Academic press optical fiber telecommunications volume b systems and networks 5th edition feb 2008 ISBN 0123741726 pdf

Electronic signal processing for dispersion compensation and error mitigation in optical transmission networks Abhijit Shanbhag, Qian Yu, and John Choma Dispersion equalization has its o

About the Editors

Ivan P Kaminow retired from Bell Labs in 1996 after a 42-year career Heconducted seminal studies on electrooptic modulators and materials, Raman scatter-ing in ferroelectrics, integrated optics, semiconductor lasers (DBR, ridge-waveguideInGaAsP, and multi-frequency), birefringent optical fibers, and WDM networks.Later, he led research on WDM components (EDFAs, AWGs, and fiber Fabry-PerotFilters), and on WDM local and wide area networks He is a member of the NationalAcademy of Engineering and a recipient of the IEEE/OSA John Tyndall, OSACharles Townes, and IEEE/LEOS Quantum Electronics Awards Since 2004, he hasbeen Adjunct Professor of Electrical Engineering at the University of California,Berkeley

Tingye Li retired from AT&T in 1998 after a 41-year career at Bell Labs andAT&T Labs His seminal work on laser resonator modes is considered a classic.Since the late 1960s, he and his groups have conducted pioneering studies onlightwave technologies and systems He led the work on amplified WDM trans-mission systems and championed their deployment for upgrading network capa-city He is a member of the National Academy of Engineering and a foreignmember of the Chinese Academy of Engineering He is also a recipient of theIEEE David Sarnoff Award, IEEE/OSA John Tyndall Award, OSA Ives Medal/Quinn Endowment, AT&T Science and Technology Medal, and IEEE PhotonicsAward

Alan E Willnerhas worked at AT&T Bell Labs and Bellcore, and he is Professor

of Electrical Engineering at the University of Southern California He received theNSF Presidential Faculty Fellows Award from the White House, Packard Founda-tion Fellowship, NSF National Young Investigator Award, Fulbright FoundationSenior Scholar, IEEE LEOS Distinguished Lecturer, and USC University-WideAward for Excellence in Teaching He is a Fellow of IEEE and OSA, and he hasbeen President of the IEEE LEOS, Editor-in-Chief of the IEEE/OSA J of LightwaveTechnology, Editor-in-Chief of Optics Letters, Co-Chair of the OSA Science &Engineering Council, and General Co-Chair of the Conference on Lasers andElectro-Optics

Optical Fiber Telecommunications V B

Systems and Networks

Edited by

Ivan P Kaminow

Tingye Li

Alan E Willner

Academic Press is an imprint of Elsevier

30 Corporate Drive, Suite 400, Burlington, MA 01803, USA

525 B Street, Suite 1900, San Diego, California 92101-4495, USA

84 Theobald’s Road, London WC1X 8RR, UK

This book is printed on acid-free paper. ¥

Copyright 2008, Elsevier Inc All rights reserved

No part of this publication may be reproduced or transmitted in any form or by anymeans, electronic or mechanical, including photocopy, recording, or any informationstorage and retrieval system, without permission in writing from the publisher

Permissions may be sought directly from Elsevier’s Science & Technology RightsDepartment in Oxford, UK: phone: (þ 44) 1865 843830, fax: (þ 44) 1865 853333,E-mail: permissions@elsevier.com You may also complete your request on-linevia the Elsevier homepage (http://elsevier.com), by selecting ‘‘Support & Contact’’then ‘‘Copyright and Permission’’ and then ‘‘Obtaining Permissions.’’

Library of Congress Cataloging-in-Publication Data

Application submitted

British Library Cataloguing-in-Publication Data

A catalogue record for this book is available from the British Library

ISBN: 978-0-12-374172-1

For information on all Academic Press publications

visit our Web site at www.books.elsevier.com

Printed in the United States of America

08 09 10 11 12 8 7 6 5 4 3 2 1

Working together to grow

libraries in developing countries

www.elsevier.com | www.bookaid.org | www.sabre.org

For Edith, Debbie, and Kathy with Love—TLFor Michelle, our Children (Moshe, Asher, Ari, Jacob), and

my Parents with Love—AEW

Ivan P Kaminow, Tingye Li, and Alan E Willner

Peter J Winzer and Rene´-Jean Essiambre

Kazuro Kikuchi

Xiang Liu, Sethumadhavan Chandrasekhar, and

Andreas Leven

Karsten Schuh and Eugen Lach

Chapter 6 Ultra-high-speed OTDM transmission technology 201

Hans-Georg Weber and Reinhold Ludwig

Alan E Willner, Zhongqi Pan, and Changyuan Yu

Mark D Feuer, Daniel C Kilper, and Sheryl L Woodward

Chapter 9 Optical Ethernet: Protocols, management, and 1–100 G

Cedric F Lam and Winston I Way

Chapter 10 Fiber-based broadband access technology and

Richard E Wagner

vii

Chapter 11 Global landscape in broadband: Politics, economics,

Richard Mack

Chapter 12 Metro networks: Services and technologies 477

Loukas Paraschis, Ori Gerstel, and Michael Y Frankel

Chapter 13 Commercial optical networks, overlay networks,

Robert Doverspike and Peter Magill

Chapter 14 Technologies for global telecommunications using

Keren Bergman, Department of Electrical Engineering, Columbia University,

1312 S.W Mudd, 500 West 120th Street, New York, NY 10027, USA,

Eugen Lach, Alcatel-Lucent Deutschland AG, Bell Labs Germany,

Lorenzstrasse 10, D-70435 Stuttgart, Germany, eugen.lach@alcatel-lucent.de

ix

Cedric F Lam, OpVista, 870 N McCarthy Blvd, Milpitas, CA, USA,

cflam@ieee.org

Andreas Leven, Bell Laboratories, Alcatel-Lucent, Holmdel, NJ, USA,

aleven@alcatel-lucent.com

Tingye Li, Boulder, CO, USA, tingyeli@aol.com

Xiang Liu, Bell Laboratories, Alcatel-Lucent, Department of Optical Networks,Holmdel, NJ, USA, xliu20@alcatel-lucent.com

Reinhold Ludwig, Fraunhofer Institute for Telecommunications,

Heinrich-Hertz-Institut, Einsteinufer 37, D-10587 Berlin, Germany,

Michael O’Mahony, Department of Electronic System Engineering,

University of Essex, Colchester, CO43SQ, UK, mikej@essex.ac.uk

Zhongqi Pan, Department of Electrical and Computer Engineering,

University of Louisiana at Lafayette, USA, zpan@louisiana.edu

Loukas Paraschis, Cisco Systems, 170 West Tasman Drive, San Jose, CA

95134, USA, loukas@cisco.com

Robert Scarmozzino, RSoft Design Group, 400 Executive Blvd., Ossining,

NY 10562, USA, rob@rsoftdesign.com

Karsten Schuh, Alcatel-Lucent Deutschland AG, Bell Labs Germany,

Lorenzstrasse 10, D-70435 Stuttgart, Germany, karsten.schuh@alcatel-lucent.deAlwyn J Seeds, Department of Electronic and Electrical Engineering,

University College London, Torrington Place, London WC1E 7JE, England,a.seeds@ee.ucl.ac.uk

Rodney S Tucker, Department of Electrical and Electronic Engineering,University of Melbourne, Vic., 3010, Australia, r.tucker@ee.unimelb.edu.auRichard E Wagner, Corning International, Corning, New York, USA,

wagnerre@corning.com

Winston I Way, OpVista, 870 N McCarthy Blvd, Milpitas, CA, USA,

wway@opvista.com

Hans-Georg Weber, Fraunhofer Institute for Telecommunications,

Heinrich-Hertz-Institut, Einsteinufer 37, D-10587 Berlin, Germany,

S J Ben Yoo, Department of Electrical and Computer Engineering,

University of California, Davis, CA, USA, yoo@ece.ucdavis.edu

Changyuan Yu, Department of Electrical and Computer Engineering,

National University of Singapore, Singapore, eleyc@nus.edu.sg

Overview of OFT V volumes A & B

†

Boulder, CO, USA

‡

University of Southern California, Los Angeles, CA, USA

Optical Fiber Telecommunications V (OFT V) is the fifth installment of the OFTseries Now 29 years old, the series is a compilation by the research and devel-opment community of progress in optical fiber communications Each editionreflects the current state-of-the-art at the time As Editors, we started with a cleanslate of chapters and authors Our goal was to update topics from OFT IV that arestill relevant as well as to elucidate topics that have emerged since the lastedition

• In the late 1980s, OFT II (Miller and Kaminow, 1988) was published after thefirst field trials and deployments of simple optical links By this time, theadvantages of multiuser optical networking had captured the imagination ofthe community and were highlighted in the book

• OFT III (Kaminow and Koch, 1997) explored the explosion in transmissioncapacity in the early-to-mid 1990s, made possible by the erbium-doped fiberamplifier (EDFA), wavelength division multiplexing (WDM), and dispersionmanagement

Optical Fiber Telecommunications V B: Systems and Networks

Copyright  2008, Elsevier Inc All rights reserved.

• By 2002, OFT IV (Kaminow and Li, 2002) dealt with extending the distanceand capacity envelope of transmission systems Subtle nonlinear and disper-sive effects, requiring mitigation or compensation in the optical and electricaldomains, were explored.

• The present edition of OFT, V, (Kaminow, Li, and Willner, 2008) moves theseries into the realm of network management and services, as well as employ-ing optical communications for ever-shorter distances Using the high band-width capacity in a cost-effective manner for customer applications takescenter stage In addition, many of the topics from earlier volumes are brought

up to date; and new areas of research which show promise of impact arefeatured

Although each edition has added new topics, it is also true that new challengesemerge as they relate to older topics Typically, certain devices may have ade-quately solved transmission problems for the systems of that era However, assystems become more complex, critical device technologies that might have beenconsidered a “solved problem” previously have new requirements placed uponthem and need a fresh technical treatment For this reason, each edition has grown

in sheer size, i.e., adding the new and, if necessary, reexamining the old

An example of this circular feedback mechanism relates to the fiber itself

At first, systems simply required low-loss fiber However, long-distance transmissionenabled by EDFAs drove research on low-dispersion fiber Further, advances in WDMand the problems of nonlinear effects necessitated development of nonzero dispersionfiber Cost considerations and ultra-high-performance systems, respectively, are driv-ing research in plastic fibers and ultra-low-polarization-dependent fibers We believethat these cycles will continue Each volume includes a CD-ROM with all the figuresfrom that volume Select figures are in color The volume B CD-ROM also has somesupplementary Powerpoint slides to accompany Chapter 19 of that volume

1.2 PERSPECTIVE OF THE PAST 6 YEARS

Our field has experienced an unprecedented upheaval since 2002 The irrationalexuberance and despair of the technology “bubble-and-bust” had poured untoldsums of money into development and supply of optical technologies, which wasfollowed by a depression-like period of over supply We are happy to say that, bynearly all accounts, the field is gaining strength again and appears to be entering astage of rational growth

What caused this upheaval? A basis seems to be related to a fundamentaldiscontinuity in economic drivers Around 2001, worldwide telecom traffic ceasedbeing dominated by the slow-growing voice traffic and was overtaken bythe rapidly growing Internet traffic The business community over-estimated the

growth rate, which generated enthusiasm and demand, leading to unsustainableexpectations Could such a discontinuity happen again? Perhaps, but chastenedinvestors now seem to be following a more gradual and sensible path Throughoutthe “bubble-and-bust” until the present, the actual demand for bandwidth hasgrown at a very healthy 80% per year globally; thus, real traffic demandexperienced no bubble at all The growth and capacity needs are real, and shouldcontinue in the future.

As a final comment, we note that optical fiber communications is firmlyentrenched as part of the global information infrastructure The only question ishow deeply will it penetrate and complement other forms of communications, e.g.,wireless, access, and on-premises networks, interconnects, satellites, etc Thisprospect is in stark contrast to the voice-based future seen by OFT, published in

1979, before the first commercial intercontinental or transatlantic cable systemswere deployed in the 1980s We now have Tb/s systems for metro and long-haulnetworks It is interesting to contemplate what topics and concerns might appear inOFT VI

1.3 OFT V VOLUME A: COMPONENTS

AND SUBSYSTEMS

1.3.1 Chapter 1 Overview of OFT V volumes A & B

(Ivan P Kaminow, Tingye Li, and Alan E Willner)

This chapter briefly reviews herewith all the chapters contained in both volumes ofOFT V

1.3.2 Chapter 2 Semiconductor quantum dots:

Genesis—The Excitonic Zoo—novel devices

for future applications (Dieter Bimberg)

The ultimate class of semiconductor nanostructures, i.e., “quantum dots” (QDs), isbased on “dots” smaller than the de Broglie wavelength in all three dimensions.They constitute nanometer-sized clusters that are embedded in the dielectricmatrix of another semiconductor They are often self-similar and can be formed

by self-organized growth on surfaces Single or few quantum dots enable noveldevices for quantum information processing, and billions of them enable activecenters in optoelectronic devices like QD lasers or QD optical amplifiers Thischapter covers the area of quantum dots from growth via various band structures tooptoelectronic device applications In addition, high-speed laser and amplifieroperations are described

1.3.3 Chapter 3 High-speed low-chirp semiconductor lasers (Shun Lien Chuang, Guobin Liu, and Piotr Konrad Kondratko)

One advantage of using quantum wells and quantum dots for the active region oflasers is the lower induced chirp when such lasers are directly modulated, permit-ting direct laser modulation that can save on the cost of separate externalmodulators This chapter provides a comparison of InAlGaAs with InGaAsPlong-wavelength quantum-well lasers in terms of high-speed performance, andextracts the important parameters such as gain, differential gain, photon lifetime,temperature dependence, and chirp Both DC characteristics and high-speed directmodulation of quantum-well lasers are presented, and a comparison with theore-tical models is made The chapter also provides insights into novel quantum-dotlasers for high-speed operation, including the ideas of p-type doping vs tunnelinginjection for broadband operation

1.3.4 Chapter 4 Recent advances in surface-emitting lasers (Fumio Koyama)

Vertical cavity surface-emitting lasers (VCSELs) have a number of special ties (compared with the more familiar edge-emitting lasers) that permit some novelapplications This chapter begins with an introduction which briefly surveys recentadvances in VCSELs, several of that are then treated in detail These includetechniques for realizing long-wavelength operation (as earlier VCSELs werelimited to operation near 850 nm), the performance of dense VCSEL arrays thatemit a range of discrete wavelengths (as large as 110 in number), and MEMS-based athermal VCSELs Also, plasmonic VCSELs that produce subwavelengthspots for high-density data storage and detection are examined Finally, work onall-optical signal processing and slow light is presented

proper-1.3.5 Chapter 5 Pump diode lasers (Christoph Harder)

Erbium-doped fiber amplifiers (EDFAs) pumped by bulky argon lasers wereknown for several years before telecom system designers took them seriously;the key development was a compact, high-power semiconductor pump laser.Considerable effort and investment have gone into today’s practical pump lasers,driven by the importance of EDFAs in realizing dense wavelength division multi-plexed (DWDM) systems The emphasis has been on high power, efficiency, andreliability The two main wavelength ranges are in the neighborhood of 980 nm forlow noise and 1400 nm for remote pumping of EDFAs The 1400-nm band is alsosuitable for Raman amplifiers, for which very high power is needed

This chapter details the many lessons learned in the design for manufacture ofcommercial pump lasers in the two bands Based on the performance developed fortelecom, numerous other commercial applications for high-power lasers haveemerged in manufacturing and printing; these applications are also discussed.

1.3.6 Chapter 6 Ultrahigh-speed laser modulation by

injection locking (Connie J Chang-Hasnain and

Xiaoxue Zhao)

It has been known for decades that one oscillator (the slave) can be locked infrequency and phase to an external oscillator (the master) coupled to it Currentstudies of injection-locked lasers show that the dynamic characteristics of thedirectly modulated slave are much improved over the same laser when freelyrunning Substantial improvements are found in modulation bandwidth for analogand digital modulation, in linearity, in chirp reduction and in noise performance

In this chapter, theoretical and experimental aspects of injection locking in alllasers are reviewed with emphasis on the authors’ research on VCSELs (verticalcavity surface-emitting lasers) A recent promising application in passive opticalnetworks for fiber to the home (FTTH) is also discussed

1.3.7 Chapter 7 Recent developments in high-speed

optical modulators (Lars Thyle´n, Urban Westergren, Petter Holmstro¨m, Richard Schatz, and Peter Ja¨nes)

Current high-speed lightwave systems make use of electro-optic modulators based

on lithium niobate or electroabsorption modulators based on semiconductor als In commercial systems, the very high-speed lithium niobate devices often require

materi-a trmateri-aveling wmateri-ave structure, while the semiconductor devices materi-are usumateri-ally lumped.This chapter reviews the theory of high-speed modulators (at rates of 100 Gb/s)and then considers practical design approaches, including comparison of lumped andtraveling-wave designs The main emphasis is on electroabsorption devices based onFranz–Keldysh effect, quantum-confined Stark effect and intersubband absorption

A number of novel designs are described and experimental results given

1.3.8 Chapter 8 Advances in photodetectors

(Joe Charles Campbell)

As a key element in optical fiber communications systems, photodetectors belong to

a well developed sector of photonics technology Silicon p–i–n and avalanche diodes deployed in first-generation lightwave transmission systems operating at0.82-mm wavelength performed very close to theory In the 1980s, InP photodiodes

photo-were developed and commercialized for systems that operated at 1.3- and 1.5-mmwavelengths, albeit the avalanche photodiodes (APDs) were expensive and nonideal.Introduction of erbium-doped fiber amplifiers and WDM technology in the 1990srelegated APDs to the background, as p–i–n photoreceivers performed well in amplifiedsystems, whereas APDs were plagued by the amplified spontaneous emission noise.Future advanced systems and special applications will require sophisticated devicesinvolving deep understanding of device physics and technology This chapter focuses onthree primary topics: high-speed waveguide photodiodes for systems that operate at

100 Gb/s and beyond, photodiodes with high saturation current for high-power tions, and recent advances of APDs for applications in telecommunications

applica-1.3.9 Chapter 9 Planar lightwave circuits in fiber-optic communications (Christopher R Doerr and

Katsunari Okamoto)

The realization of one or more optical waveguide components on a planar substratehas been under study for over 35 years Today, individual components such assplitters and arrayed waveguide grating routers (AWGRs) are in widespreadcommercial use Sophisticated functions, such as reconfigurable add–dropmultiplexers (ROADMs) and high-performance filters, have been demonstrated

by integrating elaborate combinations of such components on a single chip Forthe most part, these photonic integrated circuits (PICs), or planar lightwavecircuits (PLCs), are based on passive waveguides in lower index materials,such as silica

This chapter deals with the theory and design of such PICs The following twochapters (Chapters 11 and 12) also deal with PICs; however, they are designed to

be integrated with silicon electronic ICs, either in hybrid fashion by short wirebonds to an InP PIC or directly to a silicon PIC

1.3.10 Chapter 10 III–V photonic integrated circuits

and their impact on optical network architectures (Dave Welch, Chuck Joyner, Damien Lambert, Peter

W Evans, and Maura Raburn)

InP-based semiconductors are unique in their capability to support all the photoniccomponents required for wavelength division multiplexed (WDM) transmittersand receivers in the telecom band at 1550 nm Present subsystems have connectedthese individual components by fibers or lenses to form hybrid transmitters andreceivers for each channel

Recently, integrated InP WDM transmitter and receiver chips that provide

10 channels, each operating at 10 Gb/s, have been shown to be technically and

economically viable for deployment in commercial WDM systems The photonicintegrated circuits are wire-bonded to adjacent silicon ICs Thus a single boardprovides optoelectronic regeneration for 10 channels, dramatically reducinginterconnection complexity and equipment space In addition, as in legacy single-channel systems, the “digital” approach for transmission (as compared to “all-optical”)offers ease of network monitoring and management This chapter covers the technology

of InP photonic integrated circuits (PICs) and their commercial application The impact

on optical network architecture and operation is discussed and technology advances forfuture systems are presented

1.3.11 Chapter 11 Silicon photonics (Cary Gunn

and Thomas L Koch)

Huge amounts of money have been invested in silicon processing technology,thanks to a steady stream of applications that justified the next stage of proces-sing development In addition to investment, innovative design, process disci-pline and large-volume runs made for economic success The InP PICs described

in the previous chapter owe their success to lessons learned in silicon ICprocessing

Many people have been attracted by the prospects of fabricating PICs usingsilicon alone to capitalize on the investment and success of silicon ICs To succeedone requires a large-volume application and a design that can be made in anoperating silicon IC foundry facility A further potential advantage is the oppor-tunity to incorporate on the same photonic chip electronic signal processing Theapplication to interconnects for high-performance computers is a foremost motiva-tion for this work

While silicon has proven to be the ideal material for electronic ICs, it is far fromideal for PICs The main shortcoming is the inability so far to make a good lightsource or photodetector in silicon This chapter discusses the successes andchallenges encountered in realizing silicon PICs to date

1.3.12 Chapter 12 Photonic crystal theory: Temporal

coupled-mode formalism (Shanhui Fan)

Photonic crystal structures have an artificially created optical bandgap that isintroduced by a periodic array of perturbations, and different types of wave-guides and cavities can be fabricated that uniquely use the band gap-basedconfinement These artificially created materials have been of great interest forpotential optical information processing applications, in part because they pro-vide a common platform to miniaturize a large number of optical componentson-chip down to single wavelength scale For this purpose, many devices can be

designed using such a material with a photonic bandgap and, subsequently,introducing line and point defect states into the gap Various functional devices,such as filters, switches, modulators and delay lines, can be created by control-ling the coupling between these defect states This chapter reviews the temporalcoupled-mode theory formalism that provides the theoretical foundation ofmany of these devices.

1.3.13 Chapter 13 Photonic crystal technologies:

Experiment (Susumu Noda)

Photonic crystals belong to a class of optical nanostructures characterized by theformation of band structures with respect to photon energy In 3D photoniccrystals, a complete photonic band gap is formed; the presence of light withfrequencies lying in the band gap is not allowed This chapter describes theapplication of various types of materials engineering to photonic crystals, withparticular focus on the band gap/defect, the band edge, and the transmission bandwithin each band structure The manipulation of photons in a variety of waysbecomes possible Moreover, this chapter discusses the recent introduction of

“photonic heterostructures” as well as recent developments concerning two- andthree-dimensional photonic crystals

1.3.14 Chapter 14 Photonic crystal fibers: Basics

and applications (Philip St John Russell)

Photonic crystal fibers (PCFs)—fibers with a periodic transverse microstructure—first emerged as practical low-loss waveguides in early 1996 The initial demon-stration took 4 years of technological development, and since then the fabricationtechniques have become more and more sophisticated It is now possible tomanufacture the microstructure in air–glass PCFs to accuracies of 10 nm on thescale of 1mm, which allows remarkable control of key optical properties such asdispersion, birefringence, nonlinearity and the position and width of the photonicband gaps (PBGs) in the periodic “photonic crystal” cladding PCF has in this wayextended the range of possibilities in optical fibers, both by improving well-established properties and introducing new features such as low-loss guidance in

a hollow core

In this chapter, the properties of the various types of PCFs are introduced,followed by a detailed discussion of their established or emerging applications.The chapter describes in detail the fabrication, theory, numerical modeling,optical properties, and guiding mechanisms of PCFs Applications of photoniccrystal fibers include lasers, amplifiers, dispersion compensators, and nonlinearprocessing

1.3.15 Chapter 15 Specialty fibers for optical

communication systems (Ming-Jun Li, Xin Chen, Daniel A Nolan, Ji Wang, James A West, and

Karl W Koch)

Specialty fibers are designed by changing fiber glass composition, refractive indexprofile, or coating to achieve certain unique properties and functionalities Some ofthe common specialty fibers include active fibers, polarization control fibers,dispersion compensation fibers, highly nonlinear fibers, coupling or bridge fibers,high-numerical-aperture fibers, fiber Bragg gratings, and special single modefibers In this chapter, the design and performance of various specialty fibers arediscussed Special attention is paid to dispersion compensation fibers, polarization-maintaining and single-polarization fibers, highly nonlinear fibers, double cladfiber for high-power lasers and amplifiers, and photonic crystal fibers Moreover,there is a brief discussion of the applications of these specialty fibers

1.3.16 Chapter 16 Plastic optical fibers: Technologies

and communication links (Yasuhiro Koike and

Satoshi Takahashi)

Plastic optical fiber (POF) consists of a plastic core that is surrounded by a plasticcladding of a refractive index lower than that of the core POFs have very largecore diameters compared to glass optical fibers, and yet they are quite flexible.These features enable easy installation and safe handling Moreover, the large-corefibers can be connected without high-precision accuracy and with low cost POFshave been used extensively in short-distance datacom applications, such as indigital audio interfaces POFs are also used for data transmission within equipmentand for control signal transmission in machine tools During the late 1990s, POFswere used as the transmission medium in the data bus within automobiles As wemove into the future, high-speed communication will be required in the home, andPOFs are a promising candidate for home network wiring This chapter describesthe POF design and fabrication, the specific fiber properties of attenuation, band-width and thermal stability, and various communications applications, concludingwith a discussion of recent developments in graded-index POFs

1.3.17 Chapter 17 Polarization mode dispersion (Misha

Brodsky, Nicholas J Frigo, and Moshe Tur)

Polarization-mode dispersion (PMD) has been well recognized for sometime as animpairment factor that limits the transmission speed and distance in high-speedlightwave systems The complex properties of PMD have enjoyed scrutiny by

theorists, experimentalists, network designers, field engineers and, during the

“bubble” years, entrepreneurial technologists A comprehensive treatment of thesubject up to year 2002 is given in a chapter bearing the same title in Optical FiberTelecommunications IVB, System and Impairments The present chapter is anoverview of PMD with special emphasis on the knowledge accumulated in thepast 5 years It begins with a review of PMD concepts, and proceeds to considerthe “hinge” model used to describe field test results, which are presented andanalyzed The important subject of system penalties and outages due to first-orderPMD is then examined, followed by deliberations of higher-order PMD, andinteraction between fiber nonlinear effects and PMD

1.3.18 Chapter 18 Electronic signal processing for

dispersion compensation and error mitigation in optical transmission networks (Abhijit Shanbhag, Qian Yu, and John Choma)

Dispersion equalization has its origin in the early days of analog transmission ofvoice over copper wires where loading coils (filters) were distributed in the net-work to equalize the frequency response of the transmission line Digital transmis-sion over twisted pairs was enabled by the invention of the transversal equalizerwhich extended greatly the bandwidth and reach Sophisticated signal processingand modulation techniques have now made mobile telephones ubiquitous How-ever, it was not until the mid 1990s that wide deployment of Gigabit Ethernetrendered silicon CMOS ICs economical for application in high-speed lightwavetransmission Most, if not all lightwave transmission systems deployed today, useelectronic forward error correction and dispersion compensation to alleviate signaldegradation due to noise and fiber dispersive effects

This chapter presents an overview of various electronic equalization and tation techniques, and discusses their high-speed implementation, specificallyaddressing 10-Gb/s applications for local-area, metro, and long-haul networks

adap-It comprises a comprehensive survey of the role, scope, limitations, trends, andchallenges of this very important and compelling technology

1.3.19 Chapter 19 Microelectromechanical systems for

lightwave communication (Ming C Wu, Olav

Solgaard, and Joseph E Ford)

The earliest commercial applications of microelectromechanical systems (MEMS)were in digital displays employing arrays of tiny mirrors and in accelerometers forairbag sensors This technology has now found a host of applications in lightwavecommunications These applications usually require movable components, such asmirrors, with response times in the neighborhood of 10–6s, although fixed

elements may be called for in some applications Either a free-space or integratedlayout may be used.

This chapter describes the recent lightwave system applications of MEMS

In telecommunications, MEMS switches can provide cross-connects with largenumbers of ports A variety of wavelength selective devices, such as reconfigurableoptical add–drop multiplexers (ROADM) employ MEMS More recent devicesinclude tunable lasers and microdisk resonators

1.3.20 Chapter 20 Nonlinear optics in communications:

from crippling impairment to ultrafast tools (Stojan Radic, David J Moss, and Benjamin J Eggleton)

It is perhaps somewhat paradoxical that optical nonlinearities, whilst having posedsignificant limitations for long-haul WDM systems, also offer the promise of addres-sing the bandwidth bottleneck for signal processing for future optical networks asthey evolve beyond 40 Gb/s In particular, all-optical devices based on the 3rd order

(3)optical nonlinearity offer a significant promise in this regard, not only becausethe intrinsic nonresonant (3) is nearly instantaneous, but also because (3) isresponsible for a wide range of phenomena, including 3rd harmonic generation,stimulated Raman gain, four-wave mixing, optical phase conjugation, two-photonabsorption, and the nonlinear refractive index This plethora of physical processeshas been the basis for a wide range of activity on all-optical signal processing devices.This chapter focuses on breakthroughs in the past few years on approachesbased on highly nonlinear silica fiber as well as chalcogenide-glass-based fiber andwaveguide devices The chapter contrasts two qualitatively different approaches toall-optical signal processing based on nonphase-matched and phase-matched pro-cesses All-optical applications of 2R and 3R regeneration, wavelength conver-sion, parametric amplification, phase conjugation, delay, performance monitoring,and switching are reviewed

1.3.21 Chapter 21 Fiber-optic quantum information

technologies (Prem Kumar, Jun Chen, Paul L Voss, Xiaoying Li, Kim Fook Lee, and Jay E Sharping)

Quantum-mechanical (QM) rules are surprisingly simple: linear algebra and order partial differential equations Yet, QM predictions are unimaginably preciseand accurate when compared with experimental data A “mysterious” feature of QM

first-is the superposition principle and the ensuing quantum entanglement The mental difference between quantum entanglement and classical correlation lies in thefact that particles are quantum-mechanical objects which can exist not only in states

funda-| 0> and | 1> but also in states described by  | 0 > þ b | 1 >, while classical objects

can only exist in one of two deterministic states (i.e., “heads” or “tails”), and notsomething in between In other words, the individual particle in quantum entangle-ment does not have a well-defined pure state before measurement.

Since the beginning of the 1990s, the field of quantum information and munication has expanded rapidly, with quantum entanglement being a criticalaspect Entanglement is still an unresolved “mystery,” but a new world of

com-“quantum ideas” has been ignited and is actively being pursued The focus ofthis chapter is the generation of correlated and entangled photons in the telecomband using the Kerr nonlinearity in dispersion-shifted fiber Of particular interestare microstructure fibers, in which tailorable dispersion properties haveallowed phase-matching and entanglement to be obtained over a wide range ofwavelengths

1.4 OFT V VOLUME B: SYSTEMS AND NETWORKS 1.4.1 Chapter 1 Overview of OFT V volumes A & B

(Ivan P Kaminow, Tingye Li, and Alan E Willner)

This chapter briefly reviews herewith all the chapters contained in both volumes ofOFT V

1.4.2 Chapter 2 Advanced optical modulation formats (Peter J Winzer and Rene´-Jean Essiambre)

Today, digital radio-frequency (rf) communication equipment employs cated signal processing and communication theory technology to realize amazingperformance; wireless telephones are a prime example These implementations aremade possible by the capabilities and low cost of silicon integrated circuits in high-volume consumer applications Some of these techniques, such as forward errorcorrection (FEC) and electronic dispersion compensation (EDC) are currently inuse in lightwave communications to enhance signal-to-noise ratio and mitigatesignal degradation (See the chapter on “Electronic Signal Processing for Disper-sion Compensation and Error Mitigation in Optical Transmission Networks” byAbhijit Shanbhag, Qian Yu, and John Choma.) Advanced modulation formats thatare robust to transmission impairments or able to improve spectral efficiency arebeing considered for next-generation lightwave systems

sophisti-This chapter provides a taxonomy of optical modulation formats, along withexperimental techniques for realizing them The discussion makes clear the sub-stantial distinctions between design conditions for optical and rf applications.Demodulation concepts for coherent and delay demodulation are also coveredanalytically

1.4.3 Chapter 3 Coherent optical communication systems (Kazuro Kikuchi)

The first generation of single-channel fiber optic networks used on-off keying anddirect detection Later, coherent systems, employing homodyne and heterodynedetection, were intensely researched with the aim of taking advantage of theirimproved sensitivity and WDM frequency selectivity However, the quick success

of EDFAs in the 1990s cut short the prospects for coherent systems

Now, interest in coherent is being renewed as the need for greater spectral ciency in achieving greater bandwidth per fiber has become apparent This chapterreviews the theory of multilevel modulation formats that permit multiple bits/s of dataper Hz of bandwidth (See the chapter on “Advanced Modulation Formats” by Winzerand Essiambre.) The growing capabilities of silicon data signal processing (DSP) can

effi-be combined with digital coherent detection to provide dramatic improvements inspectral efficiency Experimental results for such receivers are presented

1.4.4 Chapter 4 Self-coherent optical transport systems (Xiang Liu, Sethumadhavan Chandrasekhar, and

Andreas Leven)

As stated above, coherent detection transmission systems were investigated in the1980s for their improved receiver sensitivity and selectivity, and for the promise ofpossible postdetection dispersion compensation However, the emergence of EDFAsand amplified WDM systems relegated the technically difficult coherent technology tothe background Now, as high-speed signal processing technology becomes technicallyand economically feasible, there is renewed interest in studying coherent and self-coherent systems, especially for their capability to increase spectral efficiency throughthe use of advanced multilevel modulation techniques and, more important, for thepossibility of implementing postdetection equalization functionalities

Self-coherent systems utilize differential direct detection that does not require alocal oscillator With high-speed analog-to-digital conversion and digital signalprocessing, both phase and amplitude of the received optical field can be recon-structed, thus offering unprecedented capability for implementing adaptive equaliza-tion of transmission impairments This chapter is a comprehensive and in-depthtreatment of self-coherent transmission systems, including theoretical considerations,receiver technologies, modulation formats, adaptive equalization techniques, andapplications for capacity upgrades and cost reduction in future optical networks

1.4.5 Chapter 5 High-bit-rate ETDM transmission

systems (Karsten Schuh and Eugen Lach)

Historically, it has been observed that the first cost of a (single-channel) sion system tended to increase as the square root of its bandwidth or bit rate This

transmis-observation has prompted the telecom industry to develop higher-speed systemsfor upgrading transport capacity Indeed, there is a relentless drive to explorehigher speed for multichannel amplified WDM transmission where, for a givenspeed of operation, the total system cost is roughly proportional to the number ofchannels plus a fixed cost It is important to note that the cost of equalizing forsignal impairment at higher speeds must be taken into account.

This chapter is an up-to-date review of high-speed transmission using electronictime division multiplexing (ETDM), a time-honored approach for upgradingsystem capacity The emphasis is on 100-Gb/s bit rate and beyond, as 40-Gb/ssystems are already being deployed and 100-Gb/s Ethernet (100 GE) is expected to

be the next dominant transport technology The chapter includes a basic treatment

of ETDM technology, followed by a description of the concepts of high-speedETDM systems Requirements of optical and electronic components and the state-of-the-art technologies are then examined in detail, and an up-to-date overview ofultra-high-speed systems experiments is presented Finally, prospects of the var-ious approaches for rendering cost-effective 100 GE are contemplated

1.4.6 Chapter 6 Ultra-high-speed OTDM transmission technology (Hans-Georg Weber and Reinhold Ludwig)

The expected increase of transmission capacity in optical fiber networks willinvolve an optimized combination of WDM and TDM TDM may be realized byelectrical multiplexing (ETDM) or by optical multiplexing (OTDM) Dispersionimpairment notwithstanding, OTDM offers a means to increase the single-channelbit rate beyond the capability of ETDM Thus OTDM transmission technology isoften considered to be a research means with which to investigate the feasibility ofultra-high-speed transmission Historically, the highest speed commercial systemshave been ETDM systems Latest examples are 40 G systems being deployed atpresent and (serial) 100 G systems expected to be commercially available in a fewyears In the past 10 years, OTDM transmission technology has made considerableprogress towards much higher bit rates and much longer transmission links.This chapter discusses ultra-high-speed data transmission in optical fibers based

on OTDM technology The chapter gives a general description of an OTDMsystem, the OTDM transmitter, the OTDM receiver, and the fiber transmissionline WDM/OTDM transmission experiments are also described

1.4.7 Chapter 7 Optical performance monitoring

(Alan E Willner, Zhongqi Pan, and Changyuan Yu)

Today’s optical networks function in a fairly static fashion and are built tooperate within well-defined specifications This scenario is quite challenging for

higher-capacity systems, since network paths are not static and channel-degradingeffects can change with temperature, component drift, aging and fiber plantmaintenance In order to enable robust and cost-effective automated operation,the network should be able to: (i) intelligently monitor the physical state of thenetwork as well as the quality of propagating data signals, (ii) automaticallydiagnose and repair the network, and (iii) redirect traffic To achieve this, opticalperformance monitoring should isolate the specific cause of the problem Further-more, it can be quite advantageous to determine when a data signal is beginning todegrade, so that the network can take action to correct the problem or to route thetraffic around the degraded area.

This chapter explores optical performance monitoring and its potential forenabling higher stability, reconfigurability, and flexibility in an optical network.Moreover, this chapter describes the specific parameters that a network might want

to monitor, such as chromatic dispersion, polarization-mode dispersion, and cal SNR Promising monitoring techniques are reviewed

opti-1.4.8 Chapter 8 ROADMs and their system applications (Mark D Feuer, Daniel C Kilper, and Sheryl L.

Woodward)

As service providers begin to offer IPTV services in addition to data and voice, theneed for fast and flexible provisioning of mixed services and for meeting unpre-dictable traffic demand becomes compelling Reconfigurable optical add/dropmultiplexers (ROADMs) have emerged as the network element that can satisfythis need Indeed, subsystem and system vendors are rapidly developing andproducing ROADMs, and carriers are installing and deploying them in theirnetworks

This chapter is a comprehensive treatment of ROADMs and their application inWDM transmission systems and networks, comprising a review of various ROADMtechnologies and architectures; analyses of their routing functionalities and economicadvantages; considerations of design features and other requirements; and discussions

of the design of ROADM transmission systems and the interplay between theROADM and transmission performance The chapter ends with some thoughts onthe remaining challenges to enable ROADMs to achieve their potential

1.4.9 Chapter 9 Optical Ethernet: Protocols, management, and 1–100 G technologies (Cedric F Lam and

Winston I Way)

As the Internet becomes the de facto platform for the delivery of voice, data, andvideo services, Ethernet has become the technology of choice for access and metro

networks, and for next-generation long-haul networks As stated concisely by theauthors, “The success of Ethernet is attributed to its simplicity, low cost, standardimplementation, and interoperability guarantee,” attributes that helped the

“networking community it serves to prosper, hence producing the economy of scale.”This chapter is an in-depth review of the evolution and development of Ethernettechnology for application in optical fiber telecommunications networks Topicscovered include: point-to-point Ethernet development, Layer-2 functions, CarrierEthernet, Ethernet in access PONs, Ethernet OAM (Operation, Administration,Maintenance), development of 10 GE for PON and 100 GE for core applications,and examples of high-speed Ethernet

1.4.10 Chapter 10 Fiber-based broadband access

technology and deployment (Richard E Wagner)

One of the earliest long-haul commercial optical fiber telecom systems was theAT&T Northeast Corridor link from Boston to New York to Washington in 1983

In this application, the large capital investment could be amortized among manyusers The prospect of economically bringing fiber all the way to a large number ofend users, where cost sharing is not available, has continuously appealed to andchallenged the telecom industry Presently, technology advances and volumemanufacture are reducing costs/user, fabulous broadband applications are luringsubscribers, and government legislation and subsidies are encouraging growthworldwide This chapter tracks the history of broadband access, compares thecompeting access technologies, and projects the roadmap to future deployment inthe US, Asia, Europe and the rest of the world The economic driver for widespreaddeployment is the explosive growth of Internet traffic, which doubles annually indeveloped countries and grows even faster in developing countries, such as China

In developed countries, growth is due to new broadband applications; in developingcountries, both new users and new applications drive traffic growth

This chapter focuses on the fiber-based approaches to broadband access worldwide,including some of the drivers for deployment, the architectural options, the capital andoperational costs, the technological advances, and the future potential of these systems.Three variants of fiber-based broadband access, collectively called FTTx in thischapter, have emerged as particularly important They are: hybrid-fiber-coax (HFC)systems, fiber-to-the-cabinet (FTTC) systems, and fiber-to-the-home (FTTH) systems

1.4.11 Chapter 11 Global landscape in broadband:

Politics, economics, and applications (Richard Mack)

The technology of choice that predominates in a specific telecom arena dependsultimately upon the competitive economics: the normalized capital and

operational costs in dollars per unit bandwidth (per unit distance) For metro,regional, and long-haul arenas, lightwave technology is indisputably theking However, in the access arena, the competitive unit cost of lightwavetechnology has not favored rational deployment Indeed, the history ofFTTH has followed a tortuous path; the early trials in the 1980s and 1990s didnot lead to massive deployment Globally, Japanese and Korean telecom com-panies have been leading the installation of FTTH (with as yet unknown eco-nomic consequences) In the meantime, the cost of FTTH equipment has beendecreasing steadily Recently, relief from “unbundling” (exemption fromrequirement for incumbent carriers to share facilities with competitive carriers,

as ruled by FCC) and competition from cable TV companies have promptedincumbent carriers in the US to install FTTH with competitive (triple-play)service offerings As the demand for broadband services grows and revenueimproves, the return from the vast investment in FTTH may be realized in thenot-to-distant future

This chapter is a fascinating, data-laden account of the history of deployment ofoptical fiber telecommunications, with emphasis on economics, growth landscape,and broadband services in the access arena The discussion includes historicalhighlights, demographics, costs and revenues, fiber installations, services scenar-ios, competition and growth, regulatory policies, applications and bandwidthrequirements, technology and network architecture choices, market scenarios,etc The interplay of these issues is discussed and summarized in the concludingsection

1.4.12 Chapter 12 Metro networks: Services and

technologies (Loukas Paraschis, Ori Gerstel,

and Michael Y Frankel)

Metropolitan networks operate in the environs of a major city, transporting andmanaging traffic to meet the diverse service needs of enterprise and residentialapplications Typically, metro networks have a reach below a few hundred kilometerswith node traffic capacities and traffic granularity that require amplified dense WDMtechnology with optical add/drop, although the more economical coarse WDMtechnology has also been deployed At present, convergence of IP services andtraditional time-division multiplexed (TDM) traffic with low operational cost is animportant issue

This chapter reviews the architecture and optical transport of metro networks,which have evolved to meet the demand of various applications and services,including discussions of the evolution of network architecture, physical buildingblocks of the WDM network layer, requirements of network automation, andconvergence of packetized IP with traditional TDM traffic, and ending with abrief perspective on the future outlook

1.4.13 Chapter 13 Commercial optical networks, overlay

networks, and services (Robert Doverspike and

Peter Magill)

As service providers are the ultimate users of novel technologies and systems intheir networks, it is important that the innovators have a sound understanding ofthe structure and workings of the carriers’ networks: the architecture and layers,traffic and capacity demands, management of reliability and services, etc Com-mercial networks are continuously upgraded to provide more capacity, new ser-vices and reduced capital and operational cost; seamless network evolution isessential for obvious economic reasons Even when “disruptive” technologiesand platforms are introduced, smooth integration within the existing infrastructure

tech-“industry focuses on advanced optical technologies for the long-distance work, most of the investment and opportunity for growth resides in the metro/access [sector].”

net-1.4.14 Chapter 14 Technologies for global telecommunications

using undersea cables (Se´bastien Bigo)

The introduction of WDM has enabled a tremendous capacity growth in underseasystems, both by the increase in the number of carrier wavelengths and by theincrease in the channel bit rate Starting from 2.5 Gbit/s in the mid-1990s, the bitrate was upgraded in commercial products to 10 Gbit/s at the end of the lastcentury The next generation of undersea systems will likely be based on 40-Gbit/sbit-rate channels However, transmission at 40 Gbit/s is significantly more chal-lenging than at 10 Gbit/s

This chapter gives an overview of the specificities of submarine links withrespect to terrestrial links and provides a few examples of recently deployedundersea systems Moreover, it describes the key technologies involved in under-sea systems, explaining particular implementations of early optical systems,today’s 10-Gbit/s systems, and future 40-Gbit/s links The key technologies arerelated to fiber selection and arrangement, to amplifier design, to modulationformats, to detection techniques, and to advanced impairment-mitigationsolutions

1.4.15 Chapter 15 Future optical networks

(Michael O’Mahony)

In the past few years Internet traffic, doubling annually, has dominated the networkcapacity demand, which has been met by the advances in lightwave communications.The transformation from a circuit-switched, voice-centric to a packet-switched,IP-centric network is well underway; amplified WDM transmission systems withterabits-per-second capacity are being deployed; rapid reconfigurable networking andautomatic service provisioning are being implemented The drive to reduce cost andincrease revenue has being inexorable In the meantime, carriers are installing FTTHand offering IPTV services, which will undoubtedly change network traffic charac-teristics and boost the traffic growth rate What will the future networks look like?This chapter reviews the growth of the data traffic and evolution of the opticalnetwork, including user communities, global regional activities, and servicerequirements Discussions cover the diversity of architectures, the evolution ofswitching, cross-connecting and routing technologies, and the transformation tocarrier-grade (100 G) Ethernet The author notes that device integration “willenable the realization of many of the key functionalities for optical networking.”

1.4.16 Chapter 16 Optical burst and packet switching

(S J Ben Yoo)

Optical switching has the potential of providing more-efficient and put networking than its electronic counterpart This chapter discusses optical burstand packet switching technologies and examines their roles in future optical net-works It covers the roles of optical circuit, burst, and packet switching systems inoptical networks, as well as their respective benefits and trade-offs A description isgiven of the networking architecture/protocols, systems, and technologies pertaining

higher-through-to optical burst and packet switching Furthermore, this chapter introduces label switching technology, which provides a unified platform for interoperatingoptical circuit, burst, and packet switching techniques By exploiting contentionresolution in wavelength, time, and space domains, the optical-label switchingrouters can achieve high-throughput without resorting to a store-and-forwardmethod associated with large buffer requirements Testbed demonstrations in sup-port of multimedia and data communications applications are reviewed

optical-1.4.17 Chapter 17 Optical and electronic technologies

for packet switching (Rodney S Tucker)

The Internet provides most of the traffic and growth in lightwave systems today.Unlike the circuit-switched telephone network, the Internet is based on packet

switching, which provides statistical multiplexing of the many data streams thatpass through a given switch node A key element of any packet switch is theability to buffer (or store) packets temporarily to avoid collisions Since it is easy

to store bits as electronic charge in silicon, commercial packet switches areelectronic

This chapter introduces the basics of practical packet routers, indicating therequirements of a switch based on either electronics or photonics The authorcompares the physical limits on routers based on storing and switching electrons vsphotons While many in the optics field would like to see photonic switchesdominate, it appears that the physical inability to provide a large optical randomaccess buffer means that packet switches will continue to be optoelectronic ratherthan all-optical for some time

1.4.18 Chapter 18 Microwave-over-fiber systems

(Alwyn J Seeds)

The low-loss, wide-bandwidth capability of optical transmission systems makesthem attractive for the transmission and processing of microwave signals, whilethe development of high-capacity optical communication systems has required theimport of microwave techniques in optical transmitters and receivers These twostrands have led to the development of the research area of microwave photonics.Following a summary of the historic development of the field and the development

of microwave photonic devices, systems applications in telecommunications, andlikely areas for future development are discussed Among the applicationsreviewed are wireless-over-fiber access systems, broadband signal distributionand access systems, and communications antenna remoting systems

1.4.19 Chapter 19 Optical interconnection networks in

advanced computing systems (Keren Bergman)

High-performance computers today use multicore architectures involving multipleparallel processors interconnected to enhance the ultimate computational power.One of the current supercomputers contains over one hundred thousand individualPowerPC nodes Thus the challenges of computer design have shifted from

“computation-bound” to “communication-bound.” Photonic technology offersthe promise of drastically extending the communications bound by using theconcepts and technologies of lightwave communications for interconnecting andnetworking the individual multi-processor chips and memory elements Siliconphotonics (see the chapter on “Silicon Photonics” by C Gunn and T L Koch)looms as a possible subsystem technology having the cost-effectiveness of siliconCMOS manufacturing process

This chapter is a review of the subject of interconnection networks for performance computers Performance issues including latency, bandwidth andpower consumption are first presented, followed by discussions of design con-siderations including technology, topology, packet switching nodes, messagestructure and formation, performance analysis, and evaluation Network designimplementation, architectures, and system demonstrations are then covered andthoughts are offered on future directions, including optical interconnection net-works on a chip.

high-1.4.20 Chapter 20 Simulation tools for devices, systems,

and networks (Robert Scarmozzino)

The ability of optical fiber telecommunications to satisfy the enormous demand fornetwork capacity comes from thorough understanding of the physics and otherdisciplines underlying the technology, as well as the abilities to recognize sourcesfor limitation, develop ideas for solution, and predict, test, and demonstrate thoseideas The field of numerical modeling has already been an important facilitator inthis process, and its influence is expected to increase further as photonics matures.This chapter discusses the broad scope of numerical modeling and specificallydescribes three overarching topics: (i) active and passive device/component-level modeling with emphasis on physical behavior, (ii) transmission-system-level modeling to evaluate data integrity, and (iii) network-level modeling forevaluating capacity planning and network protocols The chapter offers an over-view of selected numerical algorithms available to simulate photonic devices,communication systems, and networks For each method, the mathematicalformulation is presented along with application examples

ACKNOWLEDGMENTS

We wish sincerely to thank Tim Pitts and Melanie Benson of Elsevier for theirgracious and invaluable support throughout the publishing process We are alsodeeply grateful to all the authors for their laudable efforts in submitting theirscholarly works of distinction Finally, we owe a debt of appreciation to the manypeople whose insightful suggestions were of great assistance

We hope our readers learn and enjoy from all the exciting chapters

of digital optical communication techniques fall into two main areas, distinguished

by the per-wavelength symbol rates they operate at:

• At per-channel symbol rates of 10 Gbaud, electronic signal processing ging from simple feed-forward and decision-feedback equalizers (FFE, DFE)

ran-at the receiver all the way to maximum-likelihood sequence estimran-ation(MLSE) is commercially available today [1–3], and FEC is found in mostcommercial 10-Gb/s transport products [4, 5] Digital signal predistortion atthe transmitter is being actively pursued [6], and coherent detection is

Optical Fiber Telecommunications V B: Systems and Networks

Copyright  2008, Elsevier Inc All rights reserved.

experiencing renewed1interest [14–16] In contrast to widely deployed directdetection, coherent detection allows electronic signal processing to make use ofthe full optical field (magnitude and phase), which can be exploited for moreefficient mitigation of fiber transmission impairments such as chromatic dis-persion (CD) or polarization mode dispersion (PMD) Coherent detectionresearch is also aiming at increasing the bit rate through multilevel modulationand polarization multiplexing while keeping the signal bandwidth similar tothat of a 10-Gb/s binary signal to reuse 10-Gb/s optical line infrastructure and

to increase the system’s spectral efficiency (SE) In general, narrow signalbandwidths are desired in high-SE optically routed networks to minimizeexcessive filtering penalties due to the concatenation of reconfigurable opticaladd/drop multiplexers (ROADMs) and to avoid excessive cross talk betweenadjacent wavelength-division multiplexed (WDM) channels

• At per-channel symbol rates of 40 Gbaud, and up to 100 Gbaud, the ities of electronic equalization are still limited to low complexity electronic[17–21] and optical [22–24] FFE structures In this area, modulation formatsand line coding are at the center of interest They are used to mitigate linearand nonlinear impairments of fiber-optic transmission, as well as to achievehigh SEs in optically routed networks At 40 Gb/s, FEC is a standard feature

capabil-of commercially deployed optical transport systems [25]

Despite the many benefits brought by electronic signal processing and digitalcommunication techniques, it should be noted that an optical fiber network differssubstantially from an RF communication link, both fundamentally and technolo-gically Some key differences relevant to the discussions in this chapter aresummarized in Table 2.1

Fiber Nonlinearities

The most important fundamental difference between optical and RF communications

is the presence of nonlinear distortions in the optical fiber communication channel.The high transversal confinement of the optical signal field in a fiber corewith effective areas between 20 and 110mm2 causes light intensities to reach orexceed a megawatt/cm2 At such high optical intensities, the fiber’s index of refraction

is affected by the presence of optical signals through the optical Kerr effect [26], andsignal-induced refractive index changes translate into changes of the signals’ opticalphases Over optically amplified distances of many hundred or even several thousand

1 Coherent detection using advanced optical modulation formats was widely discussed in the context

of unamplified lightwave systems in the 1980s [7–12], where attenuation-limited single-span sion required utmost receiver sensitivities With the advent of efficient optical amplifiers, allowing for comparable direct-detection receiver sensitivities [9, 13], coherent systems research decayed in the early 1990s.

transmis-kilometers, these phase rotations, in conjunction with fiber dispersion, result in a host

of different waveform distortions that increase with signal power As a consequence,and in stark contrast to classical RF systems, the performance of an optical commu-nication link exhibits a maximum at a certain signal power level, which represents theoptimum trade-off between optical amplifier noise (amplified spontaneous emission,ASE) and fiber Kerr nonlinearity As an example, Figure 2.1(a) [27] shows the

Table 2.1Optical vs radio-frequency communication systems

Fundamental

–103bits) Technological

Electrical

bandwidth

60 GHz – limited by technology (bandwidth/carrier10 4 to 106)

Limited by spectral regulations (bandwidth/carrier10 2 to

104)

demodulation

Processing power per

SNR (dB) Signal launch power (dBm)

(b) (a)

Signal launch power (dBm)

measured bit error ratio (BER) as a function of launched signal power for variousoptical modulation formats transmitted over 1980 km of standard single-mode fiber(SSMF) The figure reveals the trade-off between ASE-dominated performance

at low signal launch powers and dominating fiber nonlinearity at highpowers, indicating different performance optima at format-specific optimumlaunch powers Figure 2.1(b) shows the schematic behavior of fiber channelcapacity as a function of the received signal-to-noise ratio (SNR) [28] Due toquantum-mechanical lower bounds on optical amplifier noise, the SNR canonly be increased by launching higher signal power, which eventually leads tononlinear signal distortions Note that when operated at optimum performance,every fiber-optic communication system shows signs of nonlinear signaldistortions

Electronic Bandwidth Limitations

The second major difference between optical and RF communication systems,with both fundamental and technological consequences, is the high absolutebandwidth of optical communication signals In general, and as evidenced by thehistoric evolution of lightwave systems [29], the highest possible per-channel bitrates have always yielded the lowest cost, footprint, and power consumption perend-to-end networked information bit, once the underlying technologies weresufficiently mature Therefore, optical communication systems have always beenpushing the limits of high-speed electronic and optoelectronic components, with100-Gb/s binary transmission systems representing the current limit of electronicmultiplexing and demultiplexing capabilities (see Refs [24, 30] and Chapter 5 ofthis book) The sheer per-channel bandwidth has left little room for the imple-mentation of sophisticated signal processing algorithms, including coherent detec-tion with digital phase locking (“intradyne” detection [14, 15, 31], cf Sec 2.4.1and Chapter 3 of this book) The last 5 years, however, have seen a period ofstagnation due to the burst of the telecom bubble, which has allowed electronicprocessing to catch up with 10-Gb/s line rates, still the most dominantly deployedper-channel bit rate It remains to be seen whether future systems will be based on10-Gbaud technology and its evolution to higher rates, with substantial digitalprocessing complexity, or whether systems will continue to use the highest possi-ble symbol rates in conjunction with direct detection and advanced optical tech-nologies such as adaptive dispersion management or optical phase conjugation butwith modest electronic processing In any case, optical communication technolo-gies will always be facing the limits of high-speed signal processing and modula-tion, which is an important factor to take into account when discussing advancedoptical modulation formats

In this chapter, we review, on an advanced level, important communicationengineering techniques that have been discussed in the context of optical

networking over the past 5 years and that are likely to be the subject of furtherstudies over the next 5 years We also want to point the reader to additionalreviews on the topic [27, 32–36] On a more introductory level, optical modulationformats are discussed in, e.g., [37–39] The communication engineering back-ground is covered in textbooks such as [40–42], and the background in opticalphysics can be found in, e.g., [26, 43, 44] A list of acronyms used in this chaptercan be found on pages 80-81.

2.2 COMMUNICATION ENGINEERING AND OPTICAL COMMUNICATIONS

The broad field of communication engineering has developed a variety oftechniques that can be used to beneficially impact digital data transmission and

to approach the capacity [45, 28] of a given communication channel Figure 2.2visualizes those methods and ingredients that are relevant to the understanding ofadvanced optical communication systems, and in particular to the understanding of

Detection and equalization

Maximum-likelihood sequence estimation (MLSE)

Multi-symbol phase estimation

Feed forward equalization (FFE)

Decision feedback equalization (DFE)

Optical equalization (OEQ)

Line coding

Without data correlation

With data correlation

Carrier-suppressed return-to-zero (CSRZ)

Duobinary (DB, PSBT)

Alternate mark inversion (AMI)

Error correcting coding

Forward error correction

Wavelength (Frequency)

Wavelength-division multiplexing (WDM) Coherent WDM (CoWDM) Orthogonal frequency division multiplexing (OFDM)

Network aspects

ROADM support Optical routing

Modulation

Modulation waveform Symbol constellation

[Non] return-to-zero ([N]RZ) Chirped modulation (C-NRZ, CRZ) Vestigial sideband (VSB)

Intensity/Phase (ASK, OOK, PSK) Frequency (FSK)

Polarization (PolSK) Time (PPM)

Binary, Multi-level Orthogonal or linearly dependent

Signal space

Figure 2.2 Overview of communication engineering techniques discussed in optical networks research today.