Latency-Efficient Parallelized Carrier Phase Recovery Algorithm Based on Extended Kalman Filter for Short-Reach Coherent Optical Transmission Systems

In this paper, a latency-efficient parallelized carrier phase recovery algorithm based on the extended Kalman filter with linear interpolation (LEP-EKF) is proposed and experimentally verified for short-reach coherent optical transmission systems. Compared with the traditional parallelized extended Kalman filter (EKF), the processing latency of LEP-EKF is reduced by improving the parallelism of each data block. Moreover, the linewidth tolerance of LEP-EKF has almost no penalty because the linear interpolation is used to quick convergence of each data block without extra overhead. The simulation results show that compared with the traditional parallelized EKF, the processing latency of LEP-EKF is reduced by 75% at the similar linewidth tolerance of
$\Delta {\mathcal{f}}{\mathbf{T}}=9\times10^{-5}$
. The LEP-EKF is further experimentally demonstrated under the linewidth symbol duration product of
$\Delta fT=5\times 10^{-5}$
. The experimental results show the processing latency of LEP-EKF also saves 75% with little performance penalty. In conclusion, the LEP-EKF is a potential technique in short-reach coherent optical transmission systems.