In-Service Blind Transceiver IQ Imbalance and Skew Monitoring in Long-Haul Non-Dispersion Managed Coherent Optical Systems

Cheng Ju,Na Liu,Changhong Li

doi:10.1109/access.2019.2947467

Cheng Ju, Na Liu + Show 1 more

Open Access

https://doi.org/10.1109/access.2019.2947467

Copy DOI

Abstract

Transceiver imbalance can rapidly degrade system performance, especially when high symbol rate, high modulation order and low roll-off pulse shape are used. Here, a widely linear transmission model is derived to analyze the interaction of transceiver IQ gain imbalance, IQ phase imbalance, skew and fiber link for long-haul transmission. In-service blind transceiver imbalance measurement method is designed to calibrate and monitor the transceiver status. Two $4\times 2$ complex-valued equalizers operated at twice symbol rates are designed before and after carrier phase recovery (CPR) to compensate receiver and transmitter imbalances, where the transceiver imbalance could be derived from the taps of two converged equalizers. The impact of transceiver imbalance and fiber nonlinearity on the measurement accuracy are numerically evaluated based on a 42 GBaud 16-QAM system through 1500-km standard single fiber (SSMF) transmission.

Highlights

Coherent optical communication has been rapidly developed owing to its ability to improve receiver sensitivity, spectral efficiency, as well as to mitigate the transmission impairments by using digital signal processing (DSP) [1]
Receiver imbalance compensation and calibration methods are well discussed in the literatures, including Gram-Schmidt orthogonalization procedure for gain and phase imbalance compensation [2], 4×2 complex-valued equalizer for gain, phase imbalance and skew compensation and calibration [3], [4], 4×4 real-valued equalizer for gain, phase imbalance and skew compensation [5], [6], modified interpolator in timing recovery for skew compensation and calibration [7], off-line beat frequency skew calibration method agreed by Optical Internetworking
Transmitter imbalance is more critical for transmission and some recent works have attempted for compensation and calibration, such as indirect learning architecture-based skew calibration method for back-to-back case [10], 2×2 real-valued equalizer for imbalance compensation and calibration [11], adaptive source separation method for gain and phase imbalance compensation [12], off-line skew self-calibration with integrated lowbandwidth photodetector in commercial IQ modulator [13], reconfigurable interference method for skew calibration [14], clock tone extraction based on genetic algorithm for skew calibration [15], and clustering algorithm based gain and phase imbalance estimation [16]

Summary

INTRODUCTION

Coherent optical communication has been rapidly developed owing to its ability to improve receiver sensitivity, spectral efficiency, as well as to mitigate the transmission impairments by using digital signal processing (DSP) [1]. To the author’s best knowledge, no explicit model has been given to analyze the interaction of transceiver imbalance and fiber linear interference in long-haul transmission scenario. As main contribution of this work, theory model is derived based on widely linear transformation to explore the interaction between fiber link and transceiver imbalance. Linear equalizer based on stochastic gradient descent for taps updating can only compensate quasi-static interference, such as polarization-mode dispersion (PMD), chromatic dispersion (CD) and receiver imbalance, but can not track and compensate fast time-varying interference, such as frequency offset and laser phase noise. Based on the analysis model, two 4×2 complex-valued equalizers operating at two samples per symbol are utilized before and after CPR to mitigate receiver and transmitter imbalance, and the transceiver gain imbalance, phase imbalance and skew can be extracted from the converged tap coefficients. The effect of transmitter imbalance on receiver imbalance estimation, the effect of receiver imbalance on transmitter imbalance estimation, and the effect of fiber nonlinearity on transceiver imbalance estimation are evaluated in the following

THEORETICAL MODELING AND IMBALANCE MONITORING

SIMULATION AND RESULTS

CONCLUSION