TY - JOUR
T1 - Meta Supervised Contrastive Learning for Few-Shot Open-Set Modulation Classification With Signal Constellation
AU - Zhao, Jikui
AU - Wang, Huaxia
AU - Peng, Shengliang
AU - Yao, Yu Dong
N1 - Publisher Copyright:
© 1997-2012 IEEE.
PY - 2024/4/1
Y1 - 2024/4/1
N2 - In this work, we introduce a method for few-shot open-set modulation classification utilizing signal constellation diagrams, based on a Meta Supervised Contrastive Learning (MSCL) algorithm. MSCL combines the strengths of supervised contrastive learning and meta-learning to effectively amplify inter-class distinctions and reinforce intra-class compactness. The experimental results demonstrate that MSCL exhibits superior performance in both few-shot and open-set Automatic Modulation Classification (AMC) recognition. Code available at: https://github.com/jikuizhao/MSCL
AB - In this work, we introduce a method for few-shot open-set modulation classification utilizing signal constellation diagrams, based on a Meta Supervised Contrastive Learning (MSCL) algorithm. MSCL combines the strengths of supervised contrastive learning and meta-learning to effectively amplify inter-class distinctions and reinforce intra-class compactness. The experimental results demonstrate that MSCL exhibits superior performance in both few-shot and open-set Automatic Modulation Classification (AMC) recognition. Code available at: https://github.com/jikuizhao/MSCL
KW - Few-shot learning
KW - meta-learning
KW - modulation classification
KW - open-set classification
KW - signal constellation
UR - http://www.scopus.com/inward/record.url?scp=85184795969&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85184795969&partnerID=8YFLogxK
U2 - 10.1109/LCOMM.2024.3363630
DO - 10.1109/LCOMM.2024.3363630
M3 - Article
AN - SCOPUS:85184795969
SN - 1089-7798
VL - 28
SP - 837
EP - 841
JO - IEEE Communications Letters
JF - IEEE Communications Letters
IS - 4
ER -