Joint equalization of linear impairments using two-stage cascade Kalman filter structure in coherent optical communication systems

Abstract

In this paper, instead of adopting different algorithms to deal with the different impairments separately based on different signal processing structures, we proposed a joint linear impairment equalization scheme using a two-stage cascade Kalman structure in which the first stage solves the residual chromatic dispersion (rCD), the rotation of state of polarization (RSOP) and polarization mode dispersion (PMD), and the second stage compensates for the carrier frequency offset (CFO) and carrier phase noise (CPN). The structures of the two stages are similar with a seamless connection and possess a clear logic architecture. The stages are compatible with each other and achieve excellent performance. We aim to use this two-stage Kalman scheme as a comprehensive DSP equalization scheme to replace the combined algorithms of the constant modulus algorithm (CMA), improved Mth-power algorithm (IMP), and blind phase search (BPS). Simulations based on a 28 Gbaud polarization division multiplexed quadrature-phase shift keying coherent optical communication system show that the proposed two-stage Kalman scheme comprehensively outperforms the CMA-IMP-BPS scheme in solving the problems of RSOP, PMD, rCD, CFO, and CPN. The experimental results confirm the simulation results. In addition, the complexity of the proposed Kalman scheme is relatively less than the CMA-IMP-BPS scheme.