VQ-VAE Empowered Wireless Communication for Joint Source-Channel Coding and Beyond

Vector Quantized Variational Autoencoder (VQ-VAE) has shown promise in representing diverse and complex data distributions in deep learning, making it a potential solution for various applications including wireless communications. In this paper, we propose a joint source-channel coding scheme based on VQ-VAE for point-to-point wireless communication. Our approach leverages the dependence of the encoder and decoder on a given dataset and channel conditions to develop efficient encoding and decoding schemes, leading to improved reliability and efficiency even in the presence of noisy wireless channels. We demonstrate the effectiveness of our proposed approach through extensive simulations in handling realistic wireless communication scenarios. In addition, we discuss potential connections to semantic communication and highlight the secure and energy-efficient nature of our approach.