Optimizing Bipolar Constellations for High-Rate Transmission in Short-Reach Fiber Links with Direct Detection

TL;DR

This paper enhances short-reach fiber communication by optimizing bipolar modulation and employing a neural network equalizer, achieving over 100 Gbit/s data rates with direct detection.

Contribution

It introduces a novel optimization of bipolar modulation with bias offset and a neural network equalizer for improved high-rate transmission.

Findings

Neural network equalizer outperforms standard receivers by over 100 Gbit/s.

Optimized bipolar modulation with bias offset improves data rates.

Neural network-based equalization effectively mitigates interference.

Abstract

Bipolar modulation increases the achievable information rate of communication links with direct-detection receivers. This paper optimizes bipolar transmission with a modulator bias offset for short-reach fiber links. A neural network equalizer with successive interference cancellation is shown to gain over 100 Gbit/s compared to standard receivers.