4 Out5 Out6

MIMO DSP

Orthogonal mode

coupling

Mode mixing

PD-Coh RX1

PD-Coh RX2

n1 n2 n3 n4 n5 n6

Ch1 Ch2 Ch3 Ch4 Ch5 Ch6

PD: Polarization Diversity

LP01 X-pol LP11a X-pol LP11b X-pol

Inte

nsity

LP01 Y-pol LP11a Y-pol LP11b Y-pol

[R. Ryf et al., OFC 2011; S. Randel et al., Optics Express 2011]



39

• Crosstalk between 3 cores with 2 polarizations each can be compensated using a linear 6 6 MIMO equalizer

6 6 ADAPTIVE MIMO EQUALIZER STRUCTURE

RX #3

X-Pol.

Carrier

Recovery

Carrier

Recovery

Carrier

Recovery

Carrier

Recovery

Carrier

Recovery

Carrier

Recovery

RX #2

X-Pol.

RX #1

Y-Pol.

RX #1

X-Pol.

RX #2

Y-Pol.

RX #3

Y-Pol.

TX #1

X-Pol

w1,1 w1,2 w1,3 w1,4 w1,5 w1,6

w2,1 w2,2 w2,3 w2,4 w2,5 w2,6

w3,1 w3,2 w3,3 w3,4 w3,5

w4,1 w4,2 w4,3

w5,1 w5,2 w5,3

w4,4

w5,4

w4,5

w5,5

w4,6

w5,6

w6,1 w6,2 w6,3 w6,4 w6,5 w6,6

w3,6

TX #1

Y-Pol

TX #2

X-Pol

TX #2

Y-Pol

TX #3

X-Pol

TX #3

Y-Pol

Randel et. al., Optics Express, 2011



40

IMPULSE RESPONSE MATRIX FOR 96-km 6-MODE FEW-MODE FIBER

-50

-30

-10 h 11

|h| 2

(dB

)

-50

-30

-10 h 21

|h| 2

(dB

)

-50

-30

-10 h 31

|h| 2

(dB

)

-50

-30

-10 h 41

|h| 2

(dB

)

-50

-30

-10 h 51

|h| 2

(dB

)

1 2 3 4 5 6

-50

-30

-10 h 61

t (ns)

|h| 2

(dB

)

h 12

h 22

h 32

h 42

h 52

1 2 3 4 5 6

h 62

t (ns)

h 13

h 23

h 33

h 43

h 53

1 2 3 4 5 6

h 63

t (ns)

h 14

h 24

h 34

h 44

h 54

1 2 3 4 5 6

h 64

t (ns)

h 15

h 25

h 35

h 45

h 55

1 2 3 4 5 6

h 65

t (ns)

h 16

h 26

h 36

h 46

h 56

1 2 3 4 5 6

h 66

t (ns)

•LP01x •LP01y •LP11ax •LP11ay •LP11bx •LP11by

•LP 0

1x

•LP 0

1y

•LP 1

1ax

•LP 1

1ay

•LP 1

1bx

•LP 1

1by

•Transmitted ports

•Rec

eive

d po

rts

• The impulse response was characterized for all 6 outputs as function of all 6 inputs

• Strong coupling is observed within the LP01 and the LP11 mode

• Weaker coupling is observed between the LP01 and LP11 mode

Ryf et. al., J. Lightwave Technol., 2012


A GLIMPSE INTO AN OPTICAL MIMO-SDM LAB


SOME FURTHER READING Single-mode fiber capacity limit • R.-J. Essiambre et al., “Capacity Limits of Optical Fiber Networks,” J. Lightwave Technol. 28, 662 (2010).

MIMO-SDM capacity scaling

• R. Ryf et al., “Mode-division multiplexing over 96 km of few-mode fiber using coherent 6x6 MIMO processing,” J. Lightwave Technol. 30, 521 (2012).

• S. Randel et al., “6 x 56-Gb/s Mode-Division Multiplexed Transmission over 33-km Few-Mode Fiber Enabled by 66 MIMO Equalization,” Optics Express (2011).

Overview on globally ongoing optical MIMO efforts • G. Li, X. Liu, “Focus Issue: Space Multiplexed Optical Transmission,” Opt. Ex. 19 16574 (2011). • T. Morioka et al. “Enhancing optical communications with brand new fibers,” IEEE Comm. Mag. 50, s31 (2012).

• P. J. Winzer, “Optical Networking Beyond WDM,” IEEE Photon. J. (2012).

• P. J. Winzer and G. J. Foschini, “MIMO Capacities and Outage Probabilities in Spatially Multiplexed Optical Transport Systems,” Optics Express (2011).

• P. J. Winzer “Energy-efficient optical transport capacity scaling through spatial multiplexing,” Photon. Technol. Lett. 23, 851 (2011).

First MIMO-SDM experiments



44

OPTICAL NETWORKING BEYOND WDM SPATIAL MULTIPLEXING AND MIMO FOR THE PETABIT ERAACKNOWLEDGMENTSlide Number 3HUMAN-DRIVEN TRAFFIC GROWTHMACHINE-DRIVEN TRAFFIC GROWTHSlide Number 6OPTICAL NETWORKS: WORKHORSE OF THE INTERNETHIGH-SPEED OPTICAL INTERFACE EVOLUTIONHIGH-SPEED OPTICAL INTERFACES – PRODUCTSSCALING INTERFACES TO TERABIT ETHERNETTHE SCALING OF WAVELENGTH-DIVISION MULTIPLEXINGSlide Number 12WHY IS AN OPTICAL FIBER NONLINEAR ?PHYSICAL PHENOMENA AT PLAYTHE NONLINEAR SHANNON LIMITAN LOWER BOUND ESTIMATE FOR THE SHANNON LIMITSOME EXAMPLE RESULTSSENSITIVITY ANALYSIS TO FIBER PARAMETERSREALITY CHECK: WHERE ARE WE EXPERIMENTALLY ?Slide Number 20DIMENSIONS FOR MODULATION AND MULTIPLEXINGTWO OPTIONS TO REACH OUR 2016 CAPACITY GOALCOMPARISON OF REQUIRED TRANSPONDER HARDWARECOMPARISON OF REQUIRED TRANSPONDER HARDWARECOMPARISON OF REQUIRED TRANSPONDER HARDWARESlide Number 26NEED INTEGRATION FOR ECONOMIC SUSTAINABILITYINTEGRATED 7-CORE MULTI-MODE INTERCONNECTINTEGRATED 7-CORE SINGLE-MODE RECEIVER7-CORE FIBER: LOW LOSS AND LOW CROSSTALKSPATIAL MULTIPLEXING SETS FIRST CAPACITY RECORDLONG DISTANCES AND HIGH SPECTRAL EFFICIENCIESULTIMATELY, INTEGRATION WILL LEAD TO CROSSTALKSlide Number 34MIMO-SDM IN FIBER – DIFFERENCES TO WIRELESSDISTRIBUTED NOISE LOADINGMODE-DEPENDENT LOSSTHE FIRST 6x6 OPTICAL MIMO EXPERIMENT6×6 ADAPTIVE MIMO EQUALIZER STRUCTUREIMPULSE RESPONSE MATRIX FOR 96-km 6-MODE FEW-MODE FIBERA GLIMPSE INTO AN OPTICAL MIMO-SDM LAB …OUTLOOK: THERE’S PLENTY OF WORK AHEAD !!!SOME FURTHER READINGSlide Number 44

Date post:	27-Jan-2021
Category:	Documents
Upload:	others
View:	0 times
Download:	0 times

OPTICAL NETWORKING BEYOND WDM[Nakazawa et al., OFC’10] PDM 64-QAM 21 GBaud (256 Gb/s) [Gnauck et...

Documents