Table of contents for Passive optical networks : principles and practice / Cedric F. Lam.

Bibliographic record and links to related information available from the Library of Congress catalog.

Note: Contents data are machine generated based on pre-publication provided by the publisher. Contents may have variations from the printed book or be incomplete or contain other coding.

List of Acronyms
List of Figures xiii
List of Tables xxv
Preface xxvii
Acknowledgments xxix
1 Introduction 1
Cedric F. Lam
1.1 History of Broadband Access Networks and PON 1
1.1.1 Digital Subscriber Line (DSL) 2
1.1.2 Cable Modem 3
1.1.3 Fiber Access Systems 7
1.1.4 Ethernet 8
1.1.5 WDM in Optical Access Networks 11
1.1.6 Killer Applications 11
1.2 Economic Considerations in PON Development 12
1.2.1 How Much Bandwidth is Enough? 12
1.2.2 Policy and Regulation Influence 13
1.2.3 Standardization Efforts 14
1.2.4 Cost Considerations 14
1.3 Organization of the Book 15
References 16
2 PON Architectures Review 19
Cedric F. Lam
2.1 FTTx Overview 19
2.2 TDM PON vs WDM-PON 20
2.3 Optical Transmission System 21
2.3.1 Optical Fiber 21
2.3.2 Chromatic Dispersion 23
2.3.3 Fiber Loss 25
2.3.4 Bidirectional Transmission 26
2.3.5 Wavelength Division Duplex 27
2.4 Power-Splitting Strategies in a TDM-PON 29
2.4.1 Splitting Architectures 29
2.4.2 Splitting Ratio 30
2.5 Standard Commercial TMD-PON Infrastructure 30
2.5.1 OLT and ONU Structures 31
2.5.2 Burst-Mode Operation and Ranging 33
2.5.3 C-Band Analog CATV Signal Overlay 35
2.5.4 Security Concerns in Power-Splitting PON 36
2.6 APON/BPON and G-PON 37
2.6.1 ATM-PON and ITU-T G.983 37
2.6.2 Collision Resolution in APON/G-PON 42
2.6.3 Wavelength Overlay in APON/G-PON 42
2.6.4 G-PON and ITU-T G.984 43
2.6.5 APON/G-PON Protection Switching 54
2.7 EPON 54
2.7.1 Ethernet Layering Architecture and EPON 54
2.7.2 EPON PMD Layer 56
2.7.3 Burst-Mode Operation and Loop Timing in EPON 56
2.7.4 PCS Layer and Forward Error Correction 59
2.7.5 Ethernet Framing 59
2.7.6 Multipoint Control Protocol (MPCP) 60
2.7.7 Point-to-Point Emulation in EPON 65
2.7.8 EPON Encryption and Protection 67
2.7.9 Ethernet OAM Sublayer 68
2.8 G-PON and EPON Comparison 69
2.9 Super PON 70
2.10 WDM-PON 71
2.10.1 Advantages and Challenges of WDM-PON 72
2.10.2 Arrayed Waveguide Grating (AWG) Router 72
2.10.3 Broadcast Emulation and Point-to-Point Operation 74
2.10.4 2-PONs-in-1 74
2.10.5 WDM-on-WDM 75
2.10.6 Hybrid WDM/TDM PON 76
2.10.7 Multiply Fiber Plant Utility by WDM 80
2.11 Summary 84
References 84
3 Optical Technologies in PON 87
Elaine Wong
3.1 Introduction 87
3.2 Optical Technologies in Passive Outside Plant 91
3.2.1 Planar Lightwave Circuit (PLC)-Based Optical
Power Splitter 91
3.3 Arrayed Waveguide Gratings 93
3.3.1 Introduction 93
3.3.2 Athermal Arrayed Waveguide Grating 99
3.4 PON technologies for indoor installation 108
3.4.1 Field Assembly and Indoor Connectors 108
3.4.2 Fiber for Indoor Installations 112
3.5 Transmitters Sources at Subscriber Premises 120
3.5.1 Introduction 120
3.5.2 Wavelength-Specific ONUs 122
3.5.3 Colorless ONUs 129
3.5.4 Source-free ONUs based on wavelength reuse
schemes 137
3.6 Summary 141
References 143
4 Transceivers for Passive Optical Networks 151
Yongmao Frank Chang and Badri Gomatam
4.1 Introduction 151
4.1.1 Historical background 152
4.1.2 PON transceiver evolution 153
4.2 PON system requirements 154
4.2.1 PON system power budgets and specifications 155
4.2.2 Physical-layer specifications 156
4.2.3 PON burst-mode timing requirements 157
4.3 Transceiver technologies 161
4.3.1 Transceiver building blocks 162
4.3.2 Optical transmit and receive devices 163
4.3.3 Bi-directional Optical Subassembly (BOSA) 167
4.3.4 PON transceiver modules 172
4.4 Burst-mode electronics 174
4.4.1 Conventional versus burst-mode data 174
4.4.2 Burst-Mode Transmitter (BM Tx) 175
4.4.3 Burst-Mode Receiver (BM Rx) 181
4.5 Transceiver digital-diagnostic monitoring 187
4.5.1 Module parameter monitoring 188
4.5.2 Fiber OTDR monitoring 192
4.6 PON transceiver system evaluation 196
4.6.1 G-PON transceiver system evaluation 197
4.6.2 EPON transceivers 202
4.6.3 Impact of analog CATV overlay 205
4.7 Summary and Outlook 207
References 209
5 Ranging and Dynamic Bandwidth Allocation 215
Noriki Miki and Kiyomi Kumozaki
5.1 Ranging 215
5.1.1 Purpose of Ranging 215
5.1.2 Ranging Procedures Overview 216
5.1.3 Ranging Protocol of G-PON 218
5.1.4 Ranging Protocol of EPON 220
5.2 Dynamic Bandwidth Allocation (DBA) 221
5.2.1 DBA Overview 221
5.2.2 Target Service 222
5.2.3 Requirements of DBA 224
5.2.4 Traffic Control Fundamentals - Fair Queuing 226
5.2.5 IPACT and its variants 228
5.2.6 Improved DBA 231
References 241
6 Protection Switching in PON 243
Calvin C K Chan
6.1 Introduction 243
6.2 Considerations of Protection in Passive Optical Networks 244
6.2.1 Protection or Restoration 245
6.2.2 Network Topology 245
6.2.3 Network Type 246
6.2.4 Resources to be protected 246
6.2.5 Single or Multiple failures 247
6.2.6 Automatic protection switching 247
6.2.7 Operation, administration, and management 247
6.2.8 Traffic Restoration Time 247
6.2.9 Complexity 248
6.3 Protection Architectures 248
6.3.1 Conventional Passive Optical Networks 248
6.3.2 WDM Passive Optical Networks 253
6.4 Summary 261
Acknowledgements 264
References 264
7 Optical Diagnosis, Performance Monitoring, and Characterization
for PON 267
Alan E. Willner and Zhongqi Pan
7.1 Introduction 267
7.2 Network Testing, Characterization, and Monitoring
Challenges for PON 269
7.2.1 Fiber's Key Characteristics 269
7.2.2 Characterization and Monitoring Challenges for PON 271
7.3 Methods for Characterization, Diagnosis, and Monitoring 275
7.3.1 Required Physical Layer Measurement 275
7.3.2 Basic Test Equipment 275
7.3.3 Network Testing, Characterization, and Diagnosis
Guidelines 277
7.3.4 Network Performance Monitoring 278
7.4 Conclusion 296
References 296
Appendix I G-PON PMD Characteristics 301
Appendix II EPON MPCPDU Formats 311
List of Figures
Figure 1.1 Development trend of Ethernet technologies. 9
Figure 2.1 Generic structure of a modern telecommunication
network. 20
Figure 2.2 FTTx alternatives (from reference [1]). 21
Figure 2.3 Architecture of (a) TDM-PON and (b) WDM-PON. 22
Figure 2.4 Single-mode fiber (SMF) vs multi-mode fiber (MMF). 23
Figure 2.5 Dispersion coefficient as a function of wavelengths for
various types of optical fibers. 24
Figure 2.6 Loss in an optical fiber at different wavelengths. 25
Figure 2.7 (a) One-fiber single-wavelength bi-directional
transmission. (b) Near End Cross-talk (NEXT) 27
Figure 2.8 Wavelength division duplex uses 1:3=1:5-mm coarse
WDM coupler (diplexer) to separate upstream and
downstream signals. 28
Figure 2.9 Splitting strategies in a TDM-PON: (a) one-stage
splitting, (b) multistage splitting, and (c) optical bus. 29
Figure 2.10 Standard commercial TMD-PON architecture. 31
Figure 2.11 Generic structure of a standard TDM-PON OLT. This
diagram represents a chassis with multiple OLT cards,
which are interconnected through a back plane switch.
Each OLT card with its own MAC and PMD layer
servers a separate PON. 32
Figure 2.12 Generic structure of a standard TDM-PON ONU. 33
Figure 2.13 Overlaying analog broadcast TV services on a
TDM-PON using 1:55-mm wavelength. 36
Figure 2.14 Downstream and upstream APON frame formats at
155.52-Mbps speed. For 622.08-Mbps and 1244.16-Mbps
speed, the numbers of times lots are simply multiplied
by 4 and 8 to the numbers shown in the above diagrams. 39
Figure 2.15 ATM switching examples, (a) VP and VC switching and
(b) VC switching. 41
Figure 2.16 Wavelength allocation plan in ITU-T G.983.3. 43
Figure 2.17 A T-CONT represent a logical link between the OLT
and an ONU. 45
Figure 2.18 ATM-based T-CONT vs GEM-based T-CONT. 46
Figure 2.19 Protocol stack of the G-PON transmission convergence
(GTC) layer (from [24]). 46
Figure 2.20 GTC downstream signal consists of 125-ms frames
with a PCBd header and a payload section. 48
Figure 2.21 GTC downstream frame and media access control
concept (from [24]). 48
Figure 2.22 GTC downstream frame formats. 49
Figure 2.23 GTC upstream framing: each ONT starts the upstream
transmission with PLOu. An ONT assigned with two
Alloc-ID in two consecutive upstream allocations only
need to transmit PLOu once. 50
Figure 2.24 GTC upstream frame format. 51
Figure 2.25 GEM frames in upstream payload. 52
Figure 2.26 GEM encapsulation formats. 52
Figure 2.27 Mapping and fragmentation of user data frames into
GEM payload (from [24]). 53
Figure 2.28 Multiplexing of urgent data using GEM fragmentation
process (from [24]). 54
Figure 2.29 Point-to-point (P2P) Ethernet and point-to-multipoint
(P2MP) EPON layering architecture. 55
Figure 2.30 Standard Ethernet frame format. 60
Figure 2.31 EPON ranging process. 61
Figure 2.32 EPON Gate operation. 62
Figure 2.33 EPON Report operation. 63
Figure 2.34 Generic format of MPCPDU. 63
Figure 2.35 Autodiscovery process (from [9]). 64
Figure 2.36 Although all the ONU traffic arrives at the same
physical port at the OLT, because of the directional
power-splitting coupler used at the remote node,
ONUs cannot see each other's traffic without the
forwarding aid of OLT. 66
Figure 2.37 Point-to-point emulation in EPON. 66
Figure 2.38 Modified preamble with LLID for point-to-point
emulation in EPON. 67
Figure 2.39 EPON point-to-point emulation operation. 68
Figure 2.40 Point-to-point and single-copy broadcast (SCB) MACs
in an EPON model. 68
Figure 2.41 Conventional WDM coupler vs arrayed waveguide
grating. 73
Figure 2.42 Emulation of broadcast services on a WDM-PON
with a broadband source. 74
Figure 2.43 A modified AWG for 2 PONs in 1. 75
Figure 2.44 By using CWDM devices to combine and separate
optical signals in multiple FSRs of an AWG devices,
a highly flexible WDM-on-WDM system can
be achieved. 76
Figure 2.45 Split-net test-bed demonstrating the WDM-on-WDM
concept (from reference [50]). 77
Figure 2.46 Demonstration of a hybrid WDM/TDM-PON with
wavelength-selection-free transmitters, (a) downstream
link and (b) upstream link (from reference [52]). 78
Figure 2.47 WE-PON optical layer block diagram. (Courtesy of
ETRI and Korea Telecom) 79
Figure 2.48 RSOA module used in ETRI WE-PON prototype.
(Courtesy of ETRI and Korea Telecom) 80
Figure 2.49 Feed-forward current injection to RSOA in WE-PON.
(Courtesy of ETRI and Korea Telecom) 81
Figure 2.50 PLC-ECL used in WE-PON prototype. (Courtesy of
ETRI and Korea Telecom) 82
Figure 2.51 A band sorter interleaves the optical spectrum into
bands (top) which can be further separated with
diplexers (bottom) into bands for upstream and
downstream PON connections. 83
Figure 2.52 Overlaying APON and EPON on the same fiber
plant using ''band sorters.'' 83
Figure 3.1 A PLC splitter is manufactured using two fiber arrays
and one PLC chip all aligned within one package.
(From Ref [1].) 91
Figure 3.2 Measured insertion loss of a 1_32 PLC splitter as a
function of wavelength, showing uniform insertion
loss (#1.25-dB variation) and low excess insertion less
(#1.25 dB) above the theoretical value. (From Ref [3].) 92
Figure 3.3 Schematic illustration of arrayed waveguide grating
comprising two free-propagating regions and a
waveguide array. (From Ref [7].) 94
Figure 3.4 Schematic layout of N_N arrayed waveguide grating
router. (From Ref [83].) 95
Figure 3.5 Schematic illustration of the cyclic property of an
arrayed waveguide grating router. (From Ref [83].) 95
Figure 3.6 Schematic illustration of dual-function reflective
arrayed waveguide grating. (From Ref [84].) 96
Figure 3.7 (a) Fiber-to-fiber insertion loss of 1_14 power splitter
function; (b) transmission spectra of 14-channel
wavelength-router function. (From Ref [84].) 97
Figure 3.8 Schematic illustration of 1_8 two-PONs-in-one (2P1)
device. For clarity, the vertical scale has been expanded
by four times. (From Ref [85].) 98
Figure 3.9 Spectra from the output ports of a 1_ 8 P21 device
measured around the (a) 1:3-mm waveband and
(b) 1:55-mm waveband. (From Ref [85].) 98
Figure 3.10 Schematic diagram of athermal AWG with silicone
resin-filled triangular groove. (Adapted from Ref [12].) 100
Figure 3.11 Temperature-dependent transmission spectrum of
Channel 4 over the temperature range of 0-85 8C
for (a) conventional silica-based AWG and (b)
athermal AWG with silicon resin-filled triangular
groove. (From Ref[13].) 100
Figure 3.12 Temperature-dependent wavelength shift of Channel 4
in the temperature range of 0-85 8C. (From Ref [13].) 101
Figure 3.13 Reduction of excess loss by replacing triangular groove
of wider width with several grooves of smaller widths.
(From Ref [14].) 101
Figure 3.14 Schematic diagram of modified low-loss groove and
array structure for a 1.5%-D athermal AWG. (From
Ref [KAMEI].) 102
Figure 3.15 Schematic diagram of silicon resin-filled trenches in the
slab region of 1.5%-D athermal AWG. (From Ref [17].) 102
Figure 3.16 Schematic diagram of silicon resin-filled trenches with
spot size converters based on segmented core in the slab
region of 2.5%-D athermal AWG. (From Ref [16].) 103
Figure 3.17 (a) Cross-sectional view of silica/polymer hybrid
waveguide; (b) comparison of measured TE/TM
polarization shift of the central wavelength, Channel 9,
with polymer and silica overcladding; (c) measured
temperature-dependent wavelength shift of Channel 9
with polymer and silica cladding. (Adapted from
Ref [19].) 104
Figure 3.18 All-polymer 8_8AWG with 200 GHz channel spacing:
(a) measured spectra; (b) measured TE/TM polarization
shift of Channel 4; (c) measured temperature-dependent
transmission spectrum of Channel 4 for the temperature
range of 25-65 8C. (Adapted from Ref [20].) 105
Figure 3.19 Schematic diagram of an athermal AWG with
temperature compensating plate. (From Ref [21].) 105
Figure 3.20 Optical spectra of the athermal AWG with temperature
compensating plate shown in Fig. 3.19. (From Ref [21].) 106
Figure 3.21 Schematic diagram of an athermal AWG chip with
temperature-compensating rod. (From Ref [22].) 106
Figure 3.22 (a) Optical spectra in C- and L-bands of athermal
AWG using temperature-compensating rod; (b)
temperature-dependent wavelength shift of four
channels in the C-band from _30 8C to +70 8C.
(Adapted from Ref [22].) 107
Figure 3.23 Schematic diagram of AWG with bimetal plate.
(From Ref [23].) 108
Figure 3.24 Center wavelength of AWG as a function of
temperature. (From Ref [23].) 109
Figure 3.25 Field assembly (FA) connector (a) structure of plug
and socket, (b) in optical cabinet at customer premises.
(From Refs [29] and [30].) 110
Figure 3.26 (a) Connection loss trials, and (b) return loss trials.
(From Ref [29].) 111
Figure 3.27 Bendable SC connector. (From Ref [30].) 111
Figure 3.28 Cross-sectional view of angled physical contact (APC)
connector. The 8-degree fiber end face directs reflected
light into the cladding. (From Ref [94].) 112
Figure 3.29 (a) Bend loss characteristics of bend resistant fiber;
(b) comparison of conventional optical fiber with
minimum bending radius of 30 mm and bend-resistant
with minimum bending radius of 15 mm.
(From Ref [29].) 113
Figure 3.30 (a) Typical spectral attenuation of PureAccess-Ultra;
(b) bending loss properties as a function of wavelength.
(From Ref [32].) 114
Figure 3.31 Schematic diagram of optical curl cord. (From Ref [34].) 115
Figure 3.32 Connection of laptop terminal to optical outlet in the
wall using optical curl cord. (From Ref [34].) 115
Figure 3.33 Cross section and refractive index profile of
hole-assisted fiber. (From Ref [33].) 116
Figure 3.34 Bending loss characteristics of hole assisted fibers:
(a) calculated bending loss performance as a function
of normalized hole distance; (b) wavelength dependence
of measured bending loss. (From Ref [33].) 117
Figure 3.35 (a) Calculated and experimental mode field diameter
mismatch as a function of hole arrangement; (b) splice
loss induced by MFD mismatch as a function of
refractive index difference between Nhole and Nclad.
(From Ref [33].) 118
Figure 3.36 Comparison of three GI-POFs with differing core
diameter and numerical aperture (NA): (a) refractive
index profile; and (b) bending loss as a function of
radius. (From Ref [42].) 120
Figure 3.37 (a) Core diameter and (b) NA dependence loss at
bending radius of 5 mm. (From Ref [42].) 121
Figure 3.38 (a) Bending loss of GI POF with a core
diameter 1/4 200 mm and NA 1/4 0.24; (b) optimum
waveguide parameters of GI POF to suppress the
bending loss. (From Ref [42].) 122
Figure 3.39 Schematic diagram of a index-guided Fabry-Perot
laser diode. Optical spectra showing multiple
longitudinal lasing modes. 123
Figure 3.40 Schematic diagram of a distributed feedback laser.
Optical spectrum showing excellent single-mode
behavior. 124
Figure 3.41 Schematic diagram of a vertical cavity surface-emitting
laser (VCSEL) with circular output beam. 125
Figure 3.42 WDM-PON using directly modulated OIL-VCSELs
as ONU transmitters. Each slave VCSEL, located at
the ONU is injection-locked by modulated downstream
signal transmitted from a master DFB laser located
at the CO. (From Ref [49].) 126
Figure 3.43 Upstream BER curves for different values of
(a) upstream signals to Rayleigh backscattering ratio
(SRR) at a constant downstream extinction ratio
(ER) 1/4 4.5 dB; (b) downstream ER at a constant
SRR 1/4 13.4 dB. 127
Figure 3.44 Received optical power at BER 1/4 10_9 as a function
of downstream extinction ratio. Solid line:
bidirectional 25-km transmission with Rayleigh
backscattering effects. Dash line: unidirectional
back-to-back transmission. 128
Figure 3.45 Schematic diagram of WDM-PON with broadband
optical source (LED) in ONU. (From Ref [55].) 130
Figure 3.46 Experimental setup of WDM-PON with centralized
supercontinuum broadband light source.
(From Ref [60].) 131
Figure 3.47 Schematic diagram of colorless bidirectional (BiDi)
transceivers based on FP-LDs injection-locked by
the spectrally sliced light from a broadband
superluminesence LED (SLED). (From Ref [65].) 132
Figure 3.48 Optical spectra of total injection light and spectrumsliced
injection lights measured a the input and output
ports of the (a) remote node AWG, and (b) central
office AWG. (From Ref [65].) 133
Figure 3.49 Schematic diagram of colorless reflective
semiconductor optical amplifier (RSOA) wavelength
seeded by spectrally sliced light from a broadband
superluminesence LED (SLED). (From Ref [70].) 134
Figure 3.50 Monolithic indium phosphide-based RSOA features
a curved waveguide architecture. (From Ref [89].) 135
Figure 3.51 WDM-PON with self-injection-locked Fabry-Perot
laser diodes. (From Ref [70].) 136
Figure 3.52 Optical spectra of self-injection locked FP-LD.
(From Ref [70].)
Figure 3.53 Schematic diagram of WDM-PON with self-seeding
reflective SOAs. (From Ref [72].) 137
Figure 3.54 Overlapped optical spectra of 1.25 Gbps self-seeded
upstream channels. (From Ref [72].) 138
Figure 3.55 RITE-net architecture. The CW portion of each
downstream packet is remodulated with upstream
data at the optical network unit. (From Ref [73].) 138
Figure 3.56 WDM-PON using reflective semiconductor optical
amplifiers to amplify and remodulate downstream
wavelengths carrying downstream signals
(ASK format) for upstream transmission.
(From Ref [77].) 139
Figure 3.57 Optical spectrum of adiabatic-chirped DFB laser souce
directly modulated with 1.25 Gbps NRZ signal.
Superimposed on the spectrum is the passband
characteristic of negatively detuned (minus;18.7 GHz)
demultiplexer. (From Ref [79].) 140
Figure 3.58 Wavelength reuse scheme for WDM PON using
multichannel fiber Bragg grating to filter out
downstream carriers for remodulation with upstream
data. (From Ref [80].) 141
Figure 4.1 The general PON network architecture for the FTTx
scenario. 152
Figure 4.2 Physical layer burst-mode timing definition. TON _ Tx
turn-on time; TOFF _ Tx turn-off time; TDSR _ Rx
dynamic sensitivity recovery time; TLR _ Rx level
recovery time; TCR _ Rx clock recovery time,
TDL _ Rx delimiter time. 157
Figure 4.3 Various PON diplexer and triplexer transceivers for
optical access systems. 161
Figure 4.4 Feature function blocks of optical transceivers
showing both downstream and upstream directions.
In the upstream direction, an ONU transmitter
consists of a burst-mode laser driver (BM-LD),
and an F-P laser in the form of transmit optical
subassembly (TOSA); the OLT receiver consists
of a PIN or avalanche photodiode (PIN or APD),
BM-transimpedance (BM-TIA) ROSA (receive optical
subassembly), a limiting post-amplifier (post-amp) and
a burst-mode clock/data recovery (BM-CDR) unit. 162
Figure 4.5 Typical L-I curves of (a) DFB-LD, and (b) F-P-LD. 164
Figure 4.6 APD V-I curve and multiplication factor. 165
Figure 4.7 Schematics structures of TO-CAN packaging for DFB
(a) and APD (b). Their photographs are shown on
the right. 166
Figure 4.8 Transmission spectrum characteristic (left) and
photograph (right) of the WDM filter formed on a
fiber facet [29]. The two curves represent two different
light polarizations. 166
Figure 4.9 BOSA schematics for three wavelengths based on
thin-film technology. 168
Figure 4.10 The BOSA in TO-CAN package with driving
circuit [32]. 169
Figure 4.11 Optical circuit configurations with external-filters. 170
Figure 4.12 The feasibility of grating assisted WDM filter
on a PLC platform (top) and conceptual
schematic (bottom) of a triplexer using
this structure [43]. 171
Figure 4.13 Bidirectional ONU diplexer transceiver module. 173
Figure 4.14 Bidirectional ONU triplexer transceiver module. 173
Figure 4.15 Data formats in digital communication: (a) continuousmode
data (b) burst-mode data, (c) Burst Packet
data [47]. 175
Figure 4.16 Temperature characteristics of a typical F-P laser as
L-I curves. 176
Figure 4.17 Block diagram of a typical burst-mode laser driver
IC. [49] 177
Figure 4.18 Comparison of (a) continuous and (b) burst-mode laser
driver stage. 178
Figure 4.19 Block diagrams of two typical APC circuits. 179
Figure 4.20 Optical and timing performance of an EPON BM Tx:
(top left) Optical eye diagram using 4th-order Thompson
filter at 1.25 Gbps; numbers inside box are GbE eye-mask
margin. (Bottom left) Optical bursty signal with Laser
bias on/off. The measured laser burst-on and
burst-off times are shown on the right side [22]. 180
Figure 4.21 Optical eye diagrams at _40 8C and 80 8C case
temperature with optical power above 0 dBm and
ER above 10 dB. The numbers inside box are
GbE eye-mask margin. 181
Figure 4.22 Feed-forward and feedback implementations of optical
burst-mode receivers. 182
Figure 4.23 Block diagram of a burst-mode preamplifier IC
employing ATC (left) and the response of ATC circuit
(right). [56] 183
Figure 4.24 The comparison of (a) a conventional AGC, and (b) a
cell-AGC for burst-mode inputs with low extinction
ratio. [60] 184
Figure 4.25 Configuration and operation principles of cell-AGC
based preamplifier IC. 185
Figure 4.26 Comparison of AC-coupled (left) vs DC-coupled (right)
burst-mode receivers. 186
Figure 4.27 An example of 1.25 Gbps BM APD/TIA output
eye diagram (left) and settling time (right)
measurements [22]. 187
Figure 4.28 Examining temperature error at different ADC
resolutions. The y-axis shows the absolute value of the
maximum error. The overall error for the temperature
monitoring is +y 8C. The blue, pink, and green
lines represent the 8-, 10-, and 12-bit ADCs, respectively.
The solid lines represent the cases in which the accuracy
level of both the temperature sensor and the voltage
reference is within +1%, and the dashed lines
represent a +0.5% accuracy level. The red solid line is a
10-bit ADC without oversampling [70]. 189
Figure 4.29 A Laser output power and monitoring current versus
driving current. 191
Figure 4.30 A tunable OTDR example for in-service monitoring
of the fiber fault in WDM-PON [74]. 192
Figure 4.31 Fiber fault detection results for fault locations at
(a) 3 km, (b) 3.4 km, (c) 4 km, and (d) 5.2 km from
the remote node using the setup from Fig 4.30. 194
Figure 4.32 Embedded OTDR into ONU burst-mode transmitter.
OTDR analog front-end bandwidth is limited to
5 MHz [76]. 194
Figure 4.33 ONU block diagram with OTDR functionality
integrated [77]. 195
Figure 4.34 Typical burst-mode test setup configuration consisting
of two ONUs. A weak packet from BM-Tx as
ONU#1 followed by a strong packet from commercial
SFF module Tx as ONU#2 emulates the worst-case
condition [22]. 198
Figure 4.35 A typical G-PON upstream burst-mode test pattern
when strong packets from BM-Tx as ONU#1 are
followed by the weak packets from ONU#2. The
overhead guard time consists of the mandatory 32 bits,
preamble 44 bits, and delimiter time 16-20 bits. 199
Figure 4.36 Typical measurements of burst-mode BER (top) and
sensitivity penalty (bottom) performed for a DC-coupled
G-PON OLT Rx at 1.244 Gbps. 200
Figure 4.37 1.25 Gbps GP-ON uplink performance with PLM. 201
Figure 4.38 (a) Measured continuous-mode (P2P) versus burst-mode
BER. (b) Input and output for burst-mode receiver. 203
Figure 4.39 Burst-mode BER curves for the weak packet as a
function of the power levels of the strong packet at
1.25 Gbps, indicating the degradation of OLT Rx
sensitivity due to the influence of the strong packet. 204
Figure 4.40 (a) Impact of CATV signal on BER of the digital
receiver, (b) Impact of digital signal cross talk on the
CNR of CATV receiver. 206
Figure 5.1 Time division multiple access. 216
Figure 5.2 Ranging window. 217
Figure 5.3 G-PON ranging phase 1: serial number process. 219
Figure 5.4 GPON ranging phase 2: delay measurements. 220
Figure 5.5 Dynamic bandwidth allocation. 222
Figure 5.6 An example network configuration which offers
Layer-2 link services to ISPs. 224
Figure 5.7 The principle of Fair Queuing. 226
Figure 5.8 Operation of limited service IPACT: (a) when only
ONU#3 has upstream traffic; and (b) when all ONUs
have upstream traffic. Here, R1, R2, and R3 represent
request messages from each ONU respectively; data#3
represents user data transmitted from ONU#3, G3
represents grant for ONU#3. In (a), the value of R1 and
R2 are equal to 0 and the value of R3 indicates amount
of upstream user data stored in ONU#3. 230
Figure 5.9 Deficit round-robin (DRR) scheduling. 234
Figure 5.10 Maintaining short round-robin cycle using multiple
queue request. 235
Figure 5.11 An example with two types of requests. 235
Figure 5.12 An IEEE802.3ah REPORT frame with multiple
requests. 236
Figure 5.13 Burst allocation example of DRR using multiple
queue report set. 237
Figure 5.14 Comparison of bandwidth efficiency (Emax) among
the three DBAs covered in this chapter. Here we have
chosen the number of ONU N 1/4 32, Burst overhead
1/41:4msec,Max RTD 1/4100msec, wi 1/4w,
Thi 1/4w=N,OLT processing time1/4 16msec,
and ONU processing time1/416msec. 241
Figure 6.1 Protection switching architectures suggested by
ITU-T G.983.1. 249
Figure 6.2 A 1:N protection scheme at OLT [14]. LT: Line
Terminal. 250
Figure 6.3 A PON architecture with protection of distribution
fibers by interconnecting ONUs with protection
switching [17]. OSW: optical switch, FBG: fiber Bragg
gratings, ld: downstream wavelength, lu: upstream
wavelength. 251
Figure 6.4 A PON architecture with protection of both feeder
and distribution fibers by an additional loopback
distribution fiber with protection switching [18]. ld:
downstream wavelength, lu: upstream wavelength. 252
Figure 6.5 Protection of feeder fibers in two PONs using
CWDM technology [19]. OSW: optical switch, ld:
downstream wavelength, lu: upstream wavelength. 252
Figure 6.6 An example of a four-fiber shared rings in (a)
normal operation; (b) span switching; and (c) ring
switching [25]. 254
Figure 6.7 A modified star-ring PON with protection capability [26].
Inset shows the structure of the RN to illustrate the
protection mode. OS: optical switch. 255
Figure 6.8 The structure of (a) the OLT and (b) RN of a
WDM-PON with duplicated fiber feeders for
protection [27]. AN: access node, CN: central node,
OSC: optical supervisory channel, PD: photodiode,
OSW: optical switch. 256
Figure 6.9 Self-protected network architecture for WDM-PON.
LD1-4: laser diode; PD1-4: PIN photodiode;
WC: WDM coupler; M1&M2: optical power monitors;
{Ai, Ci} for i2 {1, . . .N}: upstream wavelengths; {Bi, Di}
for i 2 {1, . . .N}: downstream wavelengths. Inset shows
the spectral response of one of the output ports of the
AWG. FSR: free-spectral range of the AWG. [28]. 258
Figure 6.10 (a) A WDM-PON survivable architecture with eight
ONUs and centralized protection switching control;
(b) OLT configuration under normal operation;
(c) wavelength assignment plan. B/R: blue/red filter;
OC: 3-dB fiber coupler; LD: laser diode; PD:
photodiode. Note: FSR1 stands for free-spectral range
of the N _ 2 (N 1/4 8) AWG at the OLT; while FSR2
stands for that of both AWG1 and AWG2 at the RN.
The wavelengths quoted in boxes are the working
upstream wavelengths. The wavelengths in blue band
are underlined but those in red band are not [31]. 259
Figure 6.11 A WDM-PON architecture with self-protection
capability against both feeder fiber and distribution
fiber failures [32]. WC: wavelength coupler, B/R:
blue-red filter, OS: optical switch, M: power
monitoring modules. 260
Figure 6.12 (a) A single-fiber CWDM optical access ring network;
(b)&(c) the structure of AN2 in (b) normal state;
(c) protection state when there was a fiber cut between
AN2 and AN3 [45]. AN: access node, Tx: transmitter,
Rx: receiver. 262
Figure 6.13 (a) A protected optical star-shaped ring network;
(b) lightpath diagram, dotted lines were the designated
protection paths, (c) the protection lightpath was
adopted when node 1 failed [47]. FBG: fiber
Bragg gratings. 263
Figure 7.1 The architecture of a passive optical-access network. 268
Figure 7.2 Points where the test required for a WDM, point-tomultipoint,
and bidirectional PON. 272
Figure 7.3 Three layers of optical performance monitoring:
transport monitoring, signal quality monitoring, and
protocol monitoring. 279
Figure 7.4 Example on the detection and localization of fiber
failures in a bidirectional WDM-PON [15]. 283
Figure 7.5 Fiber-link loss monitoring example in a
bidirectional WDM-PON using ASE-injected
FP-LD [17]. 284
Figure 7.6 Schematic of an OWDR-embedded FTTH network
from OLT to ONT [18]. 285
Figure 7.7 Optical OSNR monitoring challenges: to discriminate
the noise power within the individual channel optical
bandwidth. The noise power level adjacent to the
channel may not be as same as the noise power
level within the channel. 287
Figure 7.8 Operating principle of the polarization technique: the
polarized noise (i.e. half of the total noises) can be
measured by using the second linear polarizer,
which is aligned to be orthogonal to the signal's
polarization [25]. 287
Figure 7.9 RF pilot tone added to the channel bandwidth as
the signal quality/degradation monitor. 288
Figure 7.10 (a) Typical measured data for the logarithm of the BER
versus the decision threshold [37]; (b) the BER as a
function of the received optical SNR [41]. 290
Figure 7.11 OCDM-based PON-monitoring system where every
network leg is assigned an encoder. One tunable
decoder is employed at the CO [42]. 291
Figure 7.12 System configuration of optical fiber line testing system
using a 1650-nm testing window [43]. 292
Figure 7.13 Conceptual diagram for monitoring chromatic
dispersion using optical vestigial-sideband (VSB)
filtering: the recovered bits from either part of the
spectrum arrive at slightly different times depending
on the chromatic dispersion [50]. 294
Figure 7.14 The use of the degree of polarization (DOP) to monitor
the effects of PMD: (a) DOP measurements as a
function of instantaneous DGD. Note that the DOP
is pulse-width-dependent [52]; and (b) Measured DOP
for a 40-Gbit/s RZ signal with concatenation of 6-ps
and 4-ps DGD sections [58]. Note that higher-orders
of PMD decrease the signal's maximum DOP at the
receiver to less than unity. 295
Figure A1.1 Generic physical configuration of the optical distribution
network (reproduced from Figure 5/G.983.1). 301
Figure A2.1 Gate frame format. 311
Figure A2.2 Report frame format. 312
Figure A2.3 Register request frame format. 312
Figure A2.4 Register frame format. 313
Figure A2.5 Register acknowledgment frame format. 313
List of Tables
Table 1.1 Summary of different DSL technology performances 4
Table 1.2 Summary of DOCIS data rates 6
Table 1.3 Bandwidth requirements for different IP services 13
Table 2.1 APON downstream/upstream bit-rate combinations 38
Table 2.2 Selected prosperities of IEEE802.3ah EPON transmitters 57
Table 2.3 Selected properties of IEEE 802.3ah receiver
characteristics 58
Table 2.4 G-PON and EPON comparison 69
Table 4.1 PON power budgetsa 154
Table 4.2 ITU-T G.983.3 BPON standards 156
Table 4.3 Physical-layer requirements of ITU-T G-PON and
IEEE EPON standards 157
Table 4.4 G-PON and EPON burst-mode timing comparison 158
Table 4.5 Key PMDparameters of G-PON class B 1.244 Gbps
upstream [10, 20] 159
Table 4.6 Key PMD parameters of IEEE 802.3ah EPON
20-km 1.25 Gbps upstream [9, 22] 160
Table 4.7 State-of-the-art G-PON transceiver performance
parameters for the upstream link at 1.244 Gbps, in
comparison with the ITU-T G.984.2 specifications [10] 201
Table 4.8 A state-of-the-art EPON transceiver uplink performance
Summary in comparison with the IEEE 802.3ah standard 205
Table A1.1 G.984.2 - Physical medium dependant layer parameters
of ODN (reproduced from Table 2a/G.984.2) 302
Table A1.2 G.984.2 - Optical interface parameters of 1244 Mbit/s
downstream direction (reproduced from
Table 2b/G.984.2) 303
Table A1.3 G.984.2 - Optical interface parameters of
2488 Mbit/s downstream direction (reproduced from
Table 2c/G.984.2) 305
Table A1.4 G.984.2 - Optical interface parameters of 1244 Mbit/s
upstream direction (reproduced from Table 2f-1/G.984.2) 307
Table A1.5 G.984.2 - Optical interface parameters of 1244 Mbit/s
upstream direction,using power-levelling mechanism at
ONU Transmitter (reproduced from Table 2f-2/G.984.2) 309

Library of Congress Subject Headings for this publication:

Optical communications.