Advanced Fibre Testing: Paving the Way for High-Speed Networks Trevor Nord Application Specialist JDSU (UK) Ltd
Fibre Review Singlemode Optical Fibre
Elements of Loss Fibre Attenuation - Caused by scattering & absorption of light as it travels through the fibre - Measured as function of wavelength (db/km) Bending Losses - Microbending losses are due to microscopic fibre deformations in the core-cladding interface - Macrobending losses are due to physical bends in the fibre that are large in relation to fibre diameter
Absorption and Scattering Absorption: Light is absorbed due to chemical properties or natural impurities in the glass. Accounts for about 5% of total loss. O Si Si O O Imperfections Pure Glass=Si O 2 Scattering is the loss of a light signal from the fiber core caused by impurities or changes in the index of refraction of the fiber. Accounts for about 95% of total loss. O Si Si Si O Cu Impurities Light scattered
Useful optical phenomena OTDR depends on two types of phenomena: - Rayleigh scattering - Fresnel reflections. Rayleigh scattering and backscattering effect in a fibre Light reflection phenomenon = Fresnel reflection
Optical Return Loss (ORL) - The ORL is the amount of transmitted light reflected back to the source - The ORL is measured in db and is a positive value. - The higher the number, the smaller the reflection. Examples: +50dB is a good value +17dB is a bad value!! - Increase transmitter noise Reducing the OSNR in analog video transmission Increasing the BER in digital transmission systems - Increase light source interference Changes central wavelength and output power - Higher incidence of transmitter damage SC - PC SC - APC Angled connectors (APC) reduce the back-reflection
Chromatic Dispersion In telecom transmission, a pulse of light is composed by multiple wavelengths (colors) which travel at different speed, causing chromatic dispersion Pulse Spreading λ 1 λ 5 λ 4 λ 3 λ 2 Chromatic Dispersion In DWDM transmission, each wavelength travels at a different speed. This can have an impact on number of channels being carried, the channel spacing and the total link length
Effect - Transmission Speed and Distance Limitations Input pulses Low Data Rate Output pulses The variation of index according to the wavelength broadens the pulse and causes it to interfere with neighboring pulses 1 0 1 1 0 1 Higher Data Rate Digital pulses overlapping 1 0 1 0 1 0 1 1 0 1 0 1 0 1 CD starts being a problem at 10 Gbit/s especially for G652 fibers and is critical at 40Gbit/s Data Rate Max. Distance for G.652 at 1550nm Max. Distance for G.655 at 1550nm 2.5 Gbit/s 980km 2770km 10 Gbit/s 61km 175km 40 Gbit/s 3.8km 10.8km
Fibre Types and Compensation 20 Examples of Chromatic Dispersion values at 1550nm: - G.652 : ~ 17 ps/(nm.km) - G.653 : 0 ps/(nm.km) - G.655 : ~4 ps/(nm.km) 20 10 0-10 -20 1200 1300 1400 1500 λ (nm) As CD is a linear parameter, intrinsic to the fibre, it can be compensated. There are different techniques which enable compensating CD (such as Dispersion Compensating Fibre with negative CD values) Dispersion (ps/nm.km) SMF + DCF λ (nm) 1300 nm 1550 nm DCF Example of a dispersion compensation result on a SMF fiber
Chromatic Dispersion Fibre Classification ITU-T Fibre Type description Zero wavelength Dispersion at 1550nm Dispersion slope @1550nm G.652 Non Dispersion Shifted fibre 1310 1324 nm 17 ps/nm.km 0.057 ps/nm².km G.653 Dispersion Shifted fibre 1500 1600 nm 0 ps/nm.km 0.07 ps/nm².km G.655 Non zero Dispersion Shifted fibre Negative Non zero dispersion shifted fibre Not specified but ~1450-1480 nm G.656 Broadband NZDSF 8 ps/nm.km 4 ps/nm.km 0.045 to 0.1 ps/nm².km -5ps/nm.km 0.05 to 0.12 ps/nm².km
Polarization Mode Dispersion - Although we refer to fibres as singlemode, in reality the light energy is carried along in different modes known as polarization modes - PMD is caused because the fibre is birefringent (Silica, the base material of optical fibres is birefringent): Two principal polarization modes travel at different speed Slow axis PMD is caused by geometrical irregularities of the fiber - Elliptical and Assymetries Core stress Cladding eccentricity Elliptical fiber design Optical fiber Fast axis PMD depends on mechanical stress - Bends and Twists Fiber twist Fiber stress Fiber bend
Birefringent Crystal
Effect - Transmission Speed and Distance Limitations V 1 DGD V 1 V 2 V 2 Stress!!! ( bending, twisting) Temperature v 2 DGD Slow Fast v 1
Effect - Transmission Speed and Distance Limitations - DGD is a randomly varying quantity. It varies significantly with wavelength and times (changes may be introduced by vibration, movement or changes in temperature ) - DGD varies so much that a parameter has been defined: the PMD coefficient (average value of DGD along a range of conditions) PMD is directly linked to bit interval error 1 1 0 1 1 1 0? 1 For a PMD value of 0.5ps/ km, max distance: - 6400 km @2.5Gbit/s - 400km @10Gbit/s - 25km @40Gbit/s PMD puts limits on the transmission speed of the networks PMD puts limits on transmission distance of the networks
Limiting Fibre Parameters The standards recommend that the maximum admissible PMD delay is 10% of the bit length. SONET Transmission Rate Bit Time PMD Limit STM-4 622 Mb/s 1600 ps 160 ps STM-16 2.50 Gb/s 400 ps 40 ps STM-64 10 Gb/s 100 ps 10 ps STM-256 40 Gb/s 25 ps 2.5 ps PMD sensitivity - If the PMD measurement is bad, then the fibre is considered as PMD sensitive. Simple transmission shall apply to this fibre, as no compensation exists so far - If the PMD measurement is good, then the fibre cannot be considered as PMD insensitive or non-pmd sensitive
DGD variance over time and wavelength
Coping with PMD OFC 2007 survey data showed that about 25% of UK fibre had PMD greater than 0.5ps/ km Testing and record keeping If you do not know the size of the issue, it should be measured Agile Optical Network - Dynamic Channel Allocation If a specific wavelength is suffering because of excessive DGD then it may be possible to move the traffic to a different channel with better DGD performance or even to a different route Polarisation scrambling and FEC Covers all channels Scrambling reduces the impact of the DGD to reduce outages Remaining errors fixed by FEC Optical PMD Compensators Use polarisation controllers with a control/feedback loop
Why measure the Spectral Attenuation Profile of a fibre? The purpose of the spectral attenuation (AP) measurement is to represent the attenuation as a function of the wavelength. Historically, this measurement was required mainly for long-haul applications. With the increase of CWDM deployment and the extension of the DWDM wavelength range, it is becoming necessary to have a clear picture of the fiber attenuation other the wavelengths intended to be loaded with traffic.
Characterising the full wavelength range In long distance transmissions, as well as at very high bit rate (10G, 40G systems), Raman amplifications are becoming more widely used across the whole spectrum. In addition, distal pumping of Erbium amplifiers at 1480 nm are currently deployed. Characterising fibre at the pump wavelengths (1420, 1450 nm, 1480 nm, etc.) is of high interest to ensure amplification will occur along the required distance.
Characterising Fibre Plant Understanding Fibre Link and Network Characterisation
What is Fibre Characterisation? Simply the process of testing optical fibres to ensure that they are suitable for the type of transmission (i.e. WDM, SDH, OTN, Ethernet) for which they will be used. The type of transmission will dictate the measurement standards used Transmission type Speed PMD Max CD Max SDH 10 Gbs 10 ps 1176ps/nm Ethernet 10 Gbs 5 ps 738 ps/nm SDH 40 Gbs 2.5 ps 64 ps/nm
When should I consider testing? Std CD PMD - New cable Installs (10G+) - Planning / Upgrading existing plant - Following purchase or lease of existing plant - Periodic tests over time - After cable maintenance - Determine if dispersion compensation required
Link & Network Characterisation Link Characterisation It measures the fibre performance and the quality of any interconnections The suite of tests mostly depend on the user s methods and procedures It could be uni- or bi-directional Tests Connector Inspection, IL, ORL, OTDR, PMD, CD, AP Network Characterisation It provides the network baseline measurements before turning the transmission system up. Network Characterisation includes measurements through the optical amplifiers, dispersion compensators, and any elements in line. It is a limited suite of tests as compared to Link Characterization ROADM Point A Point B Router DWDM Optical Networ k Optical Amplifier Video Headend Optical Amp. CWDM/DWDM Optical Network
Testing the Fibre Plant @ On @ Charge Connector inspection Insertion Loss Optical Return Loss OTDR Polarization Mode Dispersion Chromatic dispersion Spectral Attenuation
Contamination and Signal Performance 1 CLEAN CONNECTION Fibre Contamination and Its Effect on Signal Performance Back Reflection = -67.5 db Total Loss = 0.250 db 3 DIRTY CONNECTION Clean Connection vs. Dirty Connection This OTDR trace illustrates a significant decrease in signal performance when dirty connectors are mated. Back Reflection = -32.5 db Total Loss = 4.87 db
d B M e n el u Measuring Insertion Loss (IL) The insertion loss measurement over a complete link requires a calibrated source and a power meter. This is a unidirectional measurement, however could be performed bi-directionally for operation purposes Calibrated Light Source Optical Power Meter d W B m >2s Perm Ca nc dw B d m B P t P r It is the difference between the transmitted power and the received power at each end of the link
Measuring Optical Return Loss (ORL) The 2 predominant test methods are available: Optical Continuous Wave Reflectometry (OCWR) An Optical Return Loss Meter (composed of a laser source and a power meter on the same test port) Optical Time Domain Reflectometry (OTDR) The OTDR is able to measure not only the total ORL of the link but also section ORL (cursor A B) OCWR Method OTDR Method
OCWR vs OTDR Optical Continuous Wave Reflectometer Display Process Controller CW Stabilized Light Source Power Meter Coupler Termination Plug Accuracy (typ.) Typical Application Strengths Weaknesses ± 0.5dB - Total link ORL & isolated event reflectance measurements during fiber installation & commissioning - Accuracy - Fast & real time info - Simple & easy results (direct value) - No localization Optical Time Domain Reflectometer Display Process Controller Pulsed Light Source Photodetector Coupler Accuracy (typ.) Typical Application Strengths Weaknesses ± 2dB - Perfect tool for troubleshooting- Spatial characterization of reflective events & estimation of the partial & total ORL - Locate reflective events - Single-end measurement - Accuracy - Long acquisition time
Measuring Loss, ORL and Distance with an OTDR It s the single most important tester used in the installation, maintenance & troubleshooting of fibre plant Most versatile of Fibre Test Tools Identifies events & impairments (splices, bends, connectors, breaks) Provides physical distance to each event/ impairment Measures fibre attenuation loss of each event or impairment Provides reflectance / return loss values for each reflective event or impairment Manages the data collected and supports data reporting.
Bi-Directional OTDR Analysis Bi-directional OTDR Analysis True splice loss measurement (gives the average loss which eliminates for example the effect of fibre mismatch). Reveals events that are hidden by dead zones in one direction.
Measuring CD Reference test method defined by IEC 60793-1-42 and ITU-T G650.1 The modulated light is sent over the Fibre Under Test. The phase of the test signal is compared to the phase of the reference signal used to modulate the input signal. The measured value is the group delay corresponding to a wavelength interval. It is calculated using an approximation formula. The chromatic dispersion is then calculated by taking the derivative of the group delay with respect to wavelength. Ref Detection Phase Shift φ Modulated Broadband source λ 1260-1640nm Fibre under test Reference Filter λ ref (1550nm) λ n Wavelength Tunable Filter @ On @ Charge Phase detection & comparison λ ref λ n Measurement Signal Detection
Measuring PMD A polarized light is sent over the Fibre Under Test and the transmitted spectrum is analyzed with an Optical Spectrum Analyzer A reference power level is taken without the polarizer in the optical path The scan is then repeated with the polarizer inserted. There will be some fluctuations in the received optical power. The system calculates the ratio of the 2 power levels scans. It s possible to shift to the time domain the analysis of the fixed-analyzer response by taking the Fourier transform of the power fluctuations with wavelength. Broadband Source Polarizer FUT Polarizer OSA @ On @ Charge
Measuring Spectral Attenuation - Shows the spectrum of the fibre (e.g. see if the water peak is present) - Provides the total loss at each wavelength (equivalent to a light source and power meter) Broadband Source Isolator Spectrum Analyzer Reference Broadband Source Isolator FUT Spectrum Analyzer After Fibre Loss Spectrum Comparison Water Peak 700 900 1100 1300 1400 1500 1600
Fibre Characterisation Results