English

End-to-End Learning of Geometrical Shaping Maximizing Generalized Mutual Information

Signal Processing 2024-01-25 v1 Artificial Intelligence Information Theory math.IT Machine Learning
Authors: Kadir Gümüs , Alex Alvarado , Bin Chen , Christian Häger , Erik Agrell
View on arXiv ↗ PDF ↗ DOI ↗

Abstract

GMI-based end-to-end learning is shown to be highly nonconvex. We apply gradient descent initialized with Gray-labeled APSK constellations directly to the constellation coordinates. State-of-the-art constellations in 2D and 4D are found providing reach increases up to 26\% w.r.t. to QAM.

Keywords

end-to-end learning geometrical shaping generalized mutual information apsk constellations qam reach increase

Cite

@article{arxiv.1912.05638,
  title  = {End-to-End Learning of Geometrical Shaping Maximizing Generalized Mutual Information},
  author = {Kadir Gümüs and Alex Alvarado and Bin Chen and Christian Häger and Erik Agrell},
  journal= {arXiv preprint arXiv:1912.05638},
  year   = {2024}
}

Related papers

View all related →
Information Theory · Computer Science
End-to-end Learning for GMI Optimized Geometric Constellation Shape
Rasmus T. Jones, Metodi P. Yankov, Darko Zibar
2019-07-22
Information Theory · Computer Science
Geometric Constellation Shaping for Fiber Optic Communication Systems via End-to-end Learning
Rasmus T. Jones, Tobias A. Eriksson, Metodi P. Yankov, Benjamin J. Puttnam +3
2018-10-02
Information Theory · Computer Science
End-to-End Learning of Joint Geometric and Probabilistic Constellation Shaping
Vahid Aref, Mathieu Chagnon
2021-12-10
Information Theory · Computer Science
End-to-end Learning of Probabilistic and Geometric Constellation Shaping with Iterative Receivers
Harindu Jayarathne, Dileepa Marasinghe, Nandana Rajatheva, Matti Latva-aho
2025-10-28
Information Theory · Computer Science
Joint Learning of Geometric and Probabilistic Constellation Shaping
Maximilian Stark, Fayçal Ait Aoudia, Jakob Hoydis
2019-08-30
Signal Processing · Electrical Eng. & Systems
End-to-end Learning of a Constellation Shape Robust to Variations in SNR and Laser Linewidth
Ognjen Jovanovic, Metodi P. Yankov, Francesco Da Ros, Darko Zibar
2022-04-26
Information Theory · Computer Science
End-to-end Waveform Learning Through Joint Optimization of Pulse and Constellation Shaping
Fayçal Ait Aoudia, Jakob Hoydis
2021-07-15
Information Theory · Computer Science
Deep Learning of Geometric Constellation Shaping including Fiber Nonlinearities
Rasmus T. Jones, Tobias A. Eriksson, Metodi P. Yankov, Darko Zibar
2018-05-11
Signal Processing · Electrical Eng. & Systems
Model-Based Deep Learning of Joint Probabilistic and Geometric Shaping for Optical Communication
Vladislav Neskorniuk, Andrea Carnio, Domenico Marsella, Sergei K. Turitsyn +2
2022-04-18
Information Theory · Computer Science
End-to-End Learning of Probabilistic Constellation Shaping through Importance Sampling
Shrinivas Chimmalgi, Laurent Schmalen, Vahid Aref
2025-06-23
Information Theory · Computer Science
End-to-end Learning for OFDM: From Neural Receivers to Pilotless Communication
Fayçal Ait Aoudia, Jakob Hoydis
2021-07-30
Signal Processing · Electrical Eng. & Systems
Rate Adaptive Autoencoder-based Geometric Constellation Shaping
Ognjen Jovanovic, Metodi P. Yankov, Francesco Da Ros, Darko Zibar
2023-01-04
Information Theory · Computer Science
High-Cardinality Geometrical Constellation Shaping for the Nonlinear Fibre Channel
Eric Sillekens, Gabriele Liga, Domaniç Lavery, Polina Bayvel +1
2022-09-19
Signal Processing · Electrical Eng. & Systems
End-to-End Deep Learning of Long-Haul Coherent Optical Fiber Communications via Regular Perturbation Model
Vladislav Neskorniuk, Andrea Carnio, Vinod Bajaj, Domenico Marsella +3
2021-07-27
Signal Processing · Electrical Eng. & Systems
End-to-end Learning of a Constellation Shape Robust to Channel Condition Uncertainties
Ognjen Jovanovic, Metodi P. Yankov, Francesco Da Ros, Darko Zibar
2022-06-08
Machine Learning · Computer Science
On the Global Convergence of Imitation Learning: A Case for Linear Quadratic Regulator
Qi Cai, Mingyi Hong, Yongxin Chen, Zhaoran Wang
2019-01-15
Machine Learning · Statistics
Stochastic Particle Gradient Descent for Infinite Ensembles
Atsushi Nitanda, Taiji Suzuki
2017-12-18
Computer Vision and Pattern Recognition · Computer Science
End-to-end training of a two-stage neural network for defect detection
Jakob Božič, Domen Tabernik, Danijel Skočaj
2020-07-16
Signal Processing · Electrical Eng. & Systems
High-Cardinality Hybrid Shaping for 4D Modulation Formats in Optical Communications Optimized via End-to-End Learning
Vinícius Oliari, Boris Karanov, Sebastiaan Goossens, Gabriele Liga +5
2021-12-21
Computer Vision and Pattern Recognition · Computer Science
End-to-end Training for Whole Image Breast Cancer Diagnosis using An All Convolutional Design
Li Shen
2019-10-08
Machine Learning · Computer Science
Neighbourhood Distillation: On the benefits of non end-to-end distillation
Laëtitia Shao, Max Moroz, Elad Eban, Yair Movshovitz-Attias
2020-10-12
Optimization and Control · Mathematics
Information Geometry of Exponentiated Gradient: Convergence beyond L-Smoothness
Yara Elshiaty, Ferdinand Vanmaele, Stefania Petra
2025-04-08
Machine Learning · Computer Science
Guided Learning of Nonconvex Models through Successive Functional Gradient Optimization
Rie Johnson, Tong Zhang
2020-07-01
Machine Learning · Computer Science
End-to-End Learning on Multimodal Knowledge Graphs
W. X. Wilcke, P. Bloem, V. de Boer, R. H. van t Veer
2023-09-06
Differential Geometry · Mathematics
An end to end gluing construction for metrics of constant Q-curvature
A. Sophie Aiken, Rayssa Caju, Jesse Ratzkin, Almir Silva Santos
2025-03-13