Photonic Extreme Learning Machine based on frequency multiplexing

TL;DR

This paper introduces a photonic Extreme Learning Machine using frequency multiplexing in fiber optics, demonstrating its potential for fast, parallel neural network processing with experimental validation on classification and channel equalization tasks.

Contribution

It presents a novel photonic ELM architecture leveraging frequency multiplexing, with both simulation and experimental results showcasing its effectiveness.

Findings

Successful classification and channel equalization with the photonic ELM

Optical multiplication by output weights demonstrated

Potential for high-speed, parallel neural processing in optics

Abstract

The optical domain is a promising field for physical implementation of neural networks, due to the speed and parallelism of optics. Extreme Learning Machines (ELMs) are feed-forward neural networks in which only output weights are trained, while internal connections are randomly selected and left untrained. Here we report on a photonic ELM based on a frequency-multiplexed fiber setup. Multiplication by output weights can be performed either offline on a computer, or optically by a programmable spectral filter. We present both numerical simulations and experimental results on classification tasks and a nonlinear channel equalization task.