1、 TIA/EIASTANDARDOFSTP-19Optical Signal-to-Noise RatioMeasurement Procedures for DenseWavelength-Division MultiplexedSystemsTIA/EIA-526-19JUNE 2000TELECOMMUNICATIONS INDUSTRY ASSOCIATIONRepresenting the telecommunications industry inassociation with the Electronic Industries Alliance ANSI/TIA/EIA-526
2、-19-2000Approved: May 19, 2000TIA/EIA-526-19NOTICETIA/EIA Engineering Standards and Publications are designed to serve the public interestthrough eliminating misunderstandings between manufacturers and purchasers, facilitatinginterchangeability and improvement of products, and assisting the purchase
3、r in selecting andobtaining with minimum delay the proper product for his particular need. Existence of suchStandards and Publications shall not in any respect preclude any member or nonmember ofTIA/EIA from manufacturing or selling products not conforming to such Standards andPublications, nor shal
4、l the existence of such Standards and Publications preclude their voluntaryuse by those other than TIA/EIA members, whether the standard is to be used either domesticallyor internationally.Standards and Publications are adopted by TIA/EIA in accordance with the American NationalStandards Institute (
5、ANSI) patent policy. By such action, TIA/EIA does not assume any liabilityto any patent owner, nor does it assume any obligation whatever to parties adopting the Standardor Publication.This Standard does not purport to address all safety problems associated with its use or allapplicable regulatory r
6、equirements. It is the responsibility of the user of this Standard toestablish appropriate safety and health practices and to determine the applicability of regulatorylimitations before its use.(From Standards Proposal No. 4290, formulated under the cognizance of the TIA FO-2.1Subcommittee on Single
7、-Mode Systems.)Published byTELECOMMUNICATIONS INDUSTRY ASSOCIATION 2000Standards and Technology Department2500 Wilson BoulevardArlington, VA 22201PRICE: Please refer to current Catalog ofEIA ELECTRONIC INDUSTRIES ALLIANCE STANDARDS and ENGINEERING PUBLICATIONS or call Global Engineering Documents, U
8、SA and Canada(1-800-854-7179) International (303-397-7956)All rights reservedPrinted in U.S.A.PLEASE!DONT VIOLATETHELAW!This document is copyrighted by the TIA and may not be reproduced withoutpermission.Organizations may obtain permission to reproduce a limited number of copiesthrough entering into
9、 a license agreement. For information, contact:Global Engineering Documents15 Inverness Way EastEnglewood, CO 80112-5704 or callU.S.A. and Canada 1-800-854-7179, International (303) 397-7956TIA/EIA-526-19iOFSTP-19Optical Signal-to-Noise Ratio Measurement Procedures forDense Wavelength-Division Multi
10、plexed SystemsContentsForewordiii1 Introduction. 12 Apparatus 43 Sampling and specimens 94 Procedure 95 Calculations . 96 Measurement uncertainty.107 Documentation10Annex A Error in measuring signal level due to signal spectral width 11Annex B Comparison between OFSTP-19 and IEC and ITU requirements
11、14Annex C References15TIA/EIA-526-19iiThis page left blank.TIA/EIA-526-19iiiOFSTP-19Optical Signal-to-Noise Ratio Measurement Procedures for DenseWavelength-Division Multiplexed SystemsForeword(This Foreword is informative only and is not part of this Standard.)From TIA Project No. 4290, this OFSTP
12、is formulated under the cognizance of TIAFO-2.1/6.6, Joint Subcommittee on Single-Mode Fiber Systems.It is part of the series of test procedures included within Recommended StandardTIA/EIA-526.There are four informative annexes.Key words: OSNR, optical signal-to-noise ratio.TIA/EIA-526-19ivThis page
13、 left blank.TIA/EIA-526-1911 Introduction1.1 IntentAt the optical interfaces within wavelength-division multiplexed (WDM) networks, itis desirable to measure parameters that provide information about the integrity ofthe physical plant. Such parameters are necessary to monitor network performanceas a
14、n integral part of network management. They are also necessary to assureproper system operation for installation and maintenance of the network.Ideally, such parameters would directly correspond to the bit error ratio (BER) ofeach channel of a multichannel carrier at the particular optical interface
15、. Relatedparameters such as Q-factor or those calculated from optical eye patterns wouldprovide similar information, that is, they would correlate to the channel BER.However, it is difficult to obtain access to these parameters at a multichannelinterface point. It is necessary to demultiplex the pot
16、entially large number ofchannels and make BER, Q-factor, or eye-diagram measurements on a per-channelbasis.In contrast, useful information about the optical properties of the multichannel carrieris readily obtained by measuring the optical spectrum. Wavelength-resolved signaland noise levels provide
17、 information on signal level, signal wavelength, andamplified spontaneous emission (ASE) for each channel. Spectral information,however, does not show signal degradation due to waveshape impairmentsresulting from polarization-mode dispersion (PMD), and chromatic dispersion.Also, intersymbol interfac
18、e and time jitter are not revealed from an OSNRmeasurement. In spite of these limitations, OSNR is listed as an interfaceparameter in ITU-T draft Recommendation G.692; “Optical interfaces formultichannel systems with optical amplifiers.” It is also proposed that OSNR belisted in ITU-T draft recommen
19、dation G.959.1, “Optical networking physical layerparameters.”This OFSTP provides a parameter definition and a test method for obtaining opticalsignal-to-noise ratio (OSNR) using apparatus that measures the optical spectrum ata multichannel interface. Three implementations for an optical spectrum an
20、alyzer(OSA) are discussed: a diffraction-grating-based OSA, a Michelson interferometer-based OSA, and a Fabry-Perot-based OSA. Performance characteristics of theOSA that affect OSNR measurement accuracy are provided.TIA/EIA-526-1921.2 BackgroundA typical optical spectrum at a multichannel interface
21、is shown in Figure 1.Important characteristics are:(a) The channels are placed nominally on the grid defined by ITU G.692.(b) Individual channels may be non-existant because it is a network designed withoptical add/drop demultiplexers or because particular channels are out ofservice.(c) Both channel
22、 power and noise power are a function of wavelength.For calculating OSNR, the most appropriate noise power value is that at the channelwavelegth. However, with a direct spectral measurement, the noise power at thechannel wavelength is included in signal power and is difficult to extract. Anestimate
23、of the channel noise power can be made by interpolating the noise powervalue between channels.WavelengthOptical powermissing channelschannels on theITU gridnoiseFigure 1. A typical optical spectrum at an optical interface in a multichanneltransmission system.1.3 OSNR DefinitionReferring to Figure 2,
24、 from the optical spectrum, the OSNR is defined as follows:rmiiBBLogNPLogOSNR 1010 += dB (1)TIA/EIA-526-193where:Pi is the optical signal power in watts at the i-th channel.Ni is the interpolated value of noise power in watts measured in noiseequivalent bandwidth, Bm, at the i-th channel:Dl is the i
25、nterpolation offset equal to or less than one-half the ITU gridspacing,Br is the reference optical bandwidth. (The units for Bm and Br may be infrequency or wavelength but must be consistent.) Typically, the referenceoptical bandwidth is 0.1 nm.Pi + NiNi N(li+Dl)N(li-Dl)Figure 2. The OSNR for each c
26、hannel is derived from direct measurements of theoptical spectrum.N N Ni i i= + +( ) ( )l l l l 2 (2)TIA/EIA-526-1942 ApparatusThe required apparatus is an optical spectrum analyzer (OSA) with the performancenecessary to measure the signal and noise powers required for equation (1). Threecommon ways
27、 to implement an OSA are with a diffraction grating, a Michelsoninterferometer, and a Fabry-Perot etalon.2.1 Diffraction grating-based OSAA simplified diagram of a diffraction grating-based OSA is shown in Figure 3. Theexpanded input light is incident on a rotatable diffraction grating. The diffract
28、ed lightcomes off at an angle proportional to wavelength and passes through an aperture toa photodetector. The size of the input and output apertures and the size of thebeam on the diffraction grating determine the spectral width of the resulting filter andtherefore the resolution of the OSA. A/D co
29、nversion and digital processing providethe familiar OSA display.PhotodiodeDiffraction GratingDisplayLight inputSLITA/DconverterDigitalprocessingFigure 3. A diffraction grating-based OSA2.2 Michelson interferometer-based OSAAnother type of OSA is based on the Michelson interferometer as shown in Figu
30、re4. The input signal is split into two paths. One path is fixed in length and one isvariable. The Michelson interferometer creates an interference pattern between thesignal and a delayed version of itself at the photodetector. The resulting waveform,called an interferogram, is the autocorrelation o
31、f the input signal. A Fouriertransform performed on the autocorrelation provides the optical spectrum. Theresolution of this type of OSA is set by the differential path delay of theinterferometer.TIA/EIA-526-195Light inputPhotodiodeBeam splitterFixed mirror Movable mirrorDisplayA/DconverterDigitalpr
32、ocessingFigure 4. A Michelson interferometer-based OSA2.3 Fabry-Perot-based OSAA third type of OSA is based on a Fabry-Perot etalon as shown in Figure 5. Thecollimated beam passes through a Fabry-Perot etalon, the free spectral range(FSR) of which is greater than the channel plan and the Finesse is
33、chosen to givethe required RBW. Piezo-electric actuators control the Fabry-Perot mirror spacingand provide spectral tuning.Digital signal processing provides any combination of spectral display or tabulardata.Figure 5. A Fabry-Perot-based OSADSP A/DConverterFabry-PerotEtalon PhotodiodeLight inputDis
34、playSpectrumor tableTIA/EIA-526-1962.4 OSA Performance RequirementsRefer to reference 1, Calibration of optical spectrum analyzers.2.4.1 Wavelength RangeThe wavelength range shall be sufficient to cover the channel plan plus one-half agrid spacing on each end of the band to measure the noise of the
35、lowest andhighest channel.2.4.2 SensitivitySensitivity of an OSA is defined as the lowest level at which spectral power can bemeasured with a specified accuracy. The OSA sensitivity must be sufficient tomeasure the lowest expected noise level. In terms of OSNR:Required Sensitivity (dBm) = Minimum ch
36、annel level (dBm) OSNR (dB) (3)For example, the sensitivity required for a minimum channel level of 10 dBm inorder to measure a 35-dB OSNR is:10 35 = 45 dBm.2.4.3 Resolution Bandwidth (RBW)The resolution bandwidth must be sufficiently wide to accurately measure the powerlevel of each modulated chann
37、el. The proper RBW setting depends on the bit rate.For example, the signal power of a laser modulated at an OC-192 (STM-16) ratewith zero chirp will measure 0.8-dB low with a 0.1-nm RBW. This results from themodulation envelope having a portion of its spectral power outside of the 0.1-nmRBW. If the
38、RBW is decreased to 0.05 nm, the signal power will measure 2.5 dBlow. This effect is made worse by the presence of laser chirp and lessened byadditional bandwidth limiting in the transmitter lasers modulation circuitry. Thissubject is treated in more detail in Annex A.TIA/EIA-526-1972.4.4 Resolution
39、 Bandwidth AccuracyThe accuracy of the noise measurement is directly impacted by the accuracy of theOSAs RBW. For best accuracy, the OSAs noise equivalent bandwidth, Bm, mustbe calibrated. RBW, in general, differs from Bm due to the non-rectangular shape ofthe optical spectrum analyzers filter chara
40、cteristic.2.4.5 Dynamic RangeThe dynamic range of an OSA is a measure of the OSAs ability to makemeasurements of low-level signals and noise that are close in wavelength to largesignals. It is important to note that narrowing the RBW does not necessarilycorrelate to better dynamic range. RBW is a me
41、asure of the 3-dB bandwidth ornoise equivalent bandwidth of its filter characteristic. Dynamic range, on the otherhand, is a measure of the steepness of the filter characteristic and the OSA noisefloor. Dynamic rage is defined as the ratio, in dB, of the filter transmissioncharacteristic at the cent
42、er wavelength, li, and at one-half a grid spacing away, lI D.Figure 6 shows two channels of a multichannel spectrum, the OSA filtercharacteristic, the OSA sensitivity limit, and the transmission system noise that is tobe measured. At the noise measurement wavelength, the dynamic range must besignifi
43、cantly higher than the OSNR for accurate measurements. The uncertaintycontribution can be predicted from the following equation:Uncertainty in OSNR = 10 log(1+10-D/10) dB (4)where D is the value in dB that the OSA dynamic range exceeds the actual OSNR.For example, for an OSNR of 30 dB, a dynamic ran
44、ge of 40 dB (at the ITU gridspacing) will cause an error of 0.42 dB.TIA/EIA-526-198OSA filtershapetransmission systemnoiseOSA dynamic rangeat 1/2 a grid spacingOSA sensitivity limitITU grid spacingOSA filtershapesignal signalFigure 6. Insufficient dynamic range is another source of measurement uncer
45、tainty.In general, either the OSA sensitivity limit or dynamic range will limit the value ofOSNR that can be measured. Typically, a Michelson interferometer-based OSA willbe limited by the sensitivity limit and a diffraction grating-based OSA by thedynamic range.2.4.6 Scale fidelityScale fidelity, a
46、lso called display linearity, is the relative error in amplitude thatoccurs over a range of input power levels. Scale fidelity directly contributes toOSNR measurement uncertainty.2.4.7 Polarization dependenceTypically, the signal, Pi, will be highly polarized while the noise, Ni, is unpolarized.OSA
47、polarization dependence will directly contribute to uncertainty in signalmeasurement.2.4.8 Wavelength data pointsThe minimum number of data points collected by the OSA shall be at least twice thewavelength span divided by the noise equivalent bandwidth.TIA/EIA-526-1993 Sampling and specimensThe devi
48、ce under test (DUT) is a multichannel fiber-optic transmission system ornetwork. The measurement apparatus is connected to the network at any point bydirectly connecting to the optical fiber or via a broadband monitoring port.4 Procedure4.1 Connect the OSA to the transmission fiber or a monitor port
49、.4.2 Choose RBW values sufficiently wide to accurately measure the signal powerand with sufficient dynamic range to measure the noise at Dl from the peakchannel wavelength where Dl is the ITU grid spacing. (See Annex A, Table 2and clause 2.4.5).4.3 Set the wavelength range to accommodate all channels plus at least a gridspacing below the lowest channel and above the highest channel.4.4 Measure the power level
