Uplink NOMA Semantic Communications: Semantic Reconstruction for SIC

Semantic communication aims to significantly improve transmission efficiency and reliability by extracting and transmitting information that best reflects the user’s intention. In this paper, we propose an uplink NOMA semantic communication framework for multi-user text transmission. In order to achieve successive interference cancellation (SIC) for signals containing semantic information, we propose three signal reconstruction schemes to remove interference from the signal through different reconstruction networks. Furthermore, a training algorithm for jointly updating receiver’s network parameters is proposed to ensure the transmission accuracy of different users. The simulation results show that the proposed channel reconstruction scheme has high reliability and transmission accuracy, and the jointly training algorithm reduces multi-access interference between users to ensure accurate data recovery of each signal at the receiver.