1 / 95

Optical Networks

Optical Networks. Josef Vojtěch josef . vojtech/zavinac/ cesnet.cz. www.ces.net. czechlight.cesnet.cz. Optical Networks Outline. History Transmission Fibers Attenuation Limited Reach - Amplifiers Doped fiber – EDFA, TDFA Raman SOA OSNR Limited Reach

danil
Download Presentation

Optical Networks

An Image/Link below is provided (as is) to download presentation Download Policy: Content on the Website is provided to you AS IS for your information and personal use and may not be sold / licensed / shared on other websites without getting consent from its author. Content is provided to you AS IS for your information and personal use only. Download presentation by click this link. While downloading, if for some reason you are not able to download a presentation, the publisher may have deleted the file from their server. During download, if you can't get a presentation, the file might be deleted by the publisher.

E N D

Presentation Transcript

  1. Optical Networks Josef Vojtěch josef.vojtech/zavinac/cesnet.cz www.ces.net czechlight.cesnet.cz Optical Networks

  2. Optical NetworksOutline • History • Transmission Fibers • Attenuation Limited Reach - Amplifiers • Doped fiber – EDFA, TDFA • Raman • SOA • OSNR Limited Reach • Impairments (CD, PMD, non-linearities) • WDM Transmission (WDM, DWDM, CWDM) • Present and Future Transmission Systems • Capacity Limits, SDM • Photonic Services, Precise Time Transfer OpticalNetworks

  3. Optical NetworksA Little Bit of History (1) Back to Antiquity (mirrors, firebeacons, smokesignals) [1] Till the end 18thcenturywithlamps, flags 1792 – Claude Chappewithmechanical „optical“ telegraph 1830 – the advent oftelegraphy (e.g. Morse) Submarinecablesstartedbeforethefirsttransatlantic in 1858 but was not working long, shortmessage (one sentence) fromQueen Victoria to President Buchanantook 30 hours! Initialassumptions: 3 words per minute but in reality lessthan 1 word per minute, only 20 daysthancablefailed, 723 messages! 1866 itssucessor Terestrial long haul in Russia, USA – telegraphalongrailways, submarinepreferable Optical Networks

  4. Optical NetworksA Little Bit of History (2) • 1876 – the invention of telephone (A.G.Bell, U.S. Patent No. 174 465) • 1895 – radio by Marconi, big competition started • 1940 – massive increase of pairs installed - 3 MHz coax-cable system (repeater spacing 1 km), high freq. dependent loss especially above 10MHz • 1948 – 4 GHz microwave system • 1975 – last most advanced coaxial system with 274Mbit/s Optical Networks

  5. Optical NetworksA Little Bit of History (2) • Evolution of optical communication systems • 1960 – the invention of laser (suitable transmission medium?) • 1960s – optical fibre 1000 dB/km, 1970 20dB/km close to 1um • 1st generation GaAs lasers 850nm, 10km regeneration • 1980 – 45 Mb/s (1st generation) multi-mode fibres • 2nd generation 1310 nm • 1980s – 1310 nm, 1 dB/km, 100 Mb/s, multi-mode fibres • Late 1980s – 2 Gb/s, single mode fibres, repeater spacing 50 km (2nd generation) • 3rd generation 1550 nm • 1990s – 1550 nm (problem with lasers, dispersion of fibres, typical repeater spacing 60 -70km), 2.5 Gb/s or 10 Gb/s (3rd generation) Optical Networks

  6. Optical NetworksA Little Bit of History (3) • 4th generation 1550 nm – WDM - DWDM, CWDM • Optical amplification, EDFAs developed late 1980s • 1990s - DWDM, optical amplification (4th generation) • Today comercially available 100G waves each consumes 50GHz~0.4nm, 160 channels in C+L bands ie 1,6 Tb/s, thousands of kilometers, coherent technology Optical Networks

  7. Optical NetworksOptical Fibres - MM • 1970: 20 dB/km (Kapron, Keck, Maurer), silica fibres • Multimode (MM), singlemode (SM) and fewmode (FM) fibres • Multimode: step-index (SI) or graded index (GI) • MM SI: modal dispersion: different rays disperse in time because of the shortest (L) and longest (L/sinΦC) paths • MM SI: 10 Mb/s • MM GI: parabolic index, lower modal dispersion, higher bit rates, 100 Mb/s • Plastic MM GI, for 1 GE (or even 10 Gb/s) • Attenuation: 1 – 4 dB/km, <10 dB/km for plastic Optical Networks

  8. Optical NetworksOptical Fibres Optical Networks

  9. Optical NetworksOptical Fibres - SM • Single mode (SM, Standard SMF, G.652) fibres • Supports only one so called „the fundamental mode of the fibre“, all higher modes are cut off @ the operating wavelength • An optical mode refers to a specific solution of the wave equation (satisfies boundary conditions, spatial distribution is constant as light travels along a fibre) • The cutoff wavelength is specified in ITU G.650, SSMF cutoff approx. 1200 nm • 0.2 dB/km@1550 nm, 0.4 dB/km@1310 nm • Best conventional fiber ever 0.1484dB/km Optical Networks

  10. Optical NetworksOptical Fibres – SM types • ITU G.651 – MMF, new types suitable for 10G • G.652 SSMF, non-dispersion-shifted fiber, optimized for 1310nm (but well usable in 1550nm and 100G coherent - big core diameter) • G.653, dispersion-shifted optical fiber, zero dispersion at 1550nm, strong nonlinear effects (FWM) • G.654, low loss, no longer supported • G.655, nonzero dispersion-shifted, small dispersion at 1550nm • G.656, for S, C and L bands • G.657, G.652-like Optical Networks

  11. Optical NetworksOptical Fibres - Special • Multicore fibres (3-20) • Photonic crystal fibres PCF • Refractive index experienced (“seen”) by propagating light depends on size and shape of air holes in every point of structure • But speed of light similar to 300 000km/s (air), perhaps promising for financial transactions Optical Networks

  12. Optical NetworksOptical Reach -Transmitter • Conversion of electrical signals into optical streams • Light source LED or laser: Fabry-Perot (FP), Distributed Feedback (DFB) • Modulation: direct or external (10 Gb/s, DWDM incl. tunable) • Output powers typicall: 0 dBm – 4 dBm OpticalNetworks

  13. Optical NetworksOptical Reach • Tx power 1mW ~ 0dBm (P [dBm] = 10 log P[mW]) • Receiver PIN diode or APD • Minimal needed Rx power– sensitivity, examples: • -45 dBm (32nW) @ 155Mbps – 180km • -36 dBm @ 1GBps – 144km • -24 dBm @ 10Gbps – ? • Reach extension – regeneration / optical amplification Optical Networks

  14. Optical NetworksOptical Amplifiers • Rare earth doped fibre, Semiconductor (SOA), Raman • Erbium Doped Fibre Amplifiers (EDFA) • Really began a revolution in the telecommunications industry • Late 1980s, Payne and Kaming (University of Southampton) • OAs can directly amplify many optical signals • Analogous principle: Protocol, bit-rate transparent • EDFAs working in the 1550 nm window (C band and later L band) Optical Networks

  15. Optical NetworksOptical Amplifiers EDFA High energy level, 1 μs Metastable energy level, 10ms 1480 nm Amplified signal 980 nm Low energy level • Light Amplification by Stimulated Emission Er atoms Optical Networks

  16. Optical NetworksOptical Amplifiers EDFA Er doped fibre, 10 m – 100 m SIGNAL SIGNAL WDM WDM PUMP PUMP • Forward, backward pumping • Forward: lowest noise • Backward: highest output power • 980 nm: low noise • 1480 nm: stronger pump sources (req. longer Er fibres) • 1480 nm & backward; 980 nm & forward • Single or dual stage (for DCF) Optical Networks

  17. Optical NetworksOptical Amplifiers EDFA • Output powers (5 Watts or more) • Gain (30 dB), not uniform across C (L) band • Input power (-36 dBm) • Noise Figure (NF): 3.6, typ 5-7dB, teoretical minimum 3 dB • ASE • For L band: long Er fibres (> 100 m) • Booster, in-line, preamplifiers Optical Networks

  18. Optical networks and equipmentOAs and Practical Deployment Praha - Pardubice Since May 17, 2002 Now gone Masarykova Univerzita, Brno

  19. Optical networksReal deployments • OSA (One Side Amplification) – all components are located at one place, for star topologies • Praha – Plzeň (OSA), 123 km, 34 dB, 1 GE Optical Networks

  20. Optical networks and equipmentOAs and Practical Deployment • Brno – České Budějovice, 308 km, 70 dB, 2,5 Gb/s, PoS, without optical filters Masarykova Univerzita, Brno

  21. Optical networks and equipmentOAs and Practical Deployment • Brno – Ostrava, 235 km, 51 dB, 1 GE (tested for 2,5 G PoS), without optical filters Masarykova Univerzita, Brno

  22. Optical NetworksOther Optical Dopped Fibre Amplifiers – O band • PraseodymiumDoped Fluoride FibreAmplifier (PDFA) • 1310 nm, not as energyefficientcompared to EDFA, higher NF • Problemswith fluoride fibres, not very widespread • Neodymium DFA • 1310 nm, fluoride fibre OpticalNetworks

  23. OpticalNetworksOtherOpticalDoppedFibreAmplifiers – S band S band • Thulium DFA (TDFA), but poor efficiency in silica, due to high non-radiative decays • TDFAs based on fluoride fibres commercially available • 1460-1500 nm 19dB gain and 19dBm output • Can’t be spliced with silica fibres, hydroscopic • Double pumped TDFA in silica with dopants – not commercially available Er Tm Optical Networks 19th Nov 2013

  24. OpticalNetworksOtherOpticalDoppedFibreAmplifiers – S band S band cont. • Modified EDFA • Can handle region about 1510-1530nm, 20dB gain, commercially available, NF higher that TDFA • Commercially unavailability of DWDM transceivers forthisregion Optical Networks 19th Nov 2013

  25. Optical NetworksOther Optical Dopped Fibre Amplifiers 2um (2000nm) band • Experimental TDFA operation • Bandwidth 1910-2020nm • Small signal gain 35dB • NF 5-7dB • Silica glass • Li Y. et al,”Thulium-doped Fiber Amplifier for Optical Communications at 2μm”, OFC2013 Optical Networks 19th Nov 2013

  26. Optical NetworksOther Optical Dopped Fibre Amplifiers 2um (2000nm) band cont. • Hollow core photonic bandgap fibers • Minimal latency, neff close to 1, propagation speed almost c, not 2/3c • Nonlinear coefficient 0.07% of SSMF • Losses still in order of 4.5dB/km in 2um • Theoretically losses can go down to 0.15 dB/km • 1966 Charles K. Kaofiber • 1000dB/km • Source: NKT Photonic Optical Networks 19th Nov 2013

  27. Optical Networks Semiconductor Optical Amplifiers • Based on conventional laser principles • Active medium (waveguide) between N and P regions • + InGaAsP – small and compact components • + Cost effective solutions, especially for O and S bands • + High gain (up to 25 dB) • - Low output powers (15 dBm) • +Wide bandwidth • - High noise figure (7-8 dB) • -- Very short life time - cross gain effects Optical Networks

  28. Optical NetworksSemiconductor Optical Amplifiers Optical Networks

  29. Optical NetworksRaman Amplification • Both discrete but mainly distributed amplifier • Stimulated Raman scattering effect • Distributed amplification, a communication fibre itself is a gain medium • Can add 40 km to increase a maximum transmitter-receiver distance • Upgrading of existing links to add more channels • A quite weak effect in silica fibre – very high powers have to be used • Safety problems (automatic laser shutdown - ALS) Optical Networks

  30. Optical NetworksRaman Amplification • Counter-directionally pumping schemes Optical Networks

  31. Optical NetworksNoise Limited Reach Optical Networks

  32. Optical NetworksNoise Limited Reach Receiver needs some minimal OSNR to achieve intended BER (e.g. 1E-12) Scenario 1: Span 80 km, Gain 18 dB, NF 6 dB, Pin 0 dBm. Scenario 2: Span 100 km, Gain 22 dB, NF 6.5 dB, Pin 0 dBm. Optical Networks

  33. Optical NetworksOptical Fibres – dispersion • MM: Intermodal dispersion (pulse broadening, the most important limiting factor) • SM: Intermodal dispersion is absent, pulse broadening is present still because of Intramodal dispersion (or Group-velocity dispersion CD), even laser pulses have finite spectral width and pulses are modulated • CD: different spectral components of the pulse travel at different speeds • Increases as the square of the bit rate Optical Networks

  34. Optical NetworksOptical Fibres – CD • D = DM + DW • Material dispersion • Waveguide dispersion Optical Networks

  35. Optical NetworksOptical Fibres – CD • G.652 (SSMF): Zero dispersion at 1310 nm • G.653 (DSF): Zero dispersion at 1550 nm • G.655 (NZDSF): Small dispersion at 1550 nm, positive/negative • Dispersion Compensating fibres (DCF) Optical Networks

  36. Optical Networks Chromatic Dispersion Compensation • Typical values (receivers can have different tolerance to CD!) Optical Networks

  37. Optical NetworksChromatic Dispersion Compensation • Dispersion compensating fibres (DCF) • A special kind of fibre, compensates all wavelengths (the only solution for „grey“ transmitters) • Adds link loss (and money), especially for long-haul applications • Stronger non-linear effects (due to a smaller core diameter) • Fibre Bragg gratings (FBG) • Typically narrow-band elements – a stabilized DWDM laser is a must • „Wide-band“ FBGs available today (for C or L band) • Signal filtering, spectrum shaping, tuneable compensators • Cost effective solution • Electronic post and pre compensation, GTE, VIPA, MZI • Tunable Optical Networks

  38. Optical nNtworksChromatic Dispersion Compensation FBG Optical Networks

  39. Optical Networks Chromatic Dispersion Compensation (FBG) Optical Networks

  40. Optical NetworksOptical Fibres – PMD • Polarization Mode Dispersion (PMD) • The stochastic phenomenon • Fibre stress, temperature, imperfections • The fundamental mode has two orthogonally polarized modes • The two components with different propagating speeds disperse along the fiber • The difference between the two propagation times is known as the Differential Group Delay (DGD) • PMD is a wavelength averaged value of DGD Optical Networks

  41. Optical NetworksOptical Fibres – PMD ITU proposed PMD values Optical Networks

  42. Optical NetworksOptical Fibres – PMD • PMD is measured and quoted in ps for a particular span and discrete components but its coefficient is in ps/(km)1/2 • PMD accumulates as the square root of distance of a link • A single span with high PMD dominates the total PMD for the whole network • A big issue for older fibres (late 1980s, 80 000 000 km) and higher bit rates (10 Gb/s and more) • Modern fibres have PMD of less than 0,5 ps/(km)1/2 • Difficult to compensate in optical domain • In coherent systems compensated in DSP Optical Networks

  43. Optical NetworksNonlinear Optical Effects • When an intesity of electromagnetic fields becomes too high, the response of materials becomes nonlinear • For optical systems, nonlinear effects can be both advantageous (Raman and Brillouin amplification) and degrading (Four Wave Mixing, Self Phase Modulation and Cross Phase modulation) Optical Networks

  44. Optical NetworksNonlinear Optical Effects - SRS • Stimulated Raman Scattering (SRS) • A signal is scattered by molecular vibrations of fibre – optical phonons • Can occur both in forward and backward directions • Shifted to longer wavelengths (lower energy) by 10 to 15 THz in the 1550 nm window • Wide bandwidth of about 7 THz (55 nm) • Maybe used for amplification (Raman fibre amplifiers), so called counter directionally pumping schemes • In DWDM systems: transfer of power from shorter wavelengths to longer ones Optical Networks

  45. Optical NetworksNonlinear Optical Effects - SBS • Stimulated Brillouin Scattering (SBS) • A signal is scattered by sound waves – acoustic phonons • Shifted to longer wavelengths (lower energy) by 11 GHz in the 1550 nm window • Narrow bandwidth of about 30 MHz • A problem for monochromatic unmodulated signals Optical Networks

  46. Optical Networks Nonlinear Optical Effects • Self-Phase Modulation (SPM) • When the intensity of the signal becomes too high, the signal can modulate its own phase • The refractive index is no longer a constant • Significant for fibres with small effective areas (G.655, DCF) • Higher bit rates (10 Gb/s and higher) Optical Networks

  47. Optical NetworksNonlinear Optical Effects (SPM1) Pin = 16,5 dBm Optical Networks

  48. Optical NetworksNonlinear Optical Effects (SPM2) Pin = 22,7 dBm Optical Networks

  49. Optical NetworksNonlinear Optical Effects (SPM3) Pin = 25,8 dBm Optical Networks

  50. Optical NetworksNonlinear Optical Effects (SPM4) Pin = 30,1 dBm Optical Networks

More Related