Deep learning with synthetic photonic lattices for equalization in optical transmission systems

Abstract

In this work we propose a new physical realization of optical neural network (ONN) based on a recently appeared technological platform of synthetic photonic lattices (SPL),1, 2 and demonstrate its capabilities for deep learning. The system operates with time series of optical pulses with ability to easily control their parameters and possesses the architecture that well suits the ONN paradigm. We have also shown that such an ONN can be potentially utilized for signal processing in optical communication lines for signal distortion compensation. Show more