A Study On Polarization Stabilization In Optical Fiber Communication System

With the growing number of communication users and variety of communication services, the demand for high transmission rate increases rapidly. As the rising of optical fiber transmission rate, polarization mode dispersion and other problems caused by polarization instabiliy have become the main factor restricting the communication quality. How to quickly and accurately detect and stabilize the polarization state of light in high-speed optical communication system becomes a very important issue.Current polarization measurement schemes mainly concentrated on the design and analysis in optical domain, ignoring the convenience of hardware and richness of features. In this dissertation, advanced virtual instrument technology is induced to study the polarization control.In order to achieve polarization stabilization, the states of polarization of signals in the fiber are required to be measured as feedback signals to control the high speed polarization controller. Therefore, polarization measurement and polarization stabilization are the two aspects of this thesis. Specific innovations and achievements are as follows,1) A polarization measurement system based on Lab VIEW is developed. It is divided into three parts:data acquisition, data processing and data display. Since the transmission matrix of polarimeter changes with the environment, it directly affect the polarization measurement accuracy, a new approach of calibrating the polarization detector based on vector projection algorithm is utilized. Expermental results show that the polarization measurement system we have developed not only has good user interface and scalable functions, but also has high efficiency and accuracy, which can be applied in researches related to polarization measurement.2) Polarization stabilization in optical fiber communication is studied and control algorithm(PSO) has been improved. Refering to ideas in genetic algorithm, particles of higher fitness are colned and particles of lower fitness are eliminated. Since the total number of particals remains the same, the convergence speed and solution precision are improved. To test the effect of the algorithms, a two-dimensional multi-polar objective functions is used for numerical simulation. Simulation results show that compared with the basic particle swarm algorithm, the improved PSO algorithm performs better in convergence and global search capability.