1、 TIA DOCUMENT FOTP124 Polarization-Mode Dispersion Measurement for Single-Mode Optical Fibers by Interferometry TIA-455-124-A (Revision of TIA/EIA-455-124) FEBRUARY 2004 TELECOMMUNICATIONS INDUSTRY ASSOCIATION The Telecommunications Industry Association represents the communications sector of NOTICE
2、 TIA Engineering Standards and Publications are designed to serve the public interest through eliminating misunderstandings between manufacturers and purchasers, facilitating interchangeability and improvement of products, and assisting the purchaser in selecting and obtaining with minimum delay the
3、 proper product for their particular need. The existence of such Publications shall not in any respect preclude any member or non-member of TIA from manufacturing or selling products not conforming to such Publications. Neither shall the existence of such Documents preclude their voluntary use by no
4、n-TIA members, either domestically or internationally. TIA DOCUMENTS TIA Documents contain information deemed to be of technical value to the industry, and are published at the request of the originating Committee without necessarily following the rigorous public review and resolution of comments wh
5、ich is a procedural part of the development of a American National Standard (ANS). Further details of the development process are available in the TIA Engineering Manual, located at http:/www.tiaonline.org/standards/sfg/engineering_manual.cfm TIA Documents shall be reviewed on a five year cycle by t
6、he formulating Committee and a decision made on whether to reaffirm, revise, withdraw, or proceed to develop an American National Standard on this subject. Suggestions for revision should be directed to: Standards FO-4.2, Subcommittee on Optical Fibers and Cables; and FO-4.6.1, Working Group on Sing
7、le-mode Fibers and Standards Harmonization.) This is part of the series of test procedures included within Recommended Standard EIA/TIA-455. 1. Introduction Intent This test method describes a procedure for measuring the polarization-mode dispersion (PMD) of single-mode optical fibers and cable asse
8、mblies. It provides a single measurement value that represents the PMD defined as the root-mean squared (RMS) differential group delay (DGD) over the measurement wavelength range of the selected source in the 1310nm or/and the 1550nm region or any other region of interest. The method can be applied
9、to any fiber length. Scope This procedure is restricted to wavelengths greater than or equal to that at which the fiber is effectively single-mode. The cutoff wavelength cfof an uncabled fiber may be determined by FOTP-80, while the cutoff wavelength ccof a cabled fiber may be determined by FOTP-170
10、. Background PMD causes an optical pulse to spread in the time domain; this dispersion could impair the performance of a single-mode fiber-optic telecommunications system. The effect can be related to differential group velocities and corresponding arrival times of different polarization components
11、of the signal. For a narrow band source, the effect can be TIA-455-124-A 7related to the DGD between pairs of orthogonal principal states of polarization (PSPs), in absence of polarization dependent loss and non-linear effects. In long fiber spans, PMD is a random effect since it depends on the deta
12、ils of the birefringence along the entire fiber length. It is also sensitive to time-dependent temperature and mechanical perturbations on the fiber. For this reason, a useful way to characterize PMD in long fibers is in terms of the expected value of the mean DGD or the RMS DGD 1/2when considering
13、the DGD distribution as a function of wavelength. The two definitions (mean or RMS DGD) mathematically hold and are accepted. In principle, the expected value does not undergo large changes for a given fiber from day to day or from source to source, unlike the parameters or . In addition, or 1/2is a
14、 useful predictor of lightwave system performance. When the DGD distribution as a function of wavelength can be approximated by a Maxwell distribution, then can be used as an easier predictor of system performance. In this case, can be easily correlated to 1/2; a maximum DGD can also be found for a
15、defined value of the probability density function (frequency of occurrence of the DGD) from the Maxwell distribution. The term “PMD” is used in the general sense of the phenomenon of the two PSPs having different group velocities, and in the specific sense of the expected value or 1/2. The DGD or pu
16、lse broadening can be averaged over wavelength, yielding , or time, yielding t, or temperature, yielding . For most purposes, it is not necessary to distinguish between these various options for obtaining . The coupling length h is the length of fiber or cable at which appreciable coupling (i.e. ene
17、rgy transfer) between the two PSPs begins to occur. Mode coupling is the physical phenomenon by which energy is exchanged between PSPs. If the fiber length L satisfies the condition L /L or = 1/2/L ( 1 ) The fiber length satisfying the condition L h regime in which case the mode coupling is random.
18、In this case or 1/2scales with the square root of fiber length, and “long-length” PMD coefficient = /L1/2or = 1/2/L1/2( 2 ) TIA-455-124-A 8Fiber lengths in the transition region L h (mixed mode coupling) cannot be described by either (1) or (2) and consequently PMD needs to be stated only by or 1/2.
19、 Typical units are ps for , and 1/2, km for L, ps/km for short-length PMD coefficient, and ps/km1/2for long-length PMD coefficient. The principle of this interferometric method (INTY) is generically based on the measurement of the time broadening of the source field cross-correlation interferogram a
20、nd then, PMD delay defined as the RMS DGD 1/2is deduced from this time broadening. This will provide direct measurement of PMD. In the case of an autocorrelation-type instrument, the resulting interferogram has a central coherence peak corresponding to the auto-correlation of the optical source. A c
21、ross-correlation-type interferometer has no central peak. A more general instrument has both autocorrelation and cross-correlation interferograms. This INTY method is based on two different kinds of analysis: The traditional analysis (TINTY) using a set of specific operating conditions for its succe
22、ssful applications and a basic set-up 1-3; and A general analysis (GINTY) using no limiting operating conditions but using a modified set-up compared to TINTY 4. 2. Normative References Test or inspection requirements may include, but are not limited to, the following references: TIA/EIA-455-A, Stan
23、dard test procedures for optic fibers, cables, transducers, sensors, connecting and terminating devices, and other fiber optic components TIA/EIA-455-57 (FOTP-57), Optical fiber end preparation and examination TIA/EIA-455-80 (FOTP-80), Cutoff wavelength of uncabled single-mode fiber by transmitted P
24、ower TIA/EIA-455-113 (FOTP-113), Polarization-Mode Dispersion Measurement for Single-Mode Fibers by Fixed Analyzer TIA-455-124-A 9TIA/EIA-455-122A (FOTP-122A), Polarization-Mode Dispersion Measurement for Single-Mode Optical Fibers by Stokes Parameter Evaluation TIA/EIA-455-170 (FOTP-170), Cable Cut
25、off Wavelength of Single-mode Fiber by Transmitted Power Users of this FOTP are encouraged to specify the most recent edition of the FOTPs referenced above. Caution: Do not make casually the decision to require the most recent edition of a referenced FOTP. There have been instances when document rev
26、isions have completely changed intent, application, use etc. of a document such that the requirement to use an edition more recent than the one originally reviewed may be totally inappropriate. 3. Apparatus Figure 1 shows the schematic diagram of a generic INTY measurement system. Broadband SourcePo
27、larizer InterferometerSo()Ss()sos()FUTJ()SOPS()AnalyzersaModulus of SumFourierSpectrumS()So()Optical Frequency ()EnvelopeE()Delay ()auto-correlationcross-correlationFringesP()(Ss)(s121)(Soa +=SpectrumS()EnvelopeE()SpectrumS()EnvelopeE()SpectrumS()SpectrumS()EnvelopeE()EnvelopeE()EnvelopeE()Figure 1
28、Generic Interferometric Method (INTY) Measurement System The parameters used in Figure 1 and throughout the document are: TIA-455-124-A 10 optical frequency ( = c); difference of round-trip delay between the two arms of the interferometer; )(Ssoptical spectrum, at fiber under test (FUT) input spectr
29、al density of )Es(r, the source spectrum; )(Sooptical spectrum, at FUT output (analyzer input); )(S optical spectrum, at analyzer output (interferometer input); )(Esroptical field complex amplitude at FUT input; )(Eoroptical field complex amplitude at FUT output (analyzer input); )(E roptical field
30、complex amplitude at analyzer output (interferometer input); ),(Edroptical field complex amplitude at interferometer output; 0sinput state of polarization (SOP) (at FUT input; a unit Stokes vector); )(soutput SOP (at FUT output); asanalyzer transmission axis; x() )(ssa, the Stokes parameter giving t
31、he projection of )(son the analyzer transmission axis. It is this parameter, x(), that contains the PMD information; )(J Jones matrix of the FUT such that )(E )(J )(Eso=rr; )(P optical power at the interferometer output, as a function of delay ; )(P-dependent part of )(P (“a.c.“ part), or)(P minus t
32、he -independent part (“d.c.“ part); or in other words, the zero-mean oscillating fringes; TIA-455-124-A 11E() envelope of the interferogram: amplitude or RMS-value of the oscillating fringes; AM-demodulation. The envelope may also be labelled “fringes visibility“; Ex() cross-correlation envelope. As
33、 illustrated in Figure 1 above, the analyzer transmission, )(S/)(S0 , consists of two parts: a constant part; and an SOP-dependent part. Therefore, the spectrum at the output of the analyzer, S(), also consists of two parts, and as a consequence, the interferogram, )(P , also consists of these two p
34、arts since it is the FT of the spectrum S(). In summary: The constant part gives: autocorrelation FT of spectrum )(S0at input of analyzer. The SOP-dependent part: cross-correlation FT of)(S)(ss0a. where )(sis the unit Stokes vector that represents the SOP at the output of the FUT (output-SOP) and as
35、is the unit Stokes vector that represents the analyzer axis. Figure 2a and Figure 2b show a schematic diagram of typical TINTY-based measurement systems. These systems may also be used with GINTY analysis. TIA-455-124-A 12ControllerFiberConnector or Splice Polarizer Mirror Coupler Moving Mirror Frin
36、ge envelope detection Optical source (A) Michelson Interferometer Polarizer MirrorDetection Optical source Fiber Connector or Splice Controller Moving arm /2 Beam Splitter Beam Splitter /2 Mirror Moveable Cube Corner (B) Mach-Zehnder Interferometer Figure 2 Basic Measurement Systems for the Interfer
37、ometric Method Typically Using TINTY Analysis Figure 3 shows a schematic diagram of a typically modified GINTY-based measurement system using a polarization diversity detection system with a polarization beam splitter (PBS) in order to simultaneously and separately obtain both the autocorrelation an
38、d TIA-455-124-A 13cross-correlation envelopes. The input/output (I/O)-SOP scramblers may also be used with TINTY-based systems. Broadband SourcePolarizerFUTScramblersFringes:Px()Py()Polarization Diversity DetectorPolarization Beam SplitterInterferometerFigure 3 Modified Measurement System for the In
39、terferometric Method Typically Using GINTY Analysis 3.1 Light Source Use a broadband light source that emits radiation at the intended measurement wavelengths, such as a light emitting diode (LED), a superfluorescent source or an amplified spontaneous emission (ASE) source. The central wavelength Cs
40、hall be within the 1310nm or 1550nm region or any other region of interest. For a TINTY-based measurement system, the source spectral shape shall be approximately Gaussian, without ripples that could influence the autocorrelation function of the emerging light. There are no such requirements for a G
41、INTY-based measurement system. The optical source spectral linewidth must be known to calculate the source coherence time tC. Determine the coherence time tCfrom the equation: ct2cc=( 3 ) where C is the central wavelength of the source, is the source linewidth, FWHM, and c is the velocity of light i
42、n free space. TIA-455-124-A 14The source linewidth should be chosen to best meet the desired PMD range. The light source should have a coherence time smaller than the PMD of the FUT. For a fixed central wavelength, the wider the optical source linewidth, the smaller the PMD which can be measured. Wh
43、en the PMD is larger than the coherence time of the source (PMD tc), the output will be depolarized and this is suited for measuring PMD. When the PMD is smaller than the coherence time (PMD 20dB. An optional half-wave plate can be used following the source to change the relative coupling to the PSP
44、s or principal axes of the FUT. This allows for optimization of the contrast of the interference function. 3.3 Polarizers/Scramblers Multiple I/O SOPs may be used with either TINTY- or GINTY-based measurement systems in order to obtain a more complete interferometric envelope, the mean envelope for
45、TINTY or the mean-squared envelope for GINTY, than would be possible for a single I/O-SOP setting. While a measurement obtained from a single I/O-SOP setting is an option, doing multiple I/O-SOP settings will improve the precision of the result. Physically, SOP scrambling consists of inserting polar
46、ization controllers as shown in Figure 3, one at the input and one at the output of the FUT, such that different input SOPs and analyzer axes are set for an interferometer scan. Conceptually, the scambling process can be viewed as follows: The input-polarizer followed by the scrambler acts as a sing
47、le unit, an equivalent polarizer which axis is set at any point on the Poincar sphere in order to define the input SOP; and The scrambler followed by the analyzer acts as an equivalent analyzer which axis is set at any point on the Poincar sphere in order to define the output SOP. One set of input-S
48、OP/analyzer-axis combination will be labelled as one I/O SOP. The aim is to get the TINTY envelopes or the GINTY squared envelopes averaged over uniformly distributed I/O SOPs. In practice, there exist a number of possible ways to achieve this goal as shown below (suggested here for the GINTY case).
49、 TIA-455-124-A 153.3.1 The 9-States Mueller Set As an example, the sum of nine squared envelopes observed with nine specific I/O SOPs is rigorously equal to the uniformly scrambled mean-squared envelope. These nine I/O SOPs are: three analyzer-axes forming a right-angled trihedron, for each three input SOPs also forming a right-angled trihedron. 3.3.2 Uniform Grid If the necessary hardware is available such as wa
