Radio-Frequency Multi-Mode OAM Detection Based on UCA Samples Learning

TL;DR

This paper introduces a machine learning-based method for detecting radio-frequency OAM modes using UCA samples, which is robust to antenna misalignments and enhances practical applicability.

Contribution

It proposes a novel learning-based OAM detection approach that handles non-parallel misalignments, outperforming classical methods in robustness and generalization.

Findings

BPNN classifier shows the best performance among tested classifiers.

Learning-based methods are robust to misalignment errors.

Simulation confirms improved detection accuracy with the proposed approach.

Abstract

Orbital angular momentum (OAM) at radio-frequency provides a novel approach of multiplexing a set of orthogonal modes on the same frequency channel to achieve high spectral efficiencies. However, classical phase gradient-based OAM mode detection methods require perfect alignment of transmit and receive antennas, which greatly challenges the practical application of OAM communications. In this paper, we first show the effect of non-parallel misalignment on the OAM phase structure, and then propose the OAM mode detection method based on uniform circular array (UCA) samples learning for the more general alignment or non-parallel misalignment case. Specifically, we applied three classifiers: K-nearest neighbor (KNN), support vector machine (SVM), and back-propagation neural network (BPNN) to both single-mode and multi-mode OAM detection. The simulation results validate that the proposed learning-based OAM mode detection methods are robust to misalignment errors and especially BPNN classifier has the best generalization performance.