International Research Journal of Engineering and Technology (IRJET)
e-ISSN: 2395-0056
Volume: 13 Issue: 02 | Feb 2026
p-ISSN: 2395-0072
www.irjet.net
Hybrid CNN-Transformer Architecture for Enhanced Signal Classification in Wireless Communications Nagla Elhaj Babiker1, Khalid Hamid Bilal2, Magdi B. M. Amien3 1Department. of Electronics Engineering (Communications and Control), University of Gezira, College of Engineering
and Technology.
2 Professor, Department of Electrical Engineering, Omdurman Islamic University, Omdurman, Sudan. 3Department of Electrical and Electronics Engineering, University of Khartoum, Khartoum, Sudan
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Signal classification in noisy environments is a common issue for wireless communication systems, particularly while working with modulation strategies. The complexity of signal data makes it difficult to use traditional methods that rely on simple convolutional neural network (CNN) architectures and conventional machine-learning models, particularly when operating in environments with high signal-to-noise ratios (SNRs). In addition, for real-world heterogeneous datasets, these methods often lack required generalizability. This work proposes a hybrid CNN transformer model to overcome these constraints. The proposed model performs better at classification when the SNR changes because it combines the sequential modelling capabilities of the transformer architecture with the feature extraction capabilities of the convolutional layers. It is trained using a dataset of 11 modulation schemes under varying SNRs to ensure the resilient performance in various noise settings of the model. The model performance is assessed using its accuracy and bit error rate (BER). The model outperforms standard methods by both accuracy and generalization for classification. The model performance under various settings was examined employing a confusion matrix visualization. Key Words: Automatic modulation recognition (AMR), Deep-learning neural networks
1.I NTRODUCTION Throughout the growth of new wireless communication technologies, the limited availability of radio spectra leads to a considerable drawback. This drawback is not a genuine deficiency of spectrum resources but rather ineffective regulatory frameworks that assign frequencies in a strict and unyielding fashion [1]. A wide range of organizations, such as the civilian, government, commercial, and military organizations, share the electromagnetic spectrum. Therefore, modern wireless communication environments require a dynamic spectrum allocation policy instead of the fixed policy that is commonly adopted today, leading to low spectrum utilization difficulties. Cognitive radio (CR) technology creation lets radios change and use unused frequency resources, it is called “spectrum holes” or “white spaces”. This work indicates the concept of dynamic access to the radio spectrum. By this, secondary users (SUs) can opportunistically access the frequency allocation to primary users (PUs) when they are not in use. This approach aims to enhance spectral efficiency by enabling transmissions on detected free bands, which addresses the issue of spectrum shortage [2]. Because spectrum sensing and detection are becoming more critical in spectrum monitoring, management, and secure communications such as 5G communications and beyond, IoT networks, and other services, CR has essential roles and capabilities in detecting active PU transmissions over the band. Deciding to transmit the sensing outcomes indicates that each PU transmitters is inactive at this band with a high probability, and CR becomes the promotion solution for scarcity and underutilization problems [3]. AMR is an essential item in digital communication systems and serves as a critical element for the effectiveness of CR. It enables dynamic radio resource management by using reconfigurable software-defined transceivers. These transceivers can reconfigure their transmission parameters by using the accessible communication resources in the electromagnetic environment to recognize modulation types of unknown signals without previous data [4]. By using AMR, CR is expected to accurately recognize or classify the modulation structure of the received signal rapidly, without any latency. This process helps to identify whether there is a PU in the channel, as all PUs use a single modulation technique for transmission over the frequency channel. This allows the corresponding demodulation process to be done on the receiving side [5], [6]. AMR is a transitional stage between signal modulation and demodulation. Because it is challenging to distinguish between numerous modulation schemes owing to several factors, including multipath fading, noise, center frequency offset, and signal
© 2026, IRJET
|
Impact Factor value: 8.315
|
ISO 9001:2008 Certified Journal
|
Page 348