MRes year on the CDT in Connected Electronic and Photonic Systems

本计画的重点在于设计与评估模式群分多工网路的新供应策略。与假定所有模式和波长的噪声系数、光纤衰减和色散相等并忽略放大段的最新配置机制不同，这些战略将考虑到这些实际方面。特别是，最近的技术，以减少差分增益（模态和频谱）的多模放大器将包括在模型中。所设计的供应策略将从自动化到深度强化学习方法。

This project focuses on the design and evaluation of novel provisioning strategies for mode group division multiplexing networks. Unlike state-of-the-art provisioning mechanisms that assume equal noise figure, fibre attenuation and chromatic dispersion for all modes and wavelengths and neglect the amplification segment, the strategies will take these practical aspects into account. In particular, recent techniques to decrease the differential gain (modal and spectral) of multi-mode amplifiers will be included in the models. The provisioning strategies designed will range from heuristics to deep reinforcement learning approaches.