Title:
DNN DECODING METHOD AND DECODING COMMUNICATION DEVICE FOR SCMA SYSTEM
Document Type and Number:
WIPO Patent Application WO/2020/125421
Kind Code:
A1
Abstract:
Disclosed in the present invention are a deep neural network (DNN) decoding method and a decoding communication device for an SCMA system, the method comprising: S1, constructing an SCMA system and obtaining a training sample data set; S2, establishing a DNN-based SCMA decoder model; S3, training said SCMA decoder model; and S4, deploying the SCMA decoder model, and decoding an SCMA signal by means of the SCMA decoder model. The present invention is able to improve the accuracy of SCMA decoding without increasing the complexity of an SCMA decoder. Compared with a traditional MPA algorithm based SCMA decoder, the DNN-based SCMA decoder in the present invention has improved performances in terms of calculation complexity and decoding bit error rate.
Inventors:
LIN JINZHI (CN)
ZHAO XIMIN (CN)
HU JINXING (CN)
ZHAO XIMIN (CN)
HU JINXING (CN)
Application Number:
PCT/CN2019/123144
Publication Date:
June 25, 2020
Filing Date:
December 05, 2019
Export Citation:
Assignee:
SHENZHEN INST OF ADV TECH CAS (CN)
International Classes:
H04L1/00; G06N3/04
Foreign References:
| CN109787715A | 2019-05-21 | |||
| CN107508775A | 2017-12-22 | |||
| CN107743103A | 2018-02-27 | |||
| CN105356971A | 2016-02-24 | |||
| CN108184128A | 2018-06-19 | |||
| US20160254937A1 | 2016-09-01 |
Attorney, Agent or Firm:
SHENZHEN KEJIN INTELLECTUAL PROPERTY OFFICE (CN)
Download PDF: