MUSA Receiver Algorithm Based on Residual Network Denoising

Multi-User Shared Access (MUSA) technology is a type of Non-Orthogonal Multiple Access (NOMA) solution based on extension sequences for future communication systems. Its characteristic is that the selection of extension sequences can comprise any combination of elements in the complex field. However, the received signal of the MUSA system is affected by noise and other factors during the transmission process, especially in task-oriented communication systems. The accuracy of signal restoration can significantly affect task completion. In this paper, we propose a complex field quintuple code and a receiver algorithm based on residual network denoising to perform signal processing and interference elimination, thereby improving the accuracy of signal restoration while conducting performance testing of the restored signals. The proposed algorithm takes inter-symbol interference into account and can effectively reduce noise. Simulation results demonstrate the proposed scheme in terms of detection error and decoding results compared to the benchmark at the MUSA receiver.