Low-Cost OAM Spatial Oversampling Receiver With 1-bit Quantized Comparators and DL-Based Detector

Abstract

For existing uniform circular array (UCA)-based orbital angular momentum (OAM) point-to-point communications, the OAM signal can only be detected when the number of antennas of the transmitter and receiver are equal. To break through this barrier, in this paper, we first validate the feasibility of spatial oversampling for OAM reception, which can bring the additional signal-to-noise ratio (SNR) gain at the receiver, thus positively influencing the system's bit error rate (BER). However, more antennas imply the use of high-precision analogto-digital converters (ADCs) in additional radio frequency (RF) chains, which would result in significant power consumption and costs. Equipping low-cost 1-bit ADCs in the RF chain is a viable approach, but it leads to the deterioration of the OAM receiver's BER. Therefore, we apply comparators and a deep learning (DL)-based detector to the 1-bit quantized OAM receiver to improve the BER. Numerical simulations validate that our proposed OAM receiver can obtain a better BER and reduce the receiver hardware costs.Index Terms-Orbital angular momentum (OAM), spatial oversampling, 1-bit analog-to-digital converter (ADC), comparator, deep learning (DL). I. INTRODUCTIONRecent studies show that vortex waves with orbital angular momentum (OAM) can be utilized in fiber and wireless communications [1]- [5]. Theoretically, OAM has an infinite number of orthogonal modes [6], which can be considered as a novel degree of freedom to increase channel capacity and spectral efficiency beyond conventional resources [7], [8]. Considering the compatibility with the existing communication network based on the antenna array, the uniform circular array (UCA) is the mainstream method for generating OAM beams in wireless communications [1], [7]. It is generally believed that only an N × N -dimensional circulant matrix can be diagonalized by the Fourier matrix, thus existing UCA-based OAM communication systems are all modeled with the same number of elements at the transmitter and receiver [8]-[10]. How to conduct spatial oversampling in OAM communications to leverage additional antennas for signal-to-noise ratio (SNR) gain remains an open issue.Z. Chen, W.-X.