Ultra-fast SOP rotation tracking and its enabled feedforward adaptive equalization based on superimposed chirp pilot

Abstract

We propose to track state of polarization (SOP) rotation by utilizing chirp pilots processed through fractional Fourier transform (FrFT). Superimposed on dual-polarization data signal, the chirp pilots with different FrFT rotation orders can converge to fractional energy peaks and then be distinguished by combining FrFT with MIMO structure. Consequently, SOP rotation resulting in energy distribution can be estimated by leveraging aggregated peaks in the matched fraction domain. The feasibility of the proposed scheme is verified by an experiment of 480Gb/s DP-16QAM and 600Gb/s DP-32QAM signals over 25 km fiber. Without sacrificing spectrum efficiency, the proposed approach can track SOP rotation up to 5 Mrad/s. To reduce the impact of superimposition, a fractional domain digital match filter (DMF) is proposed to improve the optical signal-to-noise ratio (OSNR) penalty from more than 4.0 dB to 0.9 dB. Furthermore, the SOP rotation estimated prior to MIMO facilitates the implementation of pre-equalization and adaptive step-size of subsequent equalization, thereby achieving feedforward adaptive equalization (FFD-AEQ). Experimental results demonstrate that the FFD-AEQ can cope with SOP rotation up to 4.1 Mrad/s under HD-FEC. Besides, the FFD-AEQ employing pre-equalization has a 39.1% complexity consumption decrease compared to the conventional MIMO.