C measuring polarization mode dispersion: theory – EXFO FastReporter 2 User Manual

Page 371

Advertising
background image

Post Processing Software

361

C Measuring Polarization Mode

Dispersion: Theory

The dispersion phenomenon is described as a number of
non-intensity-dependant physical occurrences resulting directly in signal
loss (pulse spreading or time jitter in a digital system; distortion in an
analog system). PMD is an important type of signal dispersion. As
unrepeated link distances are increased and transmission rates pushed
upwards, PMD can significantly reduce system performance.

To better understand the impact of PMD, consider the example of a pulse
passing through a wave plate. Upon entering the wave plate, the pulse is
decomposed into polarization components aligned with each of the two
birefringent axes of the plate (known as the fast and slow axes). The
components propagating independently through the wave plate at different
group velocities will recombine at the end of the wave plate as a
superposition of two pulses split in time.

The delay between those pulses is designated as the differential group
delay (DGD) and is written as . For a Gaussian, unchirped input pulse
of rms width , the rms width at the output is given by

where r is the fraction of the input-pulse energy launched into one of the
birefringence axes.

Worst-case spreading occurs when the signal is perfectly split in two
(r = 1/2), while no spreading occurs if the input state of polarization
(SOP) of the launched signal is aligned with one of the birefringence axes.

To generalize this example, consider a long, weakly birefringent
telecommunications fiber as a concatenation of many, randomly
orientated birefringent wave plates. Each interface between two wave
plates redistributes the optical energy along both axes of the subsequent
wave plate. This transfer of energy is called mode coupling.

2

0

2

r

0

1 r

0

 

2

+

=

Advertising