End-to-End Optimization and Equalization based on Deep-Learning for Fiber-Terahertz Integrated Communication System at 209 GHz

We proposed and experimentally demonstrated a bit-wise end-to-end autoencoder with post-equalization function for optimization based on deep learning in a fiber-THz integrated communication system at 209 GHz. Using this method, we achieved a higher transmission speed of 4 Gbps and a sensitivity gain of 1 dB compared to the conventional 32-QAM single-carrier signal.