6G MIMO transmission-oriented signal detection method based on geometric structure learning

• Main Authors: XU DADAN, ZHANG SHIJIAO, DU QINGHE, ZHANG YUHAO
• Format: Patent
• Language: Chinese; English
• Published: 03.05.2024
• Subjects: ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; TRANSMISSION

The invention belongs to the technical field of signal detection, and relates to a 6G MIMO transmission-oriented signal detection method based on geometric structure learning, which comprises the following steps of: S1, establishing an MIMO transmission system model; s2, setting an initial sphere radius; s3, executing an intra-domain sphere decoding algorithm by taking the set initial sphere radius as a radius and taking a 0 vector as a receiving signal; storing all receiving constellation point sets in a radius d range, and recording as a set phi; s4, enabling # imgabs0 # to be a solution of MMSE (Minimum Mean Square Error), finding a signal closest to the received signal y from the sum of # imgabs1 # and the set phi, updating # imgabs2 #, and S5, repeating the step S4 until convergence; according to the method, the complexity is about 1/2-1/10 of the SD algorithm and is smaller than the complexity of the SD algorithm, the detection performance is close to ML, in addition, the storage complexity needed by th