研究者所属(当時) | 資格 | 氏名 | |
---|---|---|---|
(代表者) | 理工学術院 基幹理工学部 | 教授 | 前原 文明 |
- 研究成果概要
Sparse code multiple access (SCMA) is a promising approach for massive machine type communication (mMTC) which is one of the usage scenarios for the fifth-generation mobile communications systems (5G). SCMA improves the system capacity at relatively high CNR, but in low CNR, its non-orthogonality between users makes system capacity smaller than orthogonal frequency division multiple access (OFDMA). So far, we have proposed a hybrid approach using simultaneously SCMA and OFDMA to cope with this problem, and its effectiveness has been validated. In this work, we propose a system selection method employing deep learning based on user position information. In the proposed method, the multiple access scheme that maximizes the system throughput is selected from SCMA-only, OFDMA-only, and their hybrid approach. The effectiveness of the proposed scheme is demonstrated in terms of system throughput under different user distributions via computer simulations.