This paper presents an in-depth study of deep learning-based radio signal classification for radio communications. The authors compare the performance of deep learning (DL) with traditional methods such as higher order moments and boosted gradient tree classification under various configurations and channel impairments. They evaluate both simulated and over-the-air (OTA) measurements using software radios to assess classification performance and training strategies. The study highlights the effectiveness of deep learning in accurately classifying radio signals, particularly in challenging environments. The paper discusses the use of deep learning for radio signal classification, including convolutional neural networks (CNNs), residual neural networks (ResNets), and transfer learning. It presents a comprehensive dataset of 24 different modulation types, including both analog and digital signals, and evaluates the performance of various classification models under different conditions. The results show that deep learning models, particularly ResNets, outperform traditional methods in terms of classification accuracy, especially under high signal-to-noise ratio (SNR) conditions. The study also explores the impact of various channel impairments, such as carrier frequency offset (CFO), symbol rate offset (SRO), and multi-path fading, on the performance of classification models. It demonstrates that deep learning models, particularly ResNets, are robust to these impairments and can achieve high classification accuracy even in challenging environments. The paper also discusses the importance of sufficient training data and the potential of transfer learning in improving the performance of deep learning models on real-world datasets. The authors conclude that deep learning offers significant improvements in radio signal classification, particularly for short-time observations and in environments with high signal complexity. They emphasize the need for further research to optimize synthetic impairment distributions and improve the generalization of deep learning models to real-world conditions. The study provides valuable insights into the application of deep learning in radio signal classification and highlights the potential of this approach for future communication systems.This paper presents an in-depth study of deep learning-based radio signal classification for radio communications. The authors compare the performance of deep learning (DL) with traditional methods such as higher order moments and boosted gradient tree classification under various configurations and channel impairments. They evaluate both simulated and over-the-air (OTA) measurements using software radios to assess classification performance and training strategies. The study highlights the effectiveness of deep learning in accurately classifying radio signals, particularly in challenging environments. The paper discusses the use of deep learning for radio signal classification, including convolutional neural networks (CNNs), residual neural networks (ResNets), and transfer learning. It presents a comprehensive dataset of 24 different modulation types, including both analog and digital signals, and evaluates the performance of various classification models under different conditions. The results show that deep learning models, particularly ResNets, outperform traditional methods in terms of classification accuracy, especially under high signal-to-noise ratio (SNR) conditions. The study also explores the impact of various channel impairments, such as carrier frequency offset (CFO), symbol rate offset (SRO), and multi-path fading, on the performance of classification models. It demonstrates that deep learning models, particularly ResNets, are robust to these impairments and can achieve high classification accuracy even in challenging environments. The paper also discusses the importance of sufficient training data and the potential of transfer learning in improving the performance of deep learning models on real-world datasets. The authors conclude that deep learning offers significant improvements in radio signal classification, particularly for short-time observations and in environments with high signal complexity. They emphasize the need for further research to optimize synthetic impairment distributions and improve the generalization of deep learning models to real-world conditions. The study provides valuable insights into the application of deep learning in radio signal classification and highlights the potential of this approach for future communication systems.