A nonlinear signal equalizer based on optical neuromorphic computing networks
In HHI, we employ Optical neuromorphic computing architectures to mitigate nonlinear distortion due to a long-haul optical implementation as a solution for an energy-efficient as well as low-latency applications in 6G-technologies.
Funded by Federal Ministry of Education and Research (BMBF)
Duration: August 2021 - July 2025
6G-technologies demand for chronically increasing Internet traffic. To meet challenging high-speed, -capacity, and long-haul data transmission requirements, optical neuromorphic computing architectures introduce a promising solution for energy-efficient as well as low-latency applications. However, a long-haul optical implementation requires a nonlinear signal equalizer to mitigate nonlinear distortion effects of optical fibers at inevitable required high launch powers.
Within Technical Innovation Areas (TIAs) of the 6G Research and Innovation Cluster (6G-RIC), Neuromorphic computing group (TIA6) of HHI-Fraunhofer will demonstrate the nonlinear signal equalizer through the state-of-the-art research and international networking. The demonstration aims at improving training times owing to features of optical neuromorphic reservoir computing (RC) architectures, which is promising for an energy-efficiency and low-latency performance.