首页|3M-Net: Automatic Modulation Recognition Based on Multiscale Mobile Inverted Bottleneck Convolution and Manhattan Self-Attention Network

3M-Net: Automatic Modulation Recognition Based on Multiscale Mobile Inverted Bottleneck Convolution and Manhattan Self-Attention Network

扫码查看
原文链接
  • NETL
  • NSTL
  • Wiley
The increasingly complex modern communication environment poses challenges for automatic modulation recognition (AMR)techniques. In AMR tasks, in order to more comprehensively capture signal features and improve recognition performance, wepropose a model named Multiscale Mobile Inverted Bottleneck Convolution and Manhattan Self-AttentionNetwork (3M-Net).In this 3M-Net,the MSMB block is designed to extract multiscale local features of the signals, and the MMG block is designedto enhance global information modeling of the model. Then, a hierarchical backbone that contains the two blocks is designed toextract multilevel features. Extensive experiments on the RML2016.10a and RML2018.01a datasets demonstrate that the 3M-Netmodel achieves superior recognition performance.

automatic modulation recognition (AMR)CNNdeep learning (DL)multiscaletransformer

Dan Jing、HuiLu Mo、Liang Han、HongFei Yin、Liangchao Li、Yan Zhang、Ming Li、Mian Pan、Liang Guo

展开 >

School of Telecommunications Engineering, Xidian University, Xi'an, Shaanxi, China

School of Optoelectronic Engineering, Xidian University,Xi'an, Shaanxi, China

Hangzhou Institute of Technology, Xidian University, Hangzhou, China

Guilin Changhai Development Co. Ltd, Guilin,China

School of Electronic and Information, Hangzhou Dianzi University, Hangzhou, Zhejiang, China

展开 >

2025

10.1002/dac.70109

International journal of communication systems

International journal of communication systems

ISSN：1099-1131
年,卷(期)：2025.38(9)
  • 31