Two stage extended Kalman filtering for joint compensation of frequency offset, linear and nonlinear phase noise and amplitude noise in coherent QAM systems

Abstract

We investigate our proposed two stage extended Kalman filtering (EKF) technique for the joint tracking of frequency offset (FO), laser phase noise, fiber nonlinearity as well as amplitude noise. The EKF employed in the first stage coarsely compensates the FO using a set of training data symbols. In the second stage, the employed EKF accomplishes the task of compensating the residual FO, phase noise arising from laser linewidth, fiber nonlinear effects along with simultaneous amplitude noise mitigation. The transmission performance of the proposed two stage EKF has been verified through numerical simulations on polarization multiplexed (PM) 16-quadrature amplitude modulation (QAM) system operating at 28 Gbaud. The results prove that the proposed technique is well tolerant towards the impairments of FO, linear and nonlinear phase noise and amplitude noise even up to ≈3000 km of standard single mode fiber (SSMF) transmission at a launch power of 3 dBm, giving a bit error rate (BER) of ≈2.4e-2 which is the 20% soft decision forward error correction (SD-FEC) threshold. Furthermore, the proposed two stage EKF outperforms a similar two stage approach implemented using linear Kalman filter (LKF) with improved BER.