Quantum Communication for Semantic Communication Based Image Transmission

The growing demand for reliable image transmission in noisy vehicular networks exposes the limitations of classical communication, which often suffers from bandwidth constraints and susceptibility to noise. Classical semantic communication addresses this by focusing on essential information, but it still struggles to maintain consistent image quality under adverse channel conditions. To address these limitations, this study proposes a quantum communication system for semantic communication-based image transmission by leveraging quantum superposition. The system first uses a semantic encoder to generate a compact latent representation. This representation is subsequently processed by a channel encoder and mapped into quantum superposition states using a quantum encoder. The resulting quantum states are sent through a noisy quantum channel, where they are measured and decoded at the receiver to recover the corresponding classical bitstream. The bitstream is then channel decoded to recover the latent vector, which is used by a semantic decoder to reconstruct the image. Results from the experiments confirm that the proposed system outperforms classical semantic communication systems, achieving a peak signal-to-noise ratio (PSNR) of up to 28.55 dB, a structural similarity index (SSIM) of up to 0.7996, a universal quality index (UQI) of up to 0.7836, and perfect classification accuracy at channel SNR levels as low as 6 dB, as evaluated by a convolutional neural network (CNN). These results highlight the system's effectiveness in task-oriented communication and show that combining quantum and semantic communication enhances image quality and reliability in challenging vehicular scenarios.