On the Performance of Autoencoder-Based Wireless Communication with Low-Resolution Phase Quantization

Abstract

Low-resolution analog-to-digital converters (ADCs) are a promising way to cut power consumption in future 6G receivers. However, they break many assumptions behind traditional communication system design. Instead of optimizing each block separately, this work uses an end-to-end autoencoder (AE) to jointly learn and optimize the transmitter and receiver for systems with phase quantization. Two key obstacles arise: (1) phase quantization is non-differentiable, which blocks gradientbased training back to the transmitter, and (2) standard crossentropy loss leads to poor block error rate (BLER) performance. To overcome these issues, we introduce a differentiable approximation of phase quantization and a new loss function that adds a distance-based regularization term to improve symbol separability. Simulations show that the proposed AE achieves near-optimal BLER for uncoded $M$-ary phase shift keying $(M$ -PSK) and, in coded systems, can surpass conventional designs that rely on maximum-likelihood detection and separate channel decoding. These results highlight the AE's ability to jointly learn coding, modulation, detection, and decoding under low-resolution constraints.