A hybrid transmission framework (DA-ESemCom) for post-disaster emergency wireless networks that sends scene images over UDP without retransmission. A lightweight analog deep-learning semantic codec carries the image content, while a digital distributed-source-coding branch (DCT plus quantization plus LDPC syndrome bits, QPSK modulated) corrects residual errors at a well-resourced receiver. A performance-constrained semantic coding model treats semantic and channel noise jointly, and a derived Cramer-Rao lower bound guides codec design. Over Rayleigh fading channels the framework beats classical separated source-channel coding and deep JSCC baselines on PSNR, MS-SSIM, LPIPS, and YOLO object-detection mAP.
Emergency Wireless Communication (EWC) networks adopt the User Datagram Protocol (UDP) to transmit scene images in real time for quickly assessing the extent of the damage. However, existing UDP-base…