A Silicon Photonic Neural Network for Chromatic Dispersion Compensation in 20 Gbps PAM4 Signal at 125 km and its Scalability up to 100 Gbps

2024-09-05

Journal of Lightwave Technology

AI Summary

Dispersion compensation for 20-Gbps PAM4 signals over 125-km standard single-mode fiber remains challenging. Method: This paper proposes a feedforward photonic neural network (PNN) architecture implemented on a silicon photonics platform. It employs an 8-tap delay-line-based complex perceptron for linear filtering and exploits the square-law response of photodetectors to introduce nonlinearity, forming a compact optical-domain equalizer. Contribution/Results: To our knowledge, this is the first demonstration of a programmable silicon photonic neural network for real-time dispersion equalization, revealing a physical correspondence between trained weights and the fiber’s dispersion transfer function. A hybrid training strategy—combining evolutionary algorithms with gradient-based optimization—is introduced, targeting eye-opening maximization. Experiments show significant BER reduction for 20-Gbps PAM4 after 125-km transmission. Simulations confirm scalability to 100 Gbps, demonstrating high bandwidth compatibility and hardware efficiency.

Technology Category

• Machine Learning: Hardware-aware ML
• Cognitive Modeling & Cognitive Systems: Neural Spike Coding
• Search and Optimization: Learning to Search

Application Category

• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Systems and Infrastructure for Web, Mobile and WoT: Experiences and lessons learnt from Web-based algorithms and system deployments
• Economics, Online Markets and Human Computation: Incentives in network design for Web infrastructures and ecosystems

Abstract

A feed-forward photonic neural network (PNN) is tested for chromatic dispersion compensation in Intensity Modulation/Direct Detection optical links. The PNN is based on a sequence of linear and nonlinear transformations. The linear stage is constituted by an 8-tap time-delayed complex perceptron implemented on a Silicon-On-insulator platform and acting as a tunable optical filter. The nonlinear stage is provided by the square modulus of the electrical field applied at the end-of-line photodetector. The training maximizes the separation between the optical levels (i.e. the eye diagram aperture), with consequent reduction of the Bit Error Rate. Effective equalization is experimentally demonstrated for 20Gbps 4-level Pulse Amplitude Modulated signal up to 125 km. An evolutionary algorithm and a gradient-based approach are tested for the training and then compared in terms of repeatability and convergence time. The optimal weights resulting from the training are interpreted in light of the theoretical transfer function of the optical fiber. Finally, a simulative study proves the scalability of the layout to larger bandwidths, up to 100 Gbps.

Problem

• Chromatic dispersion compensation in optical links
• Scalability of photonic neural networks to 100 Gbps
• Training methods for optical signal equalization

Innovation

• Silicon Photonic Neural Network
• Chromatic Dispersion Compensation
• Scalability to 100 Gbps