Kalman Filtering for Mitigation of Optical Fiber Transmission Impairments

In this thesis, the potential of Kalman filtering is exploited for the joint mitigation of several optical transmission impairments including laser phase noise, fiber nonlinearity, amplitude noise, frequency offset as well as polarization effects. A carrier phase and amplitude noise estimation (CPANE) algorithm is proposed and implemented using an extended Kalman filter (EKF) that estimates a complex quantity to track the phase and amplitude noise, simultaneously. Although, various DSP algorithms have been studied in this thesis, more emphasis will be given to the EKF-CPANE algorithm. Its performance is investigated in detail and compared to the conventional DSP algorithms. Approaches to enhance the nonlinear tolerance of the EKF-CPANE algorithm by incorporating with the existing techniques like digital backward propagation (DBP) will be presented. A two stage EKF approach is introduced that exhibits an improved tolerance towards phase and frequency offsets. Furthermore, an adaptive and cascaded Kalman filtering (CKF) is proposed for the joint tracking of polarization state and phase noise. A brief analysis on incorporating forward error correction (FEC) with the EKF-CPANE algorithm is also discussed. Extensive numerical investigations prove that the Kalman filters offer an attractive solution to jointly compensate several optical transmission impairments and thereby, enhance the transmission performance. Moreover, owing to their real-time feasibility and low complexity, Kalman filters seem to be a promising component of future coherent receivers.