Modulation-Transparent and Robust Frequency Offset and Phase Tracking Scheme Using Self-Learning Kalman Filter for Intelligent Receiver

A joint and robust frequency offset (FO) and phase noise tracking scheme using self-learning Kalman filter is proposed and demonstrated experimentally. In our scheme, a self-learning Kalman filter based on kurtosis error is used to track FO and phase noise simultaneously in a decision-free manner. Then, a mean-shift clustering-assisted maximum likelihood estimator is used for residual phase noise compensation. Simulation results indicate that the proposed scheme has advantages in FO estimation accuracy, OSNR sensitivity penalty, and FO jitter rate tolerance and nonlinearity tolerance, compared with the conventional FO and phase noise estimation scheme. With 7% FEC threshold, the proposed scheme can tolerate FO jitter rate up to 10 MHz for 16QAM, which is around 100 times and 20 times than that of VVPE1 +BPS and FFT+UCPE. With 1 dB OSNR penalty, the linewidth tolerance of the proposed scheme is 300 kHz for 64QAM, which is around 30 times and 5 times than that of FFT+UCPE+ML and FFT+ VVPE2. The proposed scheme also shows excellent nonlinearity tolerance over wide launched power, and it provides around 0.4 dB and 0.2 dB Q-factor improvement over VVPE1+ VVPE2, VVPE1 +BPS for 16QAM, respectively. The 28GBaud PDM 16QAM experimental results also indicate that the proposed scheme outperforms the rest of tested scheme under both back-to-back and fiber transmission cases. With optimal launched power, the proposed scheme can provide around 0.2 dB, 0.4 dB, and 0.7 dB Q-factor improvement over VVPE1 +BPS, FFT+UCPE, and VVPE1+ VVPE2 after 915 km fiber transmission, respectively. To the best of our knowledge, it is the first time that a joint and modulation transparent with robust nonlinearity and FO jitter rate tolerance is proposed and demonstrated for 64QAM.