BACKGROUND OF THE INVENTION
The present invention relates to testing of optical components
in particular for communication systems.
Measurement of polarization dependent parameters like polarization
dependent loss (PDL) and polarization dependent group delay PDGD (covering Differential
Group Delay DGD and Polarization Mode Dispersion PMD) is of increased importance
for advanced communication systems and generally described in 'Fiber Optic Test
and Measurement' by Dennis Derickson, ISBN 0-13-534330-5, 1998, pages 354ff. Especially
long-haul high-speed systems require that polarization properties of its components
fulfill certain requirements. In general, component manufacturers address this by
100% testing of components for critical parameters. PDL nowadays in many cases is
already measured 100%, PDGD may also develop to be a 100% test in manufacturing.
Today's solutions for measuring polarization dependent
loss parameters are the scrambling method (applying a random variation of polarization
states and comparing maximum with minimum determined loss) or the Mueller method,
whereby 4 defined polarization states are measured for each wavelength point and
analyzed together. The latter requires multiple measurement sweeps at predefined
polarization states. These methods are either slow, if testing at multiple wavelengths
is required (PDL measurement using the scrambling method), or require multiple measurement
sweeps at predefined polarization states (Mueller method). Multiple sweeps are disadvantageous
because measurement time is increased and require very high stability of the measurement
setup because no change of polarization properties of the whole setup (between laser
and DUT) is allowed between the sweeps.
US-B-6,229,606 discloses measuring of PMD of a dispersion
compensation grating. An optical source generates sequential optical beams with
different wavelengths. A polarization synthesizer receives the beams and produces
states of polarization for each beam. The polarized beams travel sequentially through
a DUT, which produces the PMD being measured, and then enter an analyzer. The analyzer
measures the intensity of the received beams with the different polarizations and
generates the Stokes parameters.
SUMMARY OF THE INVENTION
It is an object of the invention to provide an improved
measurement of polarization dependent parameters. The object is solved by a system
according to claim 1 and a method according to claim 5. Preferred embodiments are
defined by the dependent claims.
For measuring polarization dependent parameters of an optical
device under test (DUT), an optical source (preferably a tunable laser) provides
an optical signal through an optical polarization translator to the DUT. The polarization
translator translates the polarization of the optical signal from its input to its
output in a deterministic way dependent on the wavelength of the optical signal.
The polarization translator provides the translation of
the polarization dependent on the wavelength preferably using birefringent properties.
Accordingly, the optical source may provide a variation of the wavelength over the
time, and the polarization translator provides a 'translation' of the polarization
dependent on the wavelength. The parameters wavelength and frequency shall be regarded
here as equivalents (related by the general equation f = c/&lgr;).
When varying the wavelength of the optical source, the
polarization translator changes the polarization of the signal launched into the
DUT. Tuning the wavelength of the optical source in a way that measurement points
with different polarization states are covered thus allows determining polarization
dependent parameters of the DUT in that particular wavelength range.
Typical polarization dependent parameters that can be analyzed
by the invention are polarization dependent loss (PDL) or polarization dependent
group delay PDGD (also referred to as Differential Group Delay (DGD) or Polarization
Mode Dispersion (PMD)).
The uncertainty of the polarization state of the output
signal may be reduced by tapping off some fraction of the signal in an appropriate
way and analyzing its polarization state at each wavelength with a polarimeter or
a reduced polarization analysis device like an Analyzer.
The polarization translator may be purely passive. The
optical signal preferably does not hit a Principle State of the Polarization (PSP)
of the polarization translator, so that the output signal will follow a trajectory
(e.g. a circle) on the Poincare Sphere in a deterministic way.
The same principle of scanning the polarization can be
applied to various PMD measurement techniques: For example the Jones Matrix Eigenanalysis
(JME) or a novel method which is outlined in the European Patent Application No.
125089.3 (EP 1113250). In case of PMD measurements in general only two polarization
states are combined to a measurement value.
In case that several measurement points (defined by the
wavelength and the polarization state of the optical signal applied to the DUT)
are to be analyzed together for determining a value of a polarization dependent
parameter, the wavelength range for those measurement points is preferably selected
that a value of the polarization dependent parameter of the DUT can be considered
as substantially constant in that wavelength range.
Preferred algorithms for analyzing together several such
measurement points are interpolation of neighboring measurement points, combining
4 measurement points using the Mueller Matrix analysis, or combining 2 measurement
points using e.g. the Jones Matrix analysis.
The invention has various advantages compared to today's
standard methods (polarization scrambling and Mueller Matrix analysis). The polarization
transformation device may be purely passive, the number of measurement points can
be chosen to be much smaller compared to the scrambling method, and, most important,
the complete measurement can be performed within one sweep (instead of four for
the Mueller Matrix Analysis). Thus, the invention allows fast measurements and is
also less sensitive against e.g. environmental or mechanical disturbances.
The invention can be partly embodied or supported by one
or more suitable software programs, which can be stored on or otherwise provided
by any kind of data carrier, and which might be executed in or by any suitable data
BRIEF DESCRIPTION OF THE DRAWINGS
Other objects and many of the attendant advantages of the
present invention will be readily appreciated and become better understood by reference
to the following detailed description when considering in connection with the accompanied
drawings. Features that are substantially or functionally equal or similar will
be referred to with the same reference sign(s).
DETAILED DESCRIPTION OF THE INVENTION
- Fig. 1 shows a measurement setup according to the invention for measuring polarization
- Fig. 2 shows a representation of polarization transformation on Poincare Sphere.
- Figs. 3 and 4 show embodiments of the polarization translator.
In Fig. 1, a tunable laser 10 as an optical source provides
an optical signal through an optical polarization translator 20 to an optical device
under test (DUT) 30. A power meter 40 receives and detects the optical signal after
passing the DUT 30. The polarization translator 20 translates the polarization of
the optical signal from its input to its output in a deterministic way dependent
on the wavelength of the optical signal.
A polarization analyzer 50 might be optionally coupled
to the output of the polarization translator 20 in order to determine the polarization
state of the output signal of the polarization translator 20.
A controller 60 is coupled to the power meter 40 for analyzing
polarization dependent parameters. Preferably, the controller 60 is further coupled
to the tunable laser 10 for controlling the application and variation of the provided
optical signal and to the polarization analyzer 50 for receiving information about
the actual polarization state of the output signal from the polarization translator
When varying the wavelength of the tunable laser 10, the
polarization translator 20 changes the polarization of the optical signal launched
into the DUT 30. In operation, the wavelength of the tunable laser 10 is tuned in
a way that optical signals with different polarization states are provided to the
DUT 30. For each measuring point (defined by the wavelength and the polarization
state of optical signal applied to the DUT 30), the controller 60 receives a value
of power intensity determined by the power meter 40. Analyzing the power intensity
values for a plurality of different measuring points thus allows determining polarization
dependent parameters of the DUT 30 such as polarization dependent loss (PDL).
In a preferred embodiment, several measurement points will
be analyzed together for determining a value of a polarization dependent parameter.
The wavelength range for those measurement points is selected that a value of the
determined polarization dependent parameter of the DUT can be considered as substantially
constant in that wavelength range.
Preferably, a set of four measurement points with different
polarization states is analyzed together resulting e.g. in one PDL value for the
DUT. The wavelength range of the set of measurement points is thereby preferably
selected to be smaller than the wavelength resolution for that measurement. In an
example wherein PDL measurements are desired with a spectral resolution of 1pm,
a set of measurement points with 4 different polarizations should have a spectral
distance of 0.25pm. Within such wavelength range (even going further up to say 10
pm) it can be seen as sufficiently ensured with typical DUTs of present optical
networks that constant PDL properties are maintained in that range.
Fig. 2 shows a representation of polarization transformation
on Poincare Sphere. The polarization state Pin of the optical signal
applied to the polarization translator 20 will be transformed to polarization states
out (with i = 1, 2, 3, ...) at the output of the polarization translator
20. In case of a waveplate (which shows purely linear birefringence) as polarization
translator 20 with an orientation O, the polarization states Pi
out are all located on a circle on the Poincare Sphere dependent on the
wavelength &lgr; of the optical signal.
The polarization translator 20 can be implemented in several
ways depending on requirements of the particular measurement to be conducted. If
a PDL measurement is to be performed with a Mueller Matrix type of implementation
(see for example on pages 356ff in 'Fiber Optic Test and Measurement' by Dennis
Derickson, ISBN 0-13-534330-5, 1998), a set of at least four measurements has to
be made with polarization states fulfilling certain requirements: they have to be
significantly different, must not be located on a great circle of the Poincare Sphere,
and preferably should not be located on any circle on the Poincare Sphere. A very
high order waveplate that is excited by the linearly polarized signal of the tunable
laser 10 can be used. The angle between the polarization of the optical signal and
the optical axis of the waveplate is defined to &phgr;1. However, in
this configuration the states of polarization Pi
out are located on a circle that under certain circumstances won't allow
conducting the Mueller Matrix type of calculation. This problem can be avoided if
at least two waveplates are concatenated with their principle state of polarization
PSP not aligned. This configuration would provide second order PMD, which means
that the trajectory of the polarization translation function on the Poincare sphere
won't be a circle anymore.
In contrast to the Mueller Matrix based PDL measurement
the PMD measurement techniques mentioned before only require measurement on 2 states
of polarization. There are no requirements for these States of Polarization as long
as they are sufficiently different.
In a preferred implementation, a high order waveplate as
the polarization translator 20 creates a phase difference between the propagating
modes by its birefringence. An angle &agr;1 represents the phase difference
of the two optical signals propagating in the two Eigenmodes of the waveplate device.
When entering the device the phase difference is &agr;1=0. When exiting
the angle is given by:
with &Dgr;n, &lgr;, L representing the difference of the refractive indices
of the two propagating polarization modes, the optical wavelength and the length
of the device, respectively. If the length L of the birefringent device 20 is kept
fixed and dispersive effect are neglected (which means &Dgr;n is constant over
wavelength), the wavelength increment to increase &agr;1 by a given
amount, say &Dgr;&agr;1 is given by:
For example, if PMD measurement values are required with
a spectral resolution of 1 nm (as it may be sufficient for fused couplers as DUT
30), measurement values should be taken with an interval of 0.5nm. From Equation
2, a condition (at &lgr;=1.5µm) for the polarization transformer 20 can be
This requirement could be fulfilled with for example a
LiNbO3 waveguide or a birefringent fiber as the polarization translator
In a LiNbO3 based polarization transformer 20,
a Ti diffused waveguide can be implemented perpendicular to the c-axis (optical
axis) of the LiNbO3 crystal (typically selected: x- or y-cut). In this
configuration, the waveguide 20 has a high birefringence of &Dgr;n≈ 0.079
with a beat length LB
of the two propagating modes of:
around &lgr;=1.55 µm. Therefore the requirement mentioned in Equation 3 can
be fulfilled with a LiNbO3 waveguide 20 of a length of about 3 cm.
A polarization transformer 20 based on Polarization Maintaining
Fiber (PMF) has a typical birefringence of about 10-3. As this is much
lower than in LiNbO3, a much longer length is required: 2.25m. By increasing
the length even further, a higher spectral resolution can be achieved. However,
typical DWDM component test applications would require a spectral resolution of
the PDL and PMD measurement around 1...3pm. Therefore a PMF fiber length of more
than 1000m would be required, which might not be applicable for some applications
e.g. for price, volume and possibly stability reasons.
In a further preferred implementation, an 'artificial birefringent
device' 200 is used as polarization transformer 20 creating enough delay between
the two propagating polarization modes for very high spectral resolution. The incoming
(linearly polarized) light is split up in the artificial birefringent device 200
and guided along two different paths having different path lengths with a length
difference &Dgr;L. The artificial birefringent device 200 further provides the
light returning from the two paths with orthogonal states of polarization. This
can be done e.g. by splitting up the incoming light polarization dependent or by
changing the state of polarization at least in one of the paths. After recombining
the light returning from both paths, the state of polarization of the combined signal
depends on the wavelength (or more accurately the frequency) of the optical signal
in a deterministic, periodic way. By adjusting the length difference &Dgr;L the
periodicity can be broadly varied.
Fig. 3 shows a first embodiment of the artificial birefringent
device 200. The incoming (linearly polarized) light is split up by a beam splitter
or fiber coupler 210 and guided along the two different paths. One (typically short)
path returns the signal with its original polarization. A second path, which has
a geometric length difference &Dgr;L, returns the signal in its orthogonal state
of polarization, e.g. by providing a Faraday Mirror 220. After recombining the light
of the first and the second path, the state of polarization of the combined signal
depends on the wavelength of the optical signal.
Fig. 4 shows a further embodiment of an artificial birefringent
device 200, preferably made of PMF components. The incoming (linearly polarized)
light is split up by a polarization dependent beam splitter 250 into light beams
having orthogonal states of polarization, guided along the two different paths with
the length difference of &Dgr;L, and recombined with the still orthogonal states
of polarization. The state of polarization of the combined signal again depends
on the wavelength of the optical signal. In order to provide both paths with substantially
the same optical powers, a polarizer 260 might be inserted before the polarization
dependent beam splitter 250 in order to polarize the incoming light to 45°
with respect to polarization states provided by the polarization dependent beam
The artificial birefringent device 200 allows creating
an almost arbitrarily selectable delay difference between two signal fractions.
The delay difference is defined by the length difference of the fibers &Dgr;L.
For this setup Equation 2 changes to:
where n represents the refractive index of the fiber. As an example, getting a PDL
or PMD measurement resolution of 1 pm can be achieved by a length difference &Dgr;L=1.5m.