cpqd at optical communication ecosystem - last/next 10 years and r&d&i opportunities
TRANSCRIPT
Last/Next 10 years and R&D&I opportunitiesCPqD at Optical Communication Ecosystem
14/05/2016Juliano R. F. de Oliveira, Ph.D
Optical Communication Division Manager
AGENDA
• Motivation
• Optical communication history @ CPQD
• Last 10 years of optical communications R&D&I @ CPqD
• Last/Next 10 years technological standpoint
• Next 10 years of optical communications R&D&I @ CPqD Roadmap
• Optical Hardware, Algorithms, Networks and Emerging technologies;
• Remarks
MOTIVATION
• Today advanced communications demand
• Industries, business/banks, transportation, people and global knowledge;
• Fiber capacity crunch
• SSMF Tx/Rx of 100Tb/s hero experiments, (limit ~200Tb/s due nonlinear effects);
• CMOS ASICs limitations
• Processor frequency Moore law (i7 runs at 3.5GHz), CMOS process at 14nm (limit of 0.5nm gate size is coming);
• Tomorrow advanced communications demand
• Medical diagnosis and treatment, traffic safety and Internet of Things;
• New solutions have to meet not just capacity demand, but all economical perspective
• More capacity, reduced cost/bit, increased energy efficiency, smaller sizes;
• Emerging Technologies: Integrated electronics & photonics convergence;
MOTIVATION: TOMORROW DEMAND
M2M and IoT Smart
Cities Market
▪ Devices: +capacity, -cost/bit, +energy efficiency, -sizes;▪ Integrated electronics & photonics convergence;▪ Submarine, long-haul, metro and data center networks;
OPTICAL COMMUNICATION R&D&I HISTORY AT CPQD
• 1976 – CPqD Birth (Telebras national carrier action)
• 1976-1997 – R&D&I for import substitution
• Lasers, photodetectors, TIA, thin film devices, optical amplifiers, optical networks, etc;
• 1997-2001 – CPqD Privatization (inter. competitive integration)
• Creation of companies (optical communication: PADTEC);
• 2002-2005
• Optical communication technological transfer to PADTEC (2.5G transponders, static EDFAs and fixed WDM systems);
LAST 10 YEARS: OPTICAL COMMUNICATION R&D&I AT CPQD
2005-2010
• Optical transmission 10G, 40Gb/s (OOK), RS-FECs FPGA IPs;
• Optical amplifier (EDFA) for dynamic WDM networks (AGC) with transient suppression;
• Intelligent WDM System monitoring (IWDM);
• Eye diagram (BER) asynchronous estimation;
LAST 10 YEARS: OPTICAL COMMUNICATION R&D&I AT CPQD
2005-2010
• WDM system optimization
• Manual system alingment;
• Dispersion management;
• Optimized fiber WDM net.;
• Optical cross-connect (OXC) sub-systems;
• Planar lightwave circuit (PLC) ROADMs (for ring topologies)
LAST 10 YEARS: OPTICAL COMMUNICATION R&D&I AT CPQD
2010-2015
• Hybrid optical amplifiers
• EDFA + Dist. 1st orderRaman
• WSS ROADM sub-systems
• Linecards
• CDC add/drops
• Mesh networks architecture
LAST 10 YEARS: OPTICAL COMMUNICATION R&D&I AT CPQD
2010-2015
• Tx/Rx Coherent DSP algorithms 100/200/400Gb/s (DP-QPSK, DP-16QAM, and400G DP-64QAM)
• Tx DSP: pulse shapping, econders, interpolators
• Rx DSP: Deskew, IQ imb., CD equalizer, time recovery, dynamicequalizer, phase andfrequency estimation
LAST 10 YEARS TECHNOLOGICAL STANDPOINT
• System and subsystem Research/Development/Innovation
• Optical design and optimization IP developed @ CPqD(applied research to innovation);
• Optics and electronics devices from international companies;
• Economical standpoint
• Very fragmented market (lots of device companies);
• No local manufacturers (taxes benefits available);
• Economic growth (local R&D big investment);
NEXT 10 YEARS TECHNOLOGICAL STANDPOINT
• System and subsystem Research/Development/Innovation
• Optical design and optimization must be continually improved @ CPqD (state-of-the-art knowledge to innovation);
• Optics and electronics must be co-designed @ CPqD and local/international partners (BrPhotonics emerges);
• Economical standpoint
• Verticalized market (improve aggregated value);
• Local manufacturers (taxes benefits available, more competitive);
• Low economic growth (foreign R&D investment);
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Hardware – Fibers (systems)
• Multi-Core fiber (MCF)
• Inter core interference
• Few-mode fiber (FMF)
• Coupled core fiber (CCF);
• Hybrid between MCF & FMF;
• Photonic bandgap fiber (FBGF)
• 0.1% SSMF non-linearity, and ultimately low loss (<0.1dB/km) @ 2000nm λ;
doi:10.1088/2040-8978/18/6/063002
From: Roadmap of optical communications
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Hardware – Amplifiers
• EDFA (low cost, high energy efficiency, spectral limitation)
• Raman (high cost, low energy efficiency, no spectral limitation)
• Hybrid Amplifiers (Raman/EDFA)
• Efficient @ new modulation formats
• Dynamic operation (SDN)
• Faster controls (no power transients);
• Rare earth amplifier for FBGF
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Hardware – SDM
• Spatial division multiplexing is required to attend 40% traffic growth up to 2024 (1Pb/s);
• Arrayed integration
• Transponders, amplifiers, ROADMs, splices/connectors/ fibers;
• Initially amortize overhead cost and energy w/o sacrifice performance;
• Performance issues (induced xtalkin arrayed transceivers);
doi:10.1088/2040-8978/18/6/063002
From: Roadmap of optical communications
Integrated photonics arrayed transciever
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Hardware – Coherent transceivers• Optical modulator (LiNbO3, GaAs, InP,
or Si) non-linear effects understanding• Novel modulator design• Co-design modulator/DSP to maximize
performance
• Implementable non-linear compensation (power and area optimize);• Newer DSP or optimized ASIC IP• DBP, Volterra and evaluate NFT
• Integration DSP+FEC+non-linearcompensation (power optimized)
NEXT 10 YEARS R&D&I CPQD ROADMAP• Algorithms – Modulation formats
• Submarine and long-haul (coherent)• Performance (spectral efficiency X
distance product)
• Low implementation penalty
• Flexible transponders (rate adaptive codes/symbol/flow)
• Modulation format design (coded) aim NL mitigation;
• Inter/Intra data center (PM IM/DD)• Application dependent trade-off:
devices cost, power, simplified DSP (same access networks challenges)
56G PAM4
Submarine and long-haul with recirculation loop
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Algorithms – DSP• Bring actual systemic IPs to ASIC
(RTL and GDSII)• Power optimization to meet metro-
data center demands;• Maximum performance for
submarino/long-haul applications
• OTN and SD-FEC integratedtogether with DSP (poweroptimized)
• Co-design DSP + photonics (cost)• Best (sub) and non-ideal devices
(data-center);
• Intelligent/elastic DSP (dynamic)
DSPOTN
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Networks – long haul
• Optical automatized links (SDN like)
• Layer-0 optimization algorithms• Global WSS
equalization scheme
• Optical links auto-alignment strategies• OSNR, BER, NL
mitigation;Homemade Network
Elements
5 dB
WSS Global equalization
1
2
TOTAL
N
i
i TAA
][minTOTAL wAA TOTALw
Apply Γ
T ≤ allowed tilt?
EndYes
No
Given: N ≥ 2, W, A{1, …, N - 1} , T
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Silicon photonics
• Photonics IPs design & validation for basicbuilding blocks• Coherent receiver (3rd tapeout)
• Coherent transmitter (2nd tapeout)
• Unit cell optics (laser+tx+rx)
• Narrow LW Laser (InP + cavity design)
• Hybrid III/V-Si integration for tunnable lasers
• Hybrid Polymer (active)-Si (passive) integration for high speed + low loss + smallsize modulators
• Fabless model => great flexibility
Coherent Receiver
Coherent Transmitter
Narrow LW Laser
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Silicon photonics - Alingment• Edge coupling spot size
converter;• 1 dB tolerance misalignment
(+/- 2 um);• Wavelength independent;
• Butt coupling for hybrid III-V laser integration with passive Si photonics
• Fiber array v-groove strategies to minimize quantity of fiber alignment
2 3 4 5 6-2.5
-2
-1.5
-1
-0.5
0
X: 3.1
Y: -0.4007
Fiber MFD [m]
Tota
l eff
icie
ncy [
dB
]
TE
TM
high-efficiency edge couplers in 300-nm SOI
-4 -2 0 2 4-4
-2
0
2
4
-4 -2 0 2 4-4
-2
0
2
4
TE (log) TM (log)
NEXT 10 YEARS R&D&I CPQD ROADMAP
• Silicon photonics - Packaging• Need to leverage what has
been done in electronics industry
• BGA + optical port can be a enabler• Size reduction (smaller io)• Higher integration• Uncooled/non-hermetic• Proximity optics/electronics
• Challenge: high temperature fiber attach (~270º)
+
REMARKS
• CPqD must be prepared for the verticalization era of opticalcommunications (partner technology support)
• R&D in optical system and sub-system optimization IP;
• SDN coherent transcievers, DSP/FEC evolution;
• Co-design integrated photonics & electronics;
• Optical system & sub-system design/optimization
• D&I regarding the basic devices are needed to this new era
• Integrated photonics modulator, receiver and laser devices;
• ASIC IPs for OTN, DSP, FEC and non-linear mitigation (poweroptimized);