COM
OPTICAL TECHNOLOGIES,
COMMUNICATIONS APPLICATIONS,
AND INDUSTRY ANALYSIS WORLDWIDE
Theres
MORE INFO online
3 EDITORIAL /// Silicon photonics,
one year later
By STEPHEN HARDY
10 An optical transmitter
for every need
With complex optical modulation
schemes, transmitters may get
complex but more straightforward.
By STEPHANIE MICHEL
15 Superchannel 2.0:
What comes next?
Technology advances can mean more
bandwidth capacity and greater flexibility.
By GEOFF BENNETT
20 WDM-PON is a key component
in next generation access
Recent tests show WDM-PON can fulfill the
need for greater capacity and longer reach.
By EINAR IN DE BETOU, CHRISTIAN- ALEXANDER
BUNGE, HENRIK HLFELDT, and MAGNUS OLSON
24 Optimizing and monetizing
data-center metro networks
SDN and NFV hold significant progress.
The trick is getting there from here.
By JULIUS FRANCIS
28 Factors driving PSM4
and silicon photonics
for data-center architectures
The factors that once drove networks
from copper to MMF create
demand for newer alternatives.
By BRENT HATFIELD
Hot for 2014:
Virtualization in the optical
transport network
BY BRANDON COLLINGS, PH.D, JDSU
In 2014, as with virtualization
in data centers, control plane-
enabled virtualization of the
optical network will simplify life for
network operators considerably.
Infinera CEO predicts
systems-level consolidation
BY STEPHEN HARDY,
Lightwave
In addition to reviewing the
performance of his company
in 2013 and his expectations
for its performance in 2014, Infinera
CEO Thomas Fallon, during the analyst
call to review 4Q13 results, offered
a few opinions about the optical-
network hardware space, which he
believes will begin to get smaller.
Nominations open for 2014
FTTXcellence Award
BY LIGHTWAVE STAFF
Corning Inc. and Lightwave are pleased
to open nominations for the 11th Annual
FTTXcellence Award. The award
recognizes leadership in advancing the
deployment of optical access networks
in North America, Latin America, and
the Caribbean. The winner will receive
a glass art piece from the Corning
Museum of Glass. A $5,000 donation
in the winners name will be made
to U.S. Ignite, along with a profile of
the winner in an issue of Lightwave.
QUANTUM DOT
LASERS AND SILICON
PHOTONICS ADVANCE
DATA-CENTER
CONNECTIVITY
1403lw_1 1 2/21/14 2:10 PM
Innovative and robust JDSU 40/100 Gigabit
Ethernet test solutions unlock the performance and
revenue potential of optical transport network (OTN)
technologies. Evaluate and verify network equipment
with confidence!
Our comprehensive OTN test portfolio includes an
OTU3/4 feature set that complements 40/100 GE physical
layer tests such as photonic and electrical signal testing and
protocol testing for Layers 1, 2, and 3.
Find out moreGet a free World of 100 G poster and
Advanced Error Analysis Offers New Troubleshooting Methods for
High-Speed Data Communications white paper!
NETWORK AND SERVICE ENABLEMENT
NORTH AMERICA 1 855 275-5378
LATIN AMERICA +1 954 688 5660
ASIA PACIFIC +852 2892 0990
EMEA +49 7121 86 2222
www.jdsu.com/nse
Speed 100 GE Testing
with Future-Ready Tools
100 Gigabit Ethernet
World of 100 Gbps The Lower Layers
To learn more, www.jdsu.com/test
Product specifcations and descriptions in this document subject to change without notice. 2012 JDS Uniphase Corporation February 2012 30162982 002 0212 100GBPS.PO.LAB.TM.AE
References: 100GE IEEE 802.3ba, and 802.1Q OTU4 ITU-T G.709, G.872, G.695, and G.959.1 Line I/F OIF IA# OIF-SFI-S and IA# OIF2008.388.00
OTU4
66b Block 0 66b Block 0 66b Block 1 66b Block 1 66b Block 2 66b Block 2
PCS Block Distribution 66b Block 0 66b Block 0 66b Block 20 66b Block 20 66b Block 1 66b Block 1 66b Block 21 66b Block 21
66b Block 19 66b Block 19 66b Block 39 66b Block 39
PCS Lane 0 PCS Lane 1
PCS Lane 19
Round Robin
100 Gigabit Ethernet 103.125 Gbps 100 ppm
IEEE Layer Model
Signal Structure Signal
Structure
PCS Lane 1
PCS Receive
Descrambler
Lane Block Sync & Deskew & Alignment Marker Removal
CGMII
Input to decoder function D0 D1 D2 D3 D4 D5 D6 D7
Receive block
Sync header
S0 S1 S2 S3 S4 S5 S6 S7
RxD<0> RxD<63>
S0 S1 S2 S3 S4 S5 S6 S7 RxB<0> RxB<65>
S0S1S2S3S4S5S6S7
S0S1S2S3S4S5S6S7
RxB<0>
RxB<65>
S0S1S2S3S4S5S6S7
S0S1S2S3S4S5S6S7
RxB<65>
RxB<131>
S0S1S2S3S4S5S6S7
S0S1S2S3S4S5S6S7
RxB<1254>
RxB<1319> PMA service interface
20 PCS lanes in total
PCS Lane 0 PCS Lane 19
Input to descrambler function
<1319>
<1254>
PCS Transmit
PCS Lane 1
Sync header (2-bit wide) 01: data block 10: control block Scrambler
Block Distribution and Alignment Marker Insertion
CGMII
Output of encoder function D0 D1 D2 D3 D4 D5 D6 D7
Output of scrambler function
Transmit block
Sync header
S0 S1 S2 S3 S4 S5 S6 S7
TxD <0> TxD <63>
S0 S1 S2 S3 S4 S5 S6 S7 TxB <0> TxB <65>
S0S1S2
S4
S6S7
S0S1S2S3S4S5S6S7 TxB <65>
TxB <0> S0S1S2S3S4S5 S6S7
S0S1S2S3S4
S6S7TxB <131>
TxB <66> S0S1S2S3S4S5S6S7TxB
TxB PMA service interface
20 PCS lanes in total
PCS Lane 0 PCS Lane 19
Alignment Markers 10 M0 M1 M2 BIP 3 M4 M5 M6 BIP 7 10 M0 M1 M2 BIP 3 M4 M5 M6 BIP 7 Alignment markers are inserted after every 16383 66-bit blocks on each PCS lane
13 3 12
15 5 14 4
11 10
19 18 17 16
9876
0x1A, 0xF8, 0xBD, BIP , 3 0xE5, 0x07, 0x42, BIP7 0x4D, 0x95, 0x7B, BIP , 3 0xB2, 0x6A, 0x84, BIP7 0x5C, 0xB9, 0xB2, BIP , 3 0xA3, 0x46, 0x4D, BIP7 0x59, 0x4B, 0xE8, BIP , 3 0xA6, 0xB4, 0x17, BIP7 2
0x35, 0x36, 0xCD, BIP , 3 0xCA, 0xC9, 0x32, BIP7 0xDD, 0x14, 0xC2, BIP , 3 0x22, 0xEB, 0x3D, BIP7 0x83, 0xC7, 0xCA, BIP , 3 0x7C, 0x38, 0x35, BIP7 0xF5, 0x07, 0x09, BIP , 3 0x0A, 0xF8, 0xF6, BIP7
0xB9, 0x91, 0x55, BIP , 3 0x46, 0x6E, 0xAA, BIP7 0x9D, 0x71, 0x8E, BIP , 3 0x62, 0x8E, 0x71, BIP7 1 0xFD, 0x6C, 0x99, BIP , 3 0x02, 0x93, 0x66, BIP7 0xC1, 0x68, 0x21, BIP , 0x3E, 0x97, 0xDE, BIP 3 7 0
0xC0, 0xF0, 0xE5, BIP , 3 0x3F, 0x0F, 0x1A, BIP7 0x5F, 0x66, 0x2A, BIP , 3 0xA0, 0x99, 0xD5, BIP7 0xA0, 0x24, 0x76, BIP , 3 0x5F, 0xDB, 0x89, BIP7 0xAD, 0xD6, 0xB7, BIP , 3 0x52, 0x29, 0x48, BIP7 0x7B, 0x45, 0x66, BIP , 3 0x84, 0xBA, 0x99, BIP7 0xC4, 0x31, 0x4C, BIP , 3 0x3B, 0xCE, 0xB3, BIP7 0x9A, 0x4A, 0x26, BIP , 3 0x65, 0xB5, 0xD9, BIP7
0x68, 0xC9, 0xFB, BIP , 3 0x97, 0x36, 0x04, BIP7
PCS Lane Encoding PCS Lane Encoding
MAC 802.3 MAC Frame Structure
Type/ Length MAC Payload Source MAC address FCS Destination MAC address
6 bytes 6 bytes 2 bytes 4 bytes 46 to 1500 bytes VLAN (optional)
4 bytes Preamble SFD 7 bytes 1 byte VLAN (optional)
4 bytes
VID: VLAN ID CFI User Priority Type/ Values > 1535 are Ethernet Types Length Values 1500 are for Length of the MAC frame with preamble/SFP and typically used with LLC/SNAP in the MAC Payload
1234
1
Multilane Model
Optics for OTU4 inter-Domain Interface similar to 100GBase-LR4 and ER4 applicable to OTU4
Logical Lanes
0 1 2 3
Physical Lanes 11.181Gbps 0 1 2 8 9 ...
Physical Lanes 27.952Gbps
OTL4.10
... 0 1 2 18 19
OTM-0.4v4
OTL4.4
OTU4
... 0 1 2 18 19
OTM-0.4v4
Potential Evolution
Logical Lanes
PMA 10:4
0 1 2 3
OTU4
... 2:1 2:1 2:1
... 5:1 5:1 5 Logical lanes:1 ... 5:1 5:1
PCS 64B/66B PCS Block Format Input Data
Data Block Format Control Block Formats C C C C /C C C C 0 1 2 3 4 5 6 7
D D D D /DD D D 0 1 2 3 4 5 6 7
S D D D /DD D D 0 1 2 3 4 5 6 7 O D D2D /Z Z Z Z 0 1 3 4 5 6 7 T C C C /C C C C 0 1 2 3 4 5 6 7 D T C C /C C C C 0 1 2 3 4 5 6 7 D D T C /C C C C 0 1 2 3 4 5 6 7 D0D1D2T3/C4C5C6C7 D D D D /T C C C 0 1 2 3 4 5 6 7 D D D D /DT C C 0 1 2 3 4 5 6 7 D D D D /DD T C 0 1 2 3 4 5 6 7 D D D D /DD D T 0 1 2 3 4 5 6 7
SY NC 0 1 01
10 10 10 10 10 10 10 10 10 10 10
65 D D D D4 D5 D6 D7
C C C C3 C4 C5 C6 C7 D D D D4 D5 D6 D7 D D D O0 C C C3 C4 C5 C6 C7 D C
2
22 C3 C4 C5 C6 C7 D D C3 C4 C5 C6 C7 D D D C4 C5 C6 C7 D D D D3 C5 C6 C7 D D D D4 C6 C7 D D D D4 D5 C7 0x
Block Payload
2 D0 Block Type Field 0x1E 0x78 0x4B 0x87 0x99 0xAA 0xB4 0xCC 0xD2 0xE1 FF D
1
0110000000 D
2
1 22 1 111111 D2
3
33
2222 D4 D5 D6 D3 D3 D3
0x000_0000
Bit Position
Skew Tolerance MaximumSkew Variation (10.3 GBd) MaximumSkew SkewPoint
0.4 ns (~2 UI) 43 ns (~222 UI) SP2 0.6 ns (~4 UI) 54 ns (~278 UI) SP3 3.4 ns (~35 UI) 134 ns (~691 UI) SP4 3.6 ns (~37 UI) 145 ns (~748 UI) SP5 3.8 ns (~39 UI) 160 ns (~824 UI) SP6 4 ns (~41 bits) 180 ns (~928 UI) At PCS Rx
0.2 ns (~2 UI) 29 ns (~150 UI) SP1
SP Points shown in Multilane Model
Multilane Model
Additional optical implementations such as 10*10G can be considered
...
... 0 1 2 18 19
0 1 2 8 9 ...
PMD
10 Fibers 100GBase-SR10 850 nm MMF 100 m in OM3 125 m in OM4
PMA 20:10
PCS Lanes
Physical Lanes CPPI 10.3125 Gbps
PCS
PMA 20:10
PMA 10:4
PCS Lanes
Physical Lanes CAUI 10.3125 Gbps
0 1 2 3
... 0 1 2 18 19
0 1 2 8 9 ...
... 2:1 2:1 2:1
... 5:1 5:1 5 PCS lanes:1
PCS
SP2 SP5
SP6 SP1
100GBase-LR4 1310 nm SMF 4*25.78 Gbps 10 km 100GBase-ER4 1310 nm SMF 4*25.78 Gbps 30 km or 40 km (engineered links)
0 L 1 L 2 L 3 L 0 L 1 L 2 L 3 L
PMDSP4 SP3
PMD Service Interface
... 2:1 2:1 2:1
0 L 1 L 2 L 8 L 9 L
Potential Evolution
ODU4 Multiplexing
ODUMultiplexing
PSI for Multiplexing
ODU4 Payload ODU4 OH
OPU4 Payload OPU4 OH OPU4 Payload
ODU4
OPU4
ODTUG4 (PT=0x21)
ODU ODTUG4 (PT=0x21) ODU ODTU4.ts JOH ODTU4.ts JOH
ODTU4.ts JOH ODU (0, 1, 2, 2e, 3, 4, fex) ODTU4.ts
ODU Payload ODU OH ODU
PT=0x21 [0] [1] RES
[255]
TS1 [2] TS2 [3]
TS80 [81]
Trib Port# Trib Port#
Trib Port# TSO
TSO TSO TS Occupied 0 Unallocated 1 Allocated Tributary Port# The Tributary Port# indicates the ODUj (j=0, 1, 2, 2e, 3, fex) transported in this TS; Multiple 1.25G TS can be grouped
1 2 3 4 5 6 7 8 (Bits)
PSI
Tributary Slots (1.25G) for Multiplexing
ODU4 OH
OTU4 OH
TSOH TS1 FA
OTU4 FEC
Column (bytes) Row
Fixed Stuf
ODU4 OH
OTU4 OH
TSOH TS2 FA
OTU4 FEC Fixed Stuf
ODU4 OH
OTU4 OH
TSOH TS80 FA
OTU4 FEC Fixed Stuf
PSIOMFI
PSIOMFI
PSIOMFI
OMFI = 0 (TSOH TS1)
OMFI = 1 (TSOH TS2)
OMFI = 79 (TSOH TS80)
18 17 16 15 14 8 7 158 57 56 5598 97 96 953817 3816 38154080 3825 3824
1
141
41
2
242
42
39
3979
79
40
4080
80
41
411
1
42
422
2
...
79
7939
39
80
8040
40
1
141
41
2
242
42
...
39
3979
79
40
4080
80
...
1
141
41
2
242
42
39
3979
79
40
4080
80
41
411
1
42
422
2
...
79
7939
39
80
8040
40
1
141
41
2
242
42
...
39
3979
79
40
4080
80
...
1
141
41
2
242
42
39
3979
79
40
4080
80
41
411
1
42
422
2
...
79
7939
39
80
8040
40
1
141
41
2
242
42
...
39
3979
79
40
4080
80
...
12341234
1234
OTL4.10 OTL4.4 255/227 x 9.953280 Gb/s = 11.180997 Gbps OTL Nominal Bit Rate OTL Type
255/227 x 24.883200 Gb/s = 27.952493 Gbps
Optical Average Optical Power 100GBase- SR10 Parameter 100GBase- LR4 100GBase- ER4
8 dBm Min Launch Power/lane 4.3 dBm 2.9 dBm 2.4 dBm Max Avg Rx Power/lane 4.5 dBm 4.5 dBm 9.9 dBm Min Avg Rx Power/lane 10.6 dBm 20.9 dBm
2.4 dBm Max Launch Power/lane 4.5 dBm 2.9 dBm
LWDMWavelength Range
1295 1300 1305 1310 (nm)
100GBase-LR4 and ER4 LWDMLane Assignments
1308.09 to 1310.19 nm 1309.14 nm L3 1303.54 to 1305.63 nm 1304.58 nm L2 1299.02 to 1301.09 nm 1300.05 nm L1 1294.53 to 1296.59 nm 1295.56 nm L0 Range Center Lane
IEEE Model with CAUI (to interconnect chip devices)
MAC Reconciliation
LLC or other MAC Client Higher Layers (e.g., IP)
PMA 100GBase-R PCS CGMII
PMA PMD MDI Optical Medium
CAUI
Logical Lanes Byte Distribution from OTU4 to the 20 Logical Lanes Rotate Rotate Rotate Rotate
289:304
305:320 33:48 17:32
16289:16304 16001:16016 16305:16320 305:320 17:32
625:640 337:352 321:336 16305:16320 ... 16017:16032 16001:16016
16001:16016 16033:16048 16017:16032
305:320 17:32 1:16 (FAS) LL 0
LL 19
LL 1 Multiple Rotations
1:16 (FAS) 1:16 (FAS)
1:16 (FAS) Bytes from the OTN frame are sequentially ordered in groups of 16 bytes row by row and rotated by OTL LLMis processed at the OTL layer
PT of 0x07 for 100GBase-R client
1 1 0 0 0 0 0 0 0 1 0 0 0 0 0 0 0 0 0 0
19 1 1 0 0 1 0 0 0 19
19 18 13210
11 0
000 OA2 LLM Bits
110
0002
110
0003
001
0004
110
0005
111
0006
110
1007
10010108
239 238 20 3210 Decimal OA2 LLM Bits 2 3 4 5 6 7 8 Decimal LL 11
0 0 0 0 0 0 1 Used for logical lane identifcation and skew Logical Lane Markers
FAS Byte Values
LLM 28 28 F6 F6 F6 OA2 OA2 OA2 OA1 OA1 OA1
Information Containers
OTU4 111.809974 Gbps 20 ppm
OTU4 Section
0.4v4 (4 optical lanes)
Client (e.g.,100GE)
OPU4 ODU4 PMOH ODU4 Path
OPU4 Payload OPU4 OH OPU4
OTU4 FEC OTU4 OH
OTLCG OTLCp OTLCp OTLCp
OTM OPSM
OTL4.n #0 OTL4.n #1 OTL4.n #n-1 OT Lanes n: number of physical lanes
ODU4 TCMOH ODU4 Tandem Connection
OTU4 TC L1
ODU4 TCMOH
OTU4 TC L6 ODU4 TCMOH
Frame Structure
OPU4 OH
FAS MFAS RES TCM6 TCM5 TCM4 FTFL TCM ACT SM GCC0 RES
TCM3 TCM2 TCM1 PM EXP GCC1 GCC2 APS/PCC RES
2 3 4 5 6 7 8 9 10 11 12 13 14 15 16
ODU4 OH OPU4 Payload
OTU4 OH
OPU4 OH FA
OTU4 FEC
Column (bytes) Row
Fixed Stuf
17 16 15 14 8 7 13817 38164080 3825 3824
1234
SAPI
DAPI
Operator Specifc
TTI BIP-8
SM, PM, and TCMi (i = 1...6) (bytes)
BEI/BIAE RES BDIIAE(bits) SM
BEI/BIAE RES BDI(bits) TCMi
BEI STAT BDI(bits) PM
1 2 3
0
15 16 31 32
63
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
1 2 3 4 5 6 7 8
ODU4 SNC/S, ODU4 SNC/N ODU4 TCM3 011 ODU4 SNC/S, ODU4 SNC/N ODU4 TCM2 010 ODU4 SNC/S, ODU4 SNC/N ODU4 TCM1 001 ODU4 SNC/N ODU4 Path 000
ODU4 SNC/S, ODU4 SNC/N ODU4 TCM5 101 ODU4 SNC/S, ODU4 SNC/N ODU4 TCM4 100
ODU4 server layer trail ODU4 TCM6
APS/PCC Channel
ODU4 SNC/I ODU4 SNC/S, ODU4 SNC/N
Protection Scheme using the APS/PCC Channel
110
MFAS Bits 678
111
Justifcation via GMP
OMFI
JC1 JC2 JC3 JC4 JC5 JC6 PSI
1234
15 16
PT
Mapping & Concat. Specifc
01
255
DI II C14 C13 C12 C11 C10 C9 CRC-8 JC3 JC2 JC1 C2 C1 C8 C7 C6 C5 C4 C3
JC6 JC5 JC4 D10 D9 D8 D7 RES RES CRC-5 D6 RES D1 D5 D4 D3 D2
21 80 79 54321 TSOH
0000 000001 OMFI Bits
0011000002
0000000003
0000000004
0011000005
0011100006
0011011007
1010010108
10 79 78 43210 Decimal 1.25G TS
CAUI 100 Gbps attachment unit interface CGMII 100 Gigabit media independent interface CFI canonical format indicator CPPI 100 Gbps parallel physical interface FCS frame check sequence LLC logical link control LWDM LAN wave division multiplexing MAC media access control MDI media dependent interface PCS physical coding sublayer
PMA physical media attachment PMD physical media dependent RS reconciliation sublayer SFD start frame delimiter SP skew point VLAN virtual local area network
APS automatic protection switching BDI backward defect indication BEI backward error indication BIAE backward incoming alignment error BIP-8 bit interleaved parity-8 DAPI destination access point identifer EXP experimental FAS frame alignment signal FEC forward error correction FTFL fault type and fault location
GCC General Communication Channel GMP generic mapping procedure IAE incoming alignment error JC justifcation control JOH justifcation overhead LLM logical lane marker MFAS multi-frame alignment signal ODTU optical channel data tributary unit ODTUG optical channel data tributary unit group ODU optical channel data unit
OH overhead OMFI OPU multi-frame identifer OPSM optical physical section multilane OPU optical channel payload unit OTL optical channel transport lane OTLCG optical transport lane carrier group OTN optical transport network OTU optical channel transport unit PCC protection communication channel PM path monitoring
PMA physical media attachment PMOH performance monitoring overhead PSI payload structure identifer PT payload type RES reserved SAPI source access point identifer SM section monitoring SNC subnetwork connection SNC/I SNC protection with inherent monitoring SNC/N SNC protection with non-intrusive monitoring
SNC/S SNC protection with sublayer monitoring STAT status TC tandem connection TCM tandem connection monitoring TCMACT TCMactivation TCMOH tandem connection monitoring overhead TS tributary slot TTI trail trace identifer
GMP provides a generic mapping method for the justifcation of CBR client signals into OPU
Ci Relative change indication II Increment Indicator DI Decrement Indicator Di Ci-derived cumulative value
MAC Reconciliation
LLC or other MAC Client Higher Layers (e.g., IP)
PMA
PMD
100GBase-R PCS CGMII
MDI Optical Medium
0
Physical Lanes 25.78125Gbps
PCS Lanes
PMD
0 1 2 3
PMA 20:4
... 1 2 18 19
PCS
... 5:1 5:1
C8 for 100GBase-R client
15050.518 15052.325 15054.131 15055
Minimum Nominal Nominal Ceiling
15050 Floor
White Paper
WEBSITE: www.jdsu.com
Introduction
25 G The newstandard for I/O
Over the past decade, 10 G has become the de facto standard for both long and short-reach high-speed
(premium) data communications links. Significant resources have gone into optimizing integrat-
ed circuits (ICs) with 10 G input/output (I/O) since the late 1990s to establish a healthy ecosystem for
deploying 10 G links cost-effectively. This scales well for short inter-chip interfaces through to long
reach (LR 10 km) optical modules and beyond. The 10 G technology displaced the expensive (first) 40 G
technology in 40 GE because of its much improved cost scaling. However, it required establishing a new
defacto rate to meet the needs of newer standards, such as 100 GE.
The ideal technology for high-speed communications interfaces is a mainstream option with the fewest
parallel channels for more cost-effective implementation.
Transport choices for 100 GE might include 10 x 10 G (used as initial host interface), 4 x 25 G (standard),
2 x 50 G, or 1 x 100 G. Clearly 50 G and 100 G I/O are extremely challenging and likely to carry a signifi-
cant price-premium for several years so the choice was down to 10 x 10 G or 4 x 25 G. The option of 10 G
could leverage the existing 10 G I/O technology and would build up a body of knowledge over the three
generations of 10 G ICs while 4 x 25 G uses 40 percent of the components (hence reduced volume, cost,
connector size, and PCB trace area). Trends indicated a move toward 100 GE based on 4 x 25 G; although,
a 10 x 10 G host electrical interface was used on the first-generation (CFP) because 25 G technology was
too novel for use as a widely deployed pluggable interface.
Soon 25 G-based I/O will become the de facto I/O speed for many future technologies, including 100 G
Ethernet, OTU4, and Infiniband. Also, you can find 25 G I/O on application specific integrated circuits
(ASICs), clock and data recovery (CDR), and field-programmable gate arrays (FPGAs) today.
Challenges
Signal integrity, crosstalk, CDRand FIFO, real data signals, and jitter
The price/performance andpower capability of todays third-generation10 GI/Ousedfor most high-speed
data links can be deployed cost-effectively. Even with this established technology, first-generation 100 G
basedon10x10Gpresents manysignal integrityandperformance issues suchas jitter tolerance anddynamic
skew. The move to 25 Gwill require resolving many more issues before 100 G(using 4 x 25 G) canbecome a
true mainstreamtechnology.
The major issues, especially with the first-generation 25 G I/O ICs include:
signal integrity
CDRperformance
jitter tolerance
dynamic skew tolerance
pattern sensitivity
Advanced Error Analysis Ofers New Troubleshooting Methods
for High-Speed Data Communications
By Paul Brooks and Juan Masmela
1403lw_2 2 2/21/14 2:10 PM
EDITORIAL
F
O
L
L
O
W
S
T
E
P
H
E
N
O
N
O
ne of the bigger stories going
into last years OFC/NFOEC was
the commotion started when
Jefferies & Co.s James Kisner downgraded
Finisars stock in light of the impending
impact of silicon photonics technology on
the data-center interconnect market. Intel
had recently revealed its silicon photonics
efforts were about to reach the engineering
sampling stage, and Cisco was expected
to unveil the first fruits from its acquisition
of silicon photonics developer Lightwire
at the show (which, in fact, it did). Add
these events to the fact that companies
such as Luxtera and Kotura (now part of
Mellanox) had already produced silicon
photonics engines, and it seemed the
era of silicon photonics had arrived.
Kisner argued the potential cost benefit
of silicon photonics versus conventional
approaches was so
compelling that companies
that didnt possess this
technology yet like, say,
Finisar would quickly
become technological
dinosaurs. So down
went Finisars rating.
One year later, Kisner has
weighed in again on Finisar. This time hes
raised his rating from Hold to Buy, based on
his assessment of the opportunities for Finisar
in the data-center market. Is that because
Finisar finally jumped into silicon photonics?
Well, no silicon photonics wont begin to
play a meaningful role in shaping the price
of optical interconnects until sometime
next year at least, Kisner now reasons.
Meanwhile, theres enough momentum in
transceiver demand at 10 Gbps and below
not to mention opportunities for transceivers
and other products in carrier networks
where silicon photonics role is murky at best
to make Finisar stock an attractive buy.
As if to pour more water on last years
silicon photonics frenzy, fellow financial
industry analyst Simon M. Leopold of
Raymond James issued a note shortly
after Kisners missive that Leopold
entitled Silicon Photonics: Not Ready
for Prime Time Yet. Leopold argues
that silicon photonics likely will play its
most significant role in very short reach
data-center applications and may have
an impact in long reach links but will
need some time to mature. Meanwhile,
companies like Finisar will have plenty of
opportunity to acquire silicon photonics
expertise if/when it becomes necessary.
The reduction in expectations for
silicon photonics is a welcome occurrence
(except for companies in the space looking
for financing). Optical communications
Silicon photonics, one year later
STEPHEN HARDY
Click to view
Stephen's video blog
(Having trouble? Click here.)
1403lw_3 3 2/21/14 2:11 PM
USA/Canada 1-800-ANRITSU Europe 44 1582-433433
Japan 81 (46) 296-1208 (852) 2301-4980
South America 55 (11) 3283-2511
Network Master MT9090 OTDR
with Fiber Visualizer
Light it Up
Optical Time Domain Refectometers
with Confdence
2013 Arr|lu Corpary
Anritsus handheld Network Master MT9090A OTDR portable
design delivers 1 x 64 split completely from end-to-end on high
resolution, color 4.3-inch wide screen display.
Now with Fiber Visualizer, our one-button fault locate function, will
make your life easier as it automatically selects testing parameters
and provides a test summary within seconds.
Other Network Master MT9090A OTDR features include:
> Tri-wavelength OTDR for both installation and maintenance
> Built-in PON Power Meter, Loss Test Set and Light
Source functions
> Bluetooth
T
HE INSATIABLE DEMAND
for smart devices
with faster connectivity
coupled with cloud as a means
of storage and computing resources
has revolutionized the fabric of
data centers in recent years. They
have led to the emergence of a
new architecture focused on
faster, more resilient content
delivery networks.
This new architecture evolution in
data centers has resulted in a growth
industry thats showing no signs
of abating. Expenditures on data-
center systems, including servers,
Ethernet switches, and storage, were
estimated to be over $147 billion
in 2013, with a growth rate of 4.5%
year over year.
1
This spending
trend will continue as the need to
deliver a better user experience
drives increased investment in
new disruptive technologies. These
new technologies
will also enable a
scalable and cost-
effective network
infrastructure
for cloud, mobile, social, and
big data/analytics.
One key area of focus in recent
years has been on new optical
technologies that reduce interface
cost while supporting longer
reaches at increased data rates
for both 2-km intra-data-center
connectivity and 100-km inter-data-
center superchannel connectivity.
Classic 3-tiered
data-center architecture
Servers to
switch level 1
Servers to
switch level 2
Servers to switches
Legacy
northsouth
traffc
FIGURE 1. The traditional three-tiered
data-center architecture.
By SAEID ARAMIDEH
New pathways to low
power and cost will open
the door to more efficient
data-center network
architectures.
QUANTUM DOT
LASERS AND SILICON
PHOTONICS ADVANCE
DATA-CENTER
CONNECTIVITY
1403lw_5 5 2/21/14 2:11 PM
FEATURE Quantum dot lasers and silicon photonics
Changes in data-center
architecture
According to a study conducted by
Cisco Systems, data-center traffic
on a global scale will continue to
grow at a 25% compound annual
growth rate (CAGR), with the fastest
growing segment being the cloud
data-center traffic at 35% CAGR, a
4.5-fold growth from 2012 to 2017.
2
To further corroborate the changes
inside the data center, Ciscos 2013
Global Cloud Index report suggests
that data-center traffic will triple by
2017 and a staggering 76% of that traffic
is within the data centers server-to-
server traffic. This important trend
in data-center networking is driving
requirements for denser server
connectivity and longer network reach.
Unfortunately, a simple scaling
based on current optical-interface
technology within the data center
contradicts the requirement for
reduced network complexity and
fails to deliver a cost-effective
approach. Data-center operators
will require much higher density
optical interfaces that offer
scalability from tens or hundreds
of gigabits to terabits per second.
Data-center networks have
traditionally been designed to support
conventionally managed hosting
services, where the majority of the
data flows in and out of the data center
based on traditional northsouth traffic
flow in a three-tiered hierarchical
design (see Figure 1).
3
But hierarchical
data-center networks have grown
complex, power hungry, and costly
in terms of optical-fiber use and
optical interfaces. This approach is
ill-suited for the machine-to-machine
traffic required of todays growing
services and evolving applications.
In addition, server-to-server
traffic in a three-tiered architecture
increases network latency when
traffic traverses multiple network
layers in a northsouth direction.
This three-tiered architecture
doesnt fit with todays widespread
adoption of server virtualization and
distributed computing that drives an
increasingly eastwest traffic plan.
As such, moving to a two-tiered
spine-and-leaf architecture enables the
1403lw_6 6 2/21/14 2:11 PM
f or out door assembl i es
User friendly All-In-One System
Allowing stripping, cleaving, splicing
Complete kit with all required tools
From cable preparation to connector assembly
Enhanced cable handling for 3-8mm OD
outdoor, Mil and industrial cables
Suitable for termination of following SOCs:
E-2000, F-3000 (LC), SC, FC and ST
New modular termini
For 1.25mm and 2.5mm ferrule
DIAMOND SA
Via dei Patrizi 5, CH-6616 Losone
Tel. +41 91 785 45 45, Fax +41 91 785 45 00
sales@diamond-fo.com
www. di amond- f o. com
OD3
Visit us at the OFC/NFOEC 2014, booth #1823
X-BEAM revos E-2000
All-In-One System
FEATURE Quantum dot lasers and silicon photonics
data-center operator to address todays
changing traffic patterns and effectively
eliminate the aggregation layer from
the previous design (see Figure 2).
4
Furthermore, this architecture
provides a lower latency solution
and is better suited to accommodate
high bandwidth and delay-sensitive
inter/intra-data-center traffic flows.
Enabling a better approach
To maximize the advantages that
the new generation of spine-and-
leaf architectures offer, as well as
open the door to eventual use of
superchannel connectivity, a new
generation of optical interconnectivity
is required. This technology must:
Su|icy |l ircrucirq
demand for bandwidth.
Dlivr ciqriicur| rduc|ior ir
the power dissipation, size, and
cost of optical interconnects.
lduc r|work lu|rcy.
An extremely promising approach
toward meeting these requirements
pairs innovation in quantum dot
rul|iwuvlrq|l (D M\) lucrc
with advanced digital and photonic
integrated-circuit technologies.
A D M\ lucr ic u rurocryc|ul
semiconductor device capable of
generating multiple wavelengths
cirul|uroucly. A cirql D M\ lucr
can provide up to 96 wavelengths in
the C-band alone, which would enable
the replacement of 96 individual lasers
in an equivalent system. The overall
quir burdwid|l o |l D M\ lucr
is over 90 nm, which enables future
scalability for up to 200 wavelengths.
Tl D M\ lucr rublc u
much simpler and more compact
interface design, resulting in lower
power consumption and better
performance when compared to
an equivalent design with multiple
discrete lasers and components.
Murwlil, lo|oric ir|qru|ior
luc bcor u ky uru o rcurcl
and development in recent years
or o|iculr|work quirr|
vendors wanting to reduce the cost
urd ciz o o|icul ir|rucc. Silicor
technology and integration enables
system vendors to replace expensive
optical components with their silicon
counterparts such as modulators,
multiplexers/demultiplexers, filters,
attenuators, switches, and detectors.
A novel use of silicon photonics
has resulted in the implementation of
optical modulation on a micro-ring
resonator architecture that enables
high-speed modulation at data
rates up to 56 Gbps. These ring
rodulu|orc uc| lik u wuvlrq|l
crci|iv cwi|cl. \lr |l rirq
is not in resonance, the optical
signal passes straight through with
minimal loss. Conversely, when
the ring is resonant, the optical
New data-center
spine-and-leaf architecture
Switch
to switch
Leaf
Spine
Servers to switches
Web 2.0 eastwest traffc
FIGURE 2. The new data-center
spine-and-leaf architecture is
better suited for emerging needs.
1403lw_7 7 2/21/14 2:11 PM
>
>
>
>
>
>
>
>
Innovative solutions enable next-generation fiber optic
communications at 100Gbps and beyond
Industry-leading Portfolio Drives
Future of 100G Optical Networking
GPPO Module
Surface Mount
Die
T
X
/
R
X
C
D
R
E
M
L
D
r
i
v
e
r
L
i
m
i
t
i
n
g
T
I
A
D
M
L
D
r
i
v
e
r
S
i
n
g
l
e
-
e
n
d
e
d
M
Z
D
r
i
v
e
r
(
L
i
n
e
a
r
/
L
i
m
i
t
i
n
g
)
D
i
f
f
e
r
e
n
t
i
a
l
M
Z
D
r
i
v
e
r
(
L
i
n
e
a
r
/
L
i
m
i
t
i
n
g
)
L
i
n
e
a
r
T
I
A
100G Client Side 100G Line Side
MACOM is fueling the future of optical
communications infrastructure from long haul
to metro networks to data centers. Leveraging
Gallium Arsenide, Indium Phosphide and
Silicon Germanium technologies we are
delivering market-leading performance with
lowest power and smallest size.
We offer modulator drivers, transimpedance
ampliers (TIA) and clock/data recovery (CDR)
products supporting data rates of 100Gbps and
higher for line and client side systems in CFP,
CFP2, CFP4 and QSFP28 modules.
Drawing on deep domain knowledge in optical
communications, MACOM engineers are
delivering a portfolio of low-power, cost-effective
solutions for 100G and beyond.
Lowest power, smallest form factor drivers for
100G metro transponders
Lowest power solution for 100G Ethernet
Transceivers (CDRs, Drivers, TIAs)
Drivers for InP, Lithium Niobate and Silicon
Mach-Zehnder Modulator (single ended and
diffrential drives)
EML and DML drivers
Linear solutions for 200G and 400G
GPPO, surface mount and die solutions
Single, dual and quad channel packages
CFP, CFP2,CFP4 and QSFP28 application solutions
Order samples and get more information at
www.macomtech.com/opto
FEATURE Quantum dot lasers and silicon photonics
signal is coupled into the ring
and no longer passed through.
These modulators are fabricated
with a silicon-on-insulator process
using silicon waveguides that
provide natural vertical and lateral
confinement via a ridge. Such ring
resonators are ideal WDM building
blocks and can be cascaded to
construct WDM transceivers. In
such a construct, each ring provides
a multiplexing, demultiplexing,
and modulation operation in a
geometry thats well suited to
extreme scaling of many channels.
When we couple a single QD MW
laser source, which generates multiple
wavelengths simultaneously, with a
silicon photonics ring architecture,
we get a 1550-nm design that greatly
reduces the overall cost and networking
complexity within the data-center
optical interconnect (see Figure 3).
In contrast, Figure 4 illustrates how
existing transceiver architectures
using traditional 850/1310-nm single-
wavelength laser technologies are
ideally suited to multifiber designs
that fail to cost-effectively scale to
higher capacities primarily due to
the required fiber plant complexity.
To cost-effectively scale data-
center networks to meet growing data
rates, operators must future-proof
todays infrastructure investment.
Network economics
A study that compared a Parallel Single
Mode, 4 Lanes (PSM4) approach
against the QD MW/silicon photonics
design
described
earlier shows
a significant
advantage
for 1550-nm
designs with
single-fiber
connectivity
Resonant-ring modulators
Laser
source
Transmitter technology Receiver technology
Fiber
Single-fber architecture
1550-nm
solution
Resonant rings
FIGURE 3. Intra-data-center connectivity using technology
that combines a single QD MW laser source with silicon
photonics-based ring resonators.
1403lw_8 8 2/21/14 2:11 PM
FEATURE Quantum dot lasers and silicon photonics
as opposed to
multiple fiber
strands. The
study estimates
a 2.5X to 4X cost
advantage when
using the single-
fiber approach
as capacity
scales from 100G
to 1.6T for intra-data-center links.
Another study looked at inter-data-
center connectivity with links of up
to 120 km. The study compared the
basic transmitter layout of coherent
architectures for 100G and 400G
to the one based on the QD MW
laser with silicon photonics ring
resonators. The study demonstrated
over 5X power reduction and 24X size
reduction using the latter approach
when compared to coherent 400G
architectures. An evaluation of the full
link cost for scaling the 400G link for
120 km for data-center interconnect
based on both the coherent and QM
MW/silicon photonics approaches
showed a 2X to 3X cost advantage
for the non-coherent design.
These studies demonstrate that
the combination of the QD MW laser
and silicon photonics devices is an
ideal implementation for inter/ intra-
data-center interconnects. These
technologies have demonstrated a
significant cost and power advantage
relative to discrete implementation
and enable a much higher
scalability for the next generation
of data-center infrastructure.
References
1. Gartner report, May 2013.
2. Cisco Global Cloud Index: Forecast
and Methodology, 20122017.
3./4. James Kisner, Technology
Silicon Photonics: Data Center
Disruptor, Jeffries, 2/22/13.
Laser MZ modulator
MZ modulator
MZ modulator
MZ modulator
MZ modulator
MZ modulator
MZ modulator
MZ modulator
Fiber
1
Multifber
architectures
Fiber
2
Fiber
3
Fiber
4
Fiber
1
Fiber
2
Fiber
3
Fiber
4
Laser
Laser
Laser
Laser
1
:
s
p
l
i
t
t
e
r
n
Laser
source
1310-nm
technology
Transmitter
technology
Receiver
technology
PIN
1
PIN
2
PIN
3
PIN
4
PIN
1
PIN
2
PIN
3
PIN
4
FIGURE 4. Intra-data-center connectivity using
conventional 1310-nm technology.
1403lw_9 9 2/21/14 2:11 PM
0
FIGURE 2. Periodic relation of transmitted
power and voltage difference in a
MachZehnder modulator, with U
/
being
the voltage difference between zero
transmission and full power transmission.*
*P.J. Winzer and R.J. Essiambre:
Advanced Optical Modulation
Formats, IEEE Proceedings, 94(5),
2006. All other figures in this
article are contributed by Oliver
Funke, Stephanie Michel, Bernd
Nebendahl, and Ren Schmogrow.
1403lw_11 11 2/21/14 2:11 PM
See us at
OFC 2014
BOOTH # 721
613-599-1000 I info@jgroptics.com I www.jgroptics.com
Ask us about our short Lead-time!
o p t i c s
JGR
Optical Switches
R
1
2
3
...
...
...
N
A
R
R
1
2
3
4
5
..
.
..
.
..
.
..
.
..
.
N
2
1
A- Confguration: 1xN C- Confguration:
MxNNon-blocking
R
R
1A
1B
2A
2B
...
....
NA
NB
A
B
1-1
1-2
...
1-N
2-1
2-2
...
2-N
...
M-1
M-2
...
M-N
1
2
...
M
B- Confguration: Ganged
1xN(MxN) Confgurations
1xN or 2xN Stepper Motor Switch
Single-mode or Multimode
Low Insertion Loss 0.70dB
Backrelfection SM -60dB & MM -40dB
Excellent Repeatability 0.005dB
GPIB/RS232/USB Interface via USB-DB9 Adapter
C
u
sto
m
C
o
n
f
g
u
ra
tio
n
?
Ju
st a
sk
...
A- Confguration: 1xN C- Confguration:
n
?
t a
sk
...
See us at
BOOTH # 721
o p t i c s
C
u
sto
m
C
o
n
f
g
u
ra
tio
n
?
Ju
st
a
sk
FEATURE Optical transmitter for every need
experiences a phase
shift of
/
and on the upper
branch a phase shift of
/
.
In the green example, its
vice versa. Both examples
have in common that when
recombining the signals of
both arms, they interfere
destructively; the two
vectors sum up to a zero
vector. Therefore, in the
red and green examples,
we have no signal at the
exit of the modulator.
In the case of the yellow
signal, the MachZehnder
modulators voltages are adjusted
in such a way that on both branches
the phase of the signal is shifted
by /. When overlaying the two
signals, we get a constructive
interference. The resulting wave
is a sine wave of the original
amplitude shifted by \ = /.
MachZehnder modulator
for transmission of a QPSK signal
Now well see how the MachZehnder
modulator is used in a transmitter
setup using QPSK modulation. In
Figure 4, the complete block diagram
is given along with a recap of the
principles of the QPSK modulation
we discussed in detail in the article
Which optical modulation scheme
best fits my application?
In QPSK modulation, we double
the transmission rate relative to
OOK by encoding 2 bits into
one symbol. The four possible
symbols are in the IQ diagrams
four constellation points, which
are all situated on the same
circle. That means we operate
Q
I
Q Q Q
Q
Q
Q Q
I I
+
=
I
I
I
I I
FIGURE 3. Examples of phase shifts in a Mach
Zehnder modulator, time domain, and IQ diagram.
1403lw_12 12 2/21/14 2:11 PM
FEATURE Optical transmitter for every need
with one amplitude only. The
points are separated by
/
.
In the transmitter, the electrical
bit stream is split by a demultiplexer
into the I and Q part of the signal.
Each of the two parts directly
modulates the phase of the laser
signal on one arm of a Mach
Zehnder modulator. An additional
MachZehnder element shifts the
phase of the lower branch, the Q
branch, by
/
. After recombination
of the two branches, we get the
resultant QPSK signal as shown
at the bottom of Figure 4.
Transmitters for more complex
modulation schemes
When it comes to higher-order
modulation schemes like 16-QAM,
the transmitter setup must be able
to provide more amplitude levels
and phases, which means higher
complexity. In 16-QAM, each symbol
encodes 4 bits and we need two
optical power levels. To accomplish
this, there are several approaches that
differ in their modularity and in the
level of performing the modulation
in the electrical or optical domain.
Figure 5 offers four implementation
examples for
comparison.
On the top left is
a transmitter that
consists of discrete
components. The
digital-to-analog
conversion (DAC)
is done on the
optical signal.
Using a BERT as an
example, there are
four output channels
to electrically
Binary data
stream
11000110
1010
1001
Q
I
QPSK
constellation map
QPSK
signal
11
Q
Q
Signal
I
1 0 0 1
1 0 1 0
11 00 01 10
01
10 00
1
Demux Laser
/2
FIGURE 4. Transmitter setup for the modulation
of a QPSK signal.
1403lw_13 13 2/21/14 2:11 PM
PRODUCT SHOWCASE
SALES OFFICES
MAIN OFFICE
98 Spit Brook Road, LL-1,
Nashua, NH 03062-5737
(603) 891-0123; fax (603) 891-0574
GROUP PUBLISHER
Ernesto Burden
tel (603) 891-9137
ernestob@pennwell.com
ADVERTISING & SPONSORSHIPS MANAGER
Kathleen Skelton
tel (603) 891-9203; fax (603) 891-0587
kathleens@pennwell.com
DIRECTOR LIST RENTAL
Kelli Berry
tel (918) 831-9782;
kellib@pennwell.com
AUSTRIA, BENELUX, EASTERN EUROPE, FRANCE,
GERMANY, GREECE, PORTUGAL, SPAIN, SWITZERLAND
Holger Gerisch
49-(0)8856-8020228; fax 49-(0)8856-8020231
holgerg@pennwell.com
UK/SCANDINAVIA
Kathleen Skelton
tel (603) 891-9203; fax (603) 891-0587
kathleens@pennwell.com
ISRAEL
Dan Aronovic (Tel Aviv)
972-9-899-5813; fax 972-9-899-5815
aronovic@actcom.co.il
HONG KONG/CHINA/AUSTRALIA
Adonis Mak
852-2-838-6298; fax 852-2-838-2766
adonism@actintl.com.hk
JAPAN
Masaki Mori
81-3-321-93561
mori-masaki@ics-inc.co.jp
TAIWAN
Monica Liu
866-2-2396-5128
monica@arco.com.tw
These are the latest
products being featured
by Lightwaves partners.
For more information,
click on the link at the
end of each description.
For advertising information,
contact Kathleen Skelton
or Kelly Barker.
EQUIPMENT DESIGN
Senko Introduces New LC-HD
Connector for High Density
Applications
Senko
Senko Advanced Components, Inc. introduces
New LC-HD High Density LC Connector. The
SENKO LC-HD Connector is a totally unique design
which can increase panel density to its highest
possible. The low profile Senko LC-HD, together
with its pull tab allow LC adaptors to be stacked
with absolutely no vertical space. www.senko.com
NETWORKING
Leading the 3D light revolution
Optoscribe
Leading the 3D
light revolution
Optoscribes 3D MMux
Photonic Lantern is
the ideal waveguide
solution designed to optimise space division
multiplexing (SDM). To find out more, visit us at
OFC booth 4373, 9-13 March 2014, Moscone
Center, San Francisco. www.optoscribe.com
AFL ..........................................11
Agilent Technologies Inc. ........31
Amicra Microtechnologies
GmbH ..................................... 23
Anritsu Company...................... 4
Bristol Instruments Inc. ..........31
Corning Cable Systems ...........17
Diamond SA ............................. 7
ECRI Microelectronics .............21
JDS Uniphase Corporation ........ 2
JGR Optics ..............................12
M/A-Com Technology
Solutions Inc. ........................... 8
Optoscribe ...............................31
LW Summit Register .............. 32
Senko Advanced
Components Inc ............9, 13, 31
S.I. Tech ..................................31
Terahertz Technologies Inc. .....31
US Conec Ltd. .......................... 6
TEST AND MEASUREMENT:
New J-BERT M8020A accelerates
insight into designs up to 32 Gb/s
Agilent
The high-performance J-BERT
M8020A enables fast and accurate receiver
characterization of single and multi-lane devices. The
M8020A streamlines receiver test setup by providing
the highest level of integration, ensures accurate and
repeatable measurements by automating stressed
eye calibration, and supports interactive link training
for increased efficiency. www.agilent.com
EQUIPMENT DESIGN
Smart Decisions Start with Senkos
New Smart Cleaner
Senko
The newest member to SENKOs
product line is the Smart Cleaner,
a dry cloth cleaner specifically
designed to clean single fiber connections residing
in an adapter, faceplate or bulkhead. Made out of
anti-static material the Smart Cleaner is ideal for
working in sensitive environments and can be used
for 750+ cleanings. www.senko.com
TEST AND MEASUREMENT
Optical Wavelength Meters
Bristol Instruments
Absolute wavelength of
WDM lasers are measured
to 0.3 pm, guaranteed by
continuous calibration with a
built-in wavelength standard.
Reliable accuracy, a rugged
design, and affordability
satisfy the needs of the R&D scientist and the
manufacturing engineer. www.bristol-inst.com
TEST AND MEASUREMENT
New FTE7000A Touch Screen OTDR
Terahertz Technologies Inc.
TTI introduces the U.S. Made FTE7000A
OTDR with video scope including Pass/
Fail End face Grading Map. This touch
screen OTDR has a 36db range, 1m dead zone and
Bluetooth operation with any compatible Android
device. It displays traces in full graph, tabular event
table, and Fib-R-Map schematic modes. Please
visit our new website at www.teratec.us or call
888-U.S.-OTDRS for our latest offerings.
EQUIPMENT DESIGN
USB To Fiber Optic Bit-Driver
S.I. Tech
Supports USB 1.1 and 2.0 plug and play, multimode
or singlemode for secure communication, remote
security cameras, instruments, or other USB
devices, extended-distance. EMI/RFI immunity.
Order: 1.KIT #26 for
complete multimode
system. 11.3181/3182
for tempest version.
www.sitech-bitdriver.com
1403lw_31 31 2/21/14 2:12 PM
JULY 14-16, 2014
RENAISSANCE AUSTIN HOTEL
AUSTIN, TEXAS
During the exhibition and conference
learn from and dialogue with experts
from:
ADTRAN
ADVA Optical Networking
Alcatel-Lucent
Allied Fiber
Amphenol High Speed Interconnect
AT&T
Avago Technologies
BTI Systems
CALIENT Technologies
Calix
Ciena
Clearfeld Inc.
CommScope
Coriant
Corning Optical Communications
Cyan
Ethernet Alliance
EXFO
Facebook
Finisar
Fujitsu Network Communications
Infnera
Infonetics Research
IPC Systems, Inc
JDSU
JGR
Juniper Networks
LightCounting
Lonestar Education and Research Network
Luxtera
Marlin Equity Partners
MRV Communications
Oclaro
OneChip Photonics
Ovum
Plexxi
PMC-Sierra
Sumitomo Electric Lightwave
TE Connectivity
Tektronix, Inc.
The Ethernet Alliance
The Fiber Optic Association
The Light Brigade
TIA Fiber Optic Tech Consortium
TriQuint Semiconductor
Verizon
Walker and Associates
XKL, LLC
XO Communications
REGISTER EARLY AND SAVE Register by May 19, 2014 and SAVE $100
Register at: http://www.lightwavesummit.com/register.html
JULY 14, 2014
MONDAY
CARRIER INNOVATION TRACK ENTERPRISE/DATA CENTER INNOVATION TRACK
4:00 PM - 6:00 PM
WORKSHOP: Understanding Emerging
Technologies in Carrier Networks
WORKSHOP: Fiber Technology for Data
Center and Enterprise Networks
JULY 15, 2014
TUESDAY
CARRIER INNOVATION TRACK ENTERPRISE/DATA CENTER INNOVATION TRACK
8:30 AM - 9:30 AM PANEL: Innovation in an Era of Industry Consolidation
PANEL: Intra-Rack and Rack-to-
Rack Connectivity Innovation
9:30 AM - 10:00 AM Refreshment Break
10:00 AM - 11:00 AM
PANEL: Adding Flexibility and
Effciency to Optical Networks
PANEL: Is It Time to Consider Building
Your Own Network?
11:00 AM NOON Fujitsu Workshop Series: The xHaul Challenge
11:30 AM - 1:00 PM Delegate Lunch (for paid conference attendees only)
1:00 PM - 2:15 PM
KEYNOTE PRESENTATIONS (for paid conference attendees only):
t Dr. James King, Assistant Vice President Optical Platforms, AT&T Labs (invited)
t Steve Cheng, Principal Design Engineer and Architect,Telecommunications
and Data Cabling Infrastructure, Facebook
2:30 PM - 3:00 PM
PANEL: Enabling Transport SDN
PAPER: Classifying the New Data
Center Business Models
3:00 PM - 3:30 PM PAPER: 10/40/100GbE Migration: TheRole of Cabling
3:30 PM - 4:30 PM Refreshment Break
4:00 PM 4:45 PM
PANEL: State of 100G OTN Switched
Optical Networks in North America
PANEL: Network Security and
Optical Communications
JULY 16, 2014
WEDNESDAY
CARRIER INNOVATION TRACK ENTERPRISE/DATA CENTER INNOVATION TRACK
8:00 AM 9:00 AM Fujitsu Workshop Series: Executive Roundtable: Building the Network of the Future
9:00 AM - 10:00 AM PANEL: The Road to 1 Tbps PANEL: Data Center Design Innovations
10:00 AM - 10:30 AM Refreshment Break
10:30 AM - 11:00 AM
PAPER: OSNR Analysis in Nonlinear
Coherent Transmission Links
PAPER: How SDN Speeds Long Data
Flows from Virtualization
11:00 AM NOON Fujitsu Workshop Series: Does the Future Lie in 400G and Next Generation (NG) ROADM?
12:00 PM - 1:00 PM Delegate Lunch (for paid conference attendees only)
1:15 PM - 2:15 PM PANEL: Access Network Innovations PANEL: Optics in the LAN
2:15 PM - 2:45 PM Refreshment Break
2:45 PM - 3:15 PM PAPER: The ROADM Dilemma
PAPER: Driving Down Data Center Costs
through New Designs in Fiber Management
3:15 PM 4:00 PM PANEL: How Will IP/Optical Convergence Work, Really? PANEL: Innovation Pathways for Ethernet
4:00 PM 5:00 PM Fujitsu Workshop Series: Taking Advantage of the SDN Wave: Applications for Carriers
FOR DETAILS, VIEW
THE PRE-CONFERENCE
BROCHURE
FOR DETAILS, VIEW
THE PRE-CONFERENCE
BROCHURE
Join Us As We Foster the Conversation Between
Carriers, Data Center Managers, Systems
Developers, and Suppliers!
The pace of innovation and opportunity in optical communications is
accelerating. Now is the time for an in-depth exchange of views that
will highlight the critical innovations necessary to meet the near-term
challenges in optical networking.
The Lightwave Optical Innovation Summit Conference, brought to
you by Lightwave magazine and LightwaveOnline.com, will focus on
immediate challenges and provide innovative, practical solutions in
optical communications in the Americas.This is a must attend event
for carriers, data center and enterprise network managers, systems
developers, and suppliers of components and subsystems.
KEYNOTE SPEAKERS:
CARRIER TRACK KEYNOTE
Dr. James King
Assistant Vice President - Optical Platforms (invited)
Dr. King has served in a variety of positions since he joined AT&T
Labs in 1985. Major accomplishments include support for the initial
deployment of TCP/IP networks into AT&T central offces; core back-
bone expansions to OC-48 (1999), OC-192 (2001), and OC-768 (2007);
engineering design and implementation of AT&Ts CNI network (2001);
and design and business case development for deployment of AT&Ts
ultra-long-haul network (2005).
ENTERPRISE/DATA CENTER TRACK KEYNOTE
Steve Cheng
Principal Design Engineer and Architect,Telecommunications
and Data Cabling Infrastructure
Steve Cheng has more than 15 years of experience in data center
and telecommunications design, construction, and operations. He
has applied this experience as well as cutting-edge technology,
custom and innovative design solutions, and a focus on effciency
and modularity to develop the next level of structured telecommunica-
tions cabling to meet Facebooks growth and demand.
WWW. LI GHTWAVESUMMI T. COM WWW. LI GHTWAVESUMMI T. COM
LWOptSpd_ATDRM_140213.indd 1-2 2/13/14 1:57 PM