Emulation and Simulation of Polarization Dependent Loss and Polarization Mode Dispersion in Optical Fibre Communication Systems

  • Yangzi Liu

Student thesis: Doctoral ThesisPhD


Polarization becomes a severe obstacle in optical fibre transmission systems when the bit rate increases. Polarization affects the system performance in two ways: polarization mode dispersion (PMD) and polarization dependent loss (PDL). As PMD has been researched for decades, there are a few methods to compensate the distortion caused by it. However, even though PDL was considered a minor issue to be compared with PMD, the combined effects between PDL and PMD distorts the PMD statistics from its theoretical distribution.

In order to study the PDL statistics distribution over time and long transmis- sion distance in laboratorial environment, it is essential to have a reliable PDL source which is able to generate similar results to the real optical fibre communi- cation systems. As it is not economical to build an emulator with an ultra-long fibre link in a laboratory, a simulation model was developed as an alternative approach to accomplish the mission. To confirm the algorithm of the simulation model works properly, a PDL emulator for short fibre links can be built as a comparison.

In this thesis, a PDL emulator is designed with computer-driven polariza- tion controllers and single mode fibre. The measurement results show that the generated PDL obeys Maxwellian distribution when it is expressed in dB, as ex- pected. This was achieved with just a few polarization controllers. By adjusting the combination of the polarization controllers with different lengths of fibre links, it has been found that the positioning of polarization components can affect the PDL statistics results, even if the total length of the fibre link and the number of the involved polarization components remain the same. The results suggest that with a fixed initial state of polarization, the PDL variance reaches a minimum when each of the fibre lengths is approximately equal.

The simulation model is developed based on the approach of the Mueller matrix which is described later in the thesis. It is used to examine the statistics of PDL to be compared with the measurement from the emulator for short fibre link, in order to confirm it is reliable for long distance simulation. After that, it is extended to simulate a long distance transmission system.

A simulation model for pulse propagation has also been developed. It is used to investigate the effects of polarization in fibre communication systems. The simulation focuses on the signal attenuation caused by chromatic dispersion, nonlinear effects and polarization. The approach of optical phase conjugation (OPC) is also applied to compensate for the attenuation. Results show that the pulse widening caused by chromatic dispersion and nonlinear effects can be re- covered by OPC. However, due to randomness, the signal distortion caused by polarization cannot be recovered with OPC.

The pulse simulation model was altered to study the impact on signal-to- noise ratio caused by the combined effect of PMD and thermal noise. An optical eye diagram was formed and the results showed that PMD contributes horizon- tally reducing the eye width and noise contributes vertically reducing the eye height. A numerical approach was provided to reduce the pulse jitter caused by PMD and it worked well to increase eye width.

The pulse simulation model was also altered to study the combined effects of PMD and nonlinearity. The results showed that polarization mode dispersion can distort the pulse shape which partly prevents the optical phase conjuga- tion’s function recovering the pulse spreading from nonlinearity. The results also showed the similarity between the combined effects of PMD and PDL and the combined effect of fibre nonlinearity and PMD. That leads to an conclusion that the combined effect of PMD and PDL can be achieved with the combined effect of PMD and nonlinearity followed by an attenuator.
Date of Award3 Apr 2019
Original languageEnglish
Awarding Institution
  • University of Bath
SupervisorRobert Watson (Supervisor) & Duncan Allsopp (Supervisor)

Cite this