Deep learning (DL)-based automatic modulation classification (AMC) has made breakthroughs and is generally used for signal detection and recognition in wireless communication systems, unmanned aircraft vehicle (UAV) systems, and other fields. However, high storage and computational demands limit its use in resource-constrained UAV systems. This article presents an AMC method featuring a streamlined design with lower computational needs, using the ultralight convolutional neural network (ULCNN). This innovative model combines data augmentation, complex-valued convolution, separable convolution, channel attention, and shuffling techniques for enhanced performance. The proposed ULCNN model balances efficiency and accuracy, with simulations showing it achieves 62.47% accuracy on the RML2016.10a data set using only 9751 parameters. Furthermore, we evaluated the actual speed of ULCNN on a Raspberry Pi, an edge platform with roughly equivalent computing power to a conventional UAV, achieving an inference speed of only 0.775 ms per sample. This high performance, coupled with a significantly smaller model size, underscores the potential of ULCNN for integration into resource-constrained UAV systems, thereby enabling rapid and efficient data processing.

Automatic modulation classification (AMC) is a signal processing technology used to identify the modulation type of unknown signals without prior information such as modulation parameters for drone communications. In recent years, deep learning (DL) has been widely used in AMC methods due to its powerful feature extraction ability. The significant performance of DL-based AMC methods is highly dependent on large amount of data. However, with the increasingly complex signal environment and the emergence of new signals, several recognition tasks have difficulty obtaining sufficient high-quality signals. To address this problem, we propose an AMC method based on a deep residual neural network with masked modeling (DRMM). Specifically, masked modeling is adopted to improve the performance of a deep neural network with limited signal samples. Both complex-valued and real-valued residual neural networks (ResNet) play an important role in extracting signal features for identification. Several typical experiments are conducted to evaluate our proposed DRMM-based AMC method on the RadioML 2016.10A dataset and a simulated dataset, and comparison experiments with existing AMC methods are also conducted. The simulation results illustrate that our proposed DRMM-based AMC method achieves better performance in the case of limited signal samples with low signal-to-noise ratio (SNR) than other existing methods.

Radio spectrum prediction is an important task for dynamic spectrum management and spectrum congestion mitigation. However, due to the complexity and spatiotemporal variability of spectrum data, spectrum prediction has always been a challenging problem. To improve prediction accuracy, this paper proposes a multi-band spectrum prediction method based on attention graph convolutional recurrent neural networks (A-GCRNN), which applies temporal correlation and frequency band correlation to spectrum prediction tasks. This method represents the spectrum data of multiple bands as a graph, with each band corresponding to a node in the graph, and utilizes graph convolutional networks (GCN) to learn the correlations between different bands. Then, a gated recurrent unit (GRU) network is employed to capture temporal correlations for each band, thereby fusing the acquired feature information to predict the future spectrum of each frequency band. In addition, an attention network is used to weight the output hidden states and further enhance the model's convergence rate. We conduct experiments on a real-world spectrum dataset and compare our method with other spectrum prediction methods. The results demonstrate that our approach achieves better performance in multi-band spectrum prediction tasks. Our pytorch implementation of the A-GCRNN is available at https://github.com/TLTLHILL/A-GCRNN-for-spectrum-prediction .

Specific emitter identification (SEI) based on unavoidable hardware impairments of transmitters has emerged as a potential technology for physical layer authentication of Internet of Things (IoT) devices. The integration with deep learning has significantly accelerated the advancement of SEI in recent years, showcasing its strong application prospects. However, most SEI works are designed for close-set scenarios, leading to misclassification of unknown transmitters outside the close-set, which poses a serious security risk to the authentication system. To address this issue, a SEI framework with anomaly detection capabilities is essential. In this paper, we propose a novel open-set SEI framework that leverages the Auxiliary Classifier Generative Adversarial Network (ACGAN) to generate outlier samples, introduces OpenMax to obtain calibrated activation vectors (AV), and applies a threshold on the category confidence. Through the proposed triple anomaly detection condition, our SEI framework is endowed with enhanced capabilities for identifying anomalous transmitters. We evaluate the framework using 10 known and 6 unknown WiFi transmitters and achieve a high F1-Score of 0.911 and a low False Positive Rate (FPR) of 0.063 in an anomaly detection task.

Specific emitter identification (SEI) is pivotal for ensuring the security of the Internet of Things (IoT). Traditional deep learning-based SEI techniques often falter in real-world applications, particularly when distinguishing between legitimate and rogue devices amid noisy conditions and low-signal-to-noise ratios (SNRs). To surmount these challenges, we propose a novel open-set SEI (OS-SEI) strategy that utilizes a metric-enhanced denoising autoencoder (MeDAE) architecture. This advanced framework incorporates a deep residual shrinkage network, significantly augmenting the denoising autoencoder’s capability, thereby bolstering its resilience against noisy environments. Further, the integration of discriminative metrics, such as center loss, markedly enhances feature discrimination, resulting in heightened accuracy of device identification. Our comprehensive experimental assessments, conducted on an automatic dependent surveillance-broadcast (ADS-B) data set, underscore the superiority of our proposed OS-SEI method over existing models. The findings confirm our approach’s enhanced robustness to noise and its superior accuracy in device identification within open-set scenarios.

Specific emitter identification (SEI) has emerged as a notable device authentication technology, distinguishing various emitters through the unique radio frequency fingerprint (RFF) inherent in wireless devices. Traditional SEI methods, often hindered by time-consuming manual feature extraction, struggle with complex encrypted signals. The advent of deep learning, with its robust feature extraction capabilities, has significantly advanced SEI, yet it typically demands extensive radio frequency signal samples and falters with limited (i.e., few-shot) samples. Our proposed few-shot SEI (FS-SEI) approach, integrating neural architecture search (NAS) and advanced deep transfer learning (DTL), adeptly identifies few-shot long-range (LoRa) devices. This method begins with NAS to autonomously tailor optimal network architectures for SEI tasks, followed by pretraining on extensive auxiliary data sets to extract general RFF features of LoRa devices. Transfer learning then fine-tunes these features for distinctiveness with compact intraclass distances. By only utilizing few-shot LoRa data for final parameter adjustments, the classifier rapidly assimilates new categories. Simulations confirm our FS-SEI method’s superior accuracy over classical approaches, with visualized feature analysis underscoring its distinguishing and generalizing prowess.