High-dimensional signal encoding and decoding method based on multi-ring perfect vortex beam

Abstract

The perfect optical vortex beam (POVB), renowned for its consistent beam radius across various orbital angular momentum (OAM) modes, excels in applications such as optical tweezers and communication. In our study, we achieved a significant milestone by creating a multi-ring perfect optical vortex beam (MR-POVB) through the superposition of multiple POVBs. This beam design maintains the inherent advantages of POVB while enabling multidimensional encoding via adjustable topological charges, radii, and intensities of each ring, thereby augmenting the free-space optical communication's channel capacity. To maximize MR-POVB's encoding potential, we introduced a novel encoding method tailored for vortex beams in optical communication. This method utilizes predefined rules to control beam transmission for efficient encoding. Additionally, we implemented a machine learning-based code group selection approach, combining cosine similarity and spectral clustering algorithms, to optimize encoding and enhance decoding accuracy. During decoding, MR-POVB's coaxial interference with the Gaussian beam reveals petal-like patterns and a deep learning approach is employed to accurately identify these features, facilitating signal decoding. The introduction of the MR-POVB encoding and decoding technique in this paper not only boosts channel capacity but also improves decoding accuracy, driving advancements in the field of free-space optical communication.