New Researches in Electronic Defense Systems

New Researches in Electronic Defense Systems

Automatic Country Classification from Noisy Maritime V/UHF Communications for EW COMINT Using Transfer Learned Deep CNNs

Document Type : Original Article

Authors
seyed majid hasani azhdari
Fallah Mohamadzadeh
Mojtaba Mahmoodi
Department of Electrical Engineering, Imam Khomeini University of Marine Sciences, Nowshahr, Iran
10.22034/joeds.2026.570712.1112
Abstract
In the context of electronic warfare (EW) and communication intelligence (COMINT), automatic identification of V/UHF radio users by national origin is critical for maritime situational awareness, surveillance, and threat assessment in the North Indian Ocean theater. This study presents accent-based country classification of noisy V/UHF transmissions from 19 regional and extra-regional operators. A novel dataset was constructed by segmenting 7,126 audio samples collected from real-world operational recordings, with each segment corresponding to a specific operator within the 19 countries represented. A comprehensive preprocessing pipeline—including standardization, voice activity detection (VAD), noise reduction, fixed-length segmentation, and Log-Mel spectrogram extraction—was developed. Five ImageNet-pretrained deep convolutional neural networks (DCNNs) (VGG16, ResNet50, DenseNet121, EfficientNet-B2, ConvNeXtTiny) were systematically evaluated using a unified transfer-learning framework. ResNet50 achieved superior performance with 71.81% test accuracy, 70.36% macro F1-score, and 0.6874 MCC, demonstrating feasibility for real-time EW-COMINT integration despite severe noise and channel distortion. Results validate the effectiveness of transfer learning for automatic user nationality inference in operational maritime V/UHF environments.
Keywords
COMINT
V-UHF
Accent recognition
DCNN
New Researches in Electronic Defense Systems
Volume 5, Issue 14
Summer 2026
Pages 52-67