TY - GEN
T1 - A Meta-Classified Hybrid Fusion Model for Interference-Resilient Modulation Recognition
AU - Tahir, M. Muneeb
AU - Latif, Arbab
AU - Younis, M. Shahzad
AU - Bin Rais, Rao Naveed
AU - Ammar, Khalid
N1 - Publisher Copyright:
© 2025 IEEE.
PY - 2025
Y1 - 2025
N2 - Automatic Modulation Classification (AMC) is crucial for modern wireless systems in defense, IoT, and cognitive radio applications. Conventional AMC methods often fail under channel impairments like AWGN, Rayleigh fading, and hardware imperfections. We propose a novel Hybrid Fusion DNN combining VGG, LSTM-CNN, GRU-CNN, and CLDNN architectures to extract robust spatiotemporal features across SNR conditions. Evaluated on diverse modulations (ASK, PSK, AM, FSK, APSK, QAM) under CFO, phase noise, and fading, our model achieves 89.13% overall accuracy, with 30-40% accuracy gains at low SNRs (e.g., 66.72% at -20dB ) and near-interference-free performance at higher SNRs (greater then 97% accuracy after -2dB SNR). This demonstrates hybrid deep learning's potential for reliable AMC in real-world wireless environments.
AB - Automatic Modulation Classification (AMC) is crucial for modern wireless systems in defense, IoT, and cognitive radio applications. Conventional AMC methods often fail under channel impairments like AWGN, Rayleigh fading, and hardware imperfections. We propose a novel Hybrid Fusion DNN combining VGG, LSTM-CNN, GRU-CNN, and CLDNN architectures to extract robust spatiotemporal features across SNR conditions. Evaluated on diverse modulations (ASK, PSK, AM, FSK, APSK, QAM) under CFO, phase noise, and fading, our model achieves 89.13% overall accuracy, with 30-40% accuracy gains at low SNRs (e.g., 66.72% at -20dB ) and near-interference-free performance at higher SNRs (greater then 97% accuracy after -2dB SNR). This demonstrates hybrid deep learning's potential for reliable AMC in real-world wireless environments.
KW - Additive White Gaussian Noise (AWGN)
KW - Automatic Modulation Classification (AMC)
KW - Carrier Frequency Offset (CFO)
KW - Hybrid Fusion Deep Neural Network
KW - Rayleigh Fading
KW - Signal-to-Noise Ratio (SNR)
KW - Wireless Communication
UR - https://www.scopus.com/pages/publications/105031452611
U2 - 10.1109/IICAIET67254.2025.11264918
DO - 10.1109/IICAIET67254.2025.11264918
M3 - Conference contribution
AN - SCOPUS:105031452611
T3 - 7th IEEE International Conference on Artificial Intelligence in Engineering and Technology, IICAIET 2025
SP - 334
EP - 339
BT - 7th IEEE International Conference on Artificial Intelligence in Engineering and Technology, IICAIET 2025
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 7th IEEE International Conference on Artificial Intelligence in Engineering and Technology, IICAIET 2025
Y2 - 26 August 2025 through 28 August 2025
ER -