The Design Of High-speed Polarization Scrambler Based On FPGA

ABSTRACT:As the transmission speed of single channel increasing, polarization dependent loss starts to limit the performance of optical transmission system. One efficient way for reducing polarization dependent loss is using polarization scrambler. Scrambling speed, scrambling uniformity, insertion loss, operating wavelength and the volume are main key indications for a polarization scrambler. As signal speed rising, polarization scrambler with higher capability is needed. In this thesis, based on theories of polarized light and technology of controlling polarization states, a high speed polarization scrambler based on FPGA was proposed for 160Gbit/s optical time division multiplexing (OTDM) systems. This thesis consists of following parts:1. Based on related theories of polarized light and technology of controlling polarization states, physic mechanisms and key properties of many kinds of polarization controller were studied. Compared advantages and disadvantages of many polarization controllers, LiNiO3 based polarization controller was selected to fabricate high-speed polarization scrambler for 160Gbit/s optical systems.2. Key properties of polarization controller used in 160Gbit/s optical system were studied, especially performances of polarization controller with different polarization scrambling speeds. According to requirements of polarization controller, control circuits for polarization controller were designed.3. The generation algorithms of pseudo-random numbers based on FPGA were studied, especially linear feedback shift register (LFSR) algorithm and cellular automata (CA) algorithm. LSFR algorithm, CA90 algorithm and CA150 algorithm were realized and simulated using FPGA by Verilog language respectively. Statistical properties and randomness of pseudo-random numbers generated by different algorithms were studied. At last, CA algorithm which can realize better Statistical properties and randomness was selected for generating pseudo-random numbers for polarization scrambler.4. Maxwell distribution function was studied and program for generating data function in keeping with Maxwell distribution was composed and simulated by Matlab. The key part of polarization controller scrambler was generation of random numbers in keeping with Maxwell distribution as fast as we can. We realized this control part by Verilog language and simulate it using Modelsim. At last, high-speed polarization scrambling was implemented using our designed polarization scrambler. 5. Compared with fiber-squeezer-based polarization scrambler, performances of home made polarization scrambler were experimentally studied. Results indicate that no phase noise and amplitude modulation was demonstrated using proposed polarization scrambler. Lower nonlinear phase perturbations were achieved. All these indicate that proposed polarization scrambler was a good choice for high-speed optical systems. At last, we concluded our research works and some advanced research would be implemented in future.