Quality of Transmission Estimation for Multi-User Free Space Optical Communication Using Supervised Machine Learning
Authors:
- 21 June 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-5
Abstract
A methodology based on supervised machine learning (e.g., support vector machine [SVM], artificial neural network [ANN], random forest [RF]) is proposed and experimentally demonstrated for signal-to-noise ratio (SNR) estimation in a multi-user free-space optical communication (FSOC) link. The proposed methodology relies on amplitude information of received mixed signals when users share time, bandwidth, and space resources. Two approaches are investigated to identify input features for classification models for jointly estimating SNRs from two simultaneously communicating users overlapping into a single node receiver. The first approach uses the empirical histogram of the received mixed signal amplitude, while the second approach uses locations of local maxima of the empirical histogram. The latter method demonstrated highest performance; 92% classification accuracy was achieved using the SVM algorithm. Scalability analysis was conducted to identify required training data and time series samples for feature extraction to guarantee high classification performance.Keywords
This publication has 11 references indexed in Scilit:
- An Optical Communication's Perspective on Machine Learning and Its ApplicationsJournal of Lightwave Technology, 2019
- Nonlinear Signal-to-Noise Ratio Estimation in Coherent Optical Fiber Transmission Systems Using Artificial Neural NetworksJournal of Lightwave Technology, 2018
- Machine Learning Based Noise Estimation in Optical Fiber Communication Networks2018
- Artificial intelligence (AI) methods in optical networks: A comprehensive surveyOptical Switching and Networking, 2018
- Learning Process for Reducing Uncertainties on Network Parameters and Design MarginsJournal of Optical Communications and Networking, 2018
- A Comparative Survey of Optical Wireless Technologies: Architectures and ApplicationsIEEE Access, 2018
- Modulation Format Recognition and OSNR Estimation Using CNN-Based Deep LearningIEEE Photonics Technology Letters, 2017
- QoT Estimation for Unestablished Lighpaths using Machine Learning2017
- Survey on Free Space Optical Communication: A Communication Theory PerspectiveIEEE Communications Surveys & Tutorials, 2014
- Optical Performance Monitoring Using Artificial Neural Networks Trained With Empirical Moments of Asynchronously Sampled Signal AmplitudesIEEE Photonics Technology Letters, 2012