Vector Quantized Semantic Communication System

Although analog semantic communication systems have received considerable attention in the literature, there is less work on digital systems. In this paper, we develop a deep learning (DL)-enabled vector quantized (VQ) system for image transmission, named VQ-DeepSC. Specifically, propose convolutional neural network (CNN)-based transceiver to extract multi-scale features of images and introduce embedding spaces perform feature quantization, rendering data compatible with Furthermore, employ adversarial training improve quality by introducing PatchGAN discriminator. Experimental results demonstrate that proposed VQ-DeepSC more robustness than BPG has comparable MS-SSIM performance DeepJSCC method.