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.

© 2020 The Author(s)

PDF Article
More Like This
Fiber Nonlinearity Mitigation Scheme based on Geometric Constellation Shaping via End-to-end Auto-encoder Learning and KNN Deciding

Yuanru Zang, Yan Zhao, and Xue Chen
M4A.276 Asia Communications and Photonics Conference (ACPC) 2020

Maximization of the Achievable Mutual Information using Probabilistically Shaped Squared-QAM Constellations

D. Pilori, F. Forghieri, and G. Bosco
W2A.57 Optical Fiber Communication Conference (OFC) 2017

Comparison of Modulation Formats Based on Generalized Mutual Information

Shaoliang Zhang, Fatih Yaman, Eduardo Mateo, and Yoshihisa Inada
M2A.3 Optical Fiber Communication Conference (OFC) 2016

References

You do not have subscription access to this journal. Citation lists with outbound citation links are available to subscribers only. You may subscribe either as an OSA member, or as an authorized user of your institution.

Contact your librarian or system administrator
or
Login to access OSA Member Subscription