Deep learning-based detection for moderate-density code multiple access in IoT networks

Han, Yu, Wang, Zhenyong, Guo, Qing, and Xiang, Wei (2020) Deep learning-based detection for moderate-density code multiple access in IoT networks. IEEE Communications Letters, 24 (1). pp. 122-125.

[img] PDF (Published Version) - Published Version
Restricted to Repository staff only

View at Publisher Website:


In the era of the Internet of Things, massive devices communications become a fundamental problem. To improve spectral efficiency and reduce latency, a new non-orthogonal multiple access scheme dubbed moderate-density code multiple access (MCMA) is presented. We also propose a new deep learning-based multi-user detection algorithm for MCMA systems, which is based upon a new graphic representation of the Tanner graph for the message passing algorithm (MPA). The proposed algorithm learns to adjust the weights of the edges of the neural network to realize multi-user detection without iterations as required in conventional MPA algorithms. Experimental results show that with an increase in the overloading factor and the number of users, the BER performance of the proposed scheme is better than that of deep learning-aided SCMA (DL-SCMA) with a lower computational complexity.

Item ID: 62258
Item Type: Article (Research - C1)
ISSN: 1558-2558
Keywords: Internet of Things, MCMA, multi-user detection, deep learning, low complexity
Copyright Information: © 2019 IEEE.
Funders: National Natural Science Foundation of China (NNSFC), Fundamental Research Funds of Shenzhen Innovation of Science and Technology Commitee (FRFSISTC)
Projects and Grants: NNSFC 61601147, NNSFC 61371100, FRFSISTC JCYJ20160331141634788
Date Deposited: 12 Feb 2020 07:41
FoR Codes: 40 ENGINEERING > 4006 Communications engineering > 400608 Wireless communication systems and technologies (incl. microwave and millimetrewave) @ 100%
More Statistics

Actions (Repository Staff Only)

Item Control Page Item Control Page