Optical communication, optical information science, and optical computations

Detection of superposition of twisted beams under noise with the use of artificial neural networks

V. V. Chechetkin^a, A. V. Semenova^bc, M. B. Chernyaeva^ab, V. L. Vaks<sup>abc

a</sup> National Research Lobachevsky State University of Nizhny Novgorod
^b Institute for Physics of Microstructures, Russian Academy of Sciences, Nizhnii Novgorod
^c Ioffe Institute, St. Petersburg

Abstract: Twisted beams with non-zero orbital angular momentum projection onto the direction of motion have attracted sufficient interest in recent years. Such beams with non-zero orbital angular momentum can be used in wireless communication networks to improve efficiency and throughput.The work is aimed to evaluate the efficiency of the artificial neural network for the recognition of the superposition of twisted beam under conditions of noise in the receiving channel, as well as variations in the orientation and spatial positioning of images of intensity profiles of the superposition of two twisted beams. The results could be applied for analysis of the potential of the artificial networks in wireless optical communications.

Keywords: Laguerre–Gaussian beams, orbital angular momentum, topological charge, optical communications, free space optics, optical angular momentum shift-keying, multiplexing, machine learning, convolutional neural networks, LeNet.

Received: 02.06.2025
Revised: 22.08.2025
Accepted: 03.09.2025

DOI: 10.61011/OS.2025.09.61765.8233-25