Digital Signal Processing for Transmission Systems with Fibre Optical Parametric Devices

Abstract

Fibre-optical parametric devices, including two candidates fibre-optical parametric amplifier and optical phase conjugator, play crucial roles in the future of optical communications. They offer solutions for high gain with ultra-wide bandwidth, low noise and all-optical mitigation of optical channel impairments. However, the stimulated Brillouin scattering effect poses a fundamental challenge as it limits the pump power these devices need. Among suppression techniques, pump-phase modulation is widely used because of its modest cost although it introduces a temporal variation in the devices’ transfer functions. In this thesis, we propose digital signal processing algorithms, including both parametric and non-parametric approaches, which can compensate for penalties induced by pump-phase modulation schemes. We focus on the linear and nonlinear regression tasks, where the latter is solved by kernel methods. By carrying out numerical simulations, we prove the effectiveness of our proposed schemes against the existing compensation algorithms. Therefore, we believe the proposed algorithms will unlock the potentials of optical parametric devices, simplify the experimental design and become a vital part of parametric devices enabled transmission systems.