To overcome the limitations of traditional satellite navigation technologies in complex and signal-obstructed industrial environments, this paper presents MSM-Pnet, a novel semi-supervised FM-based positioning framework leveraging FM signals of opportunity. By integrating wavelet packet decomposition with a multi-scale Vision Transformer and a hybrid masking strategy that combines random and time–frequency-aware masking, MSM-Pnet introduces a masked autoencoder architecture capable of robust positioning with limited labeled data. Experimental results demonstrate that MSM-Pnet consistently outperforms conventional supervised learning methods in both indoor and outdoor environments, while also significantly reducing model complexity. These results highlight the method’s potential as a cost-effective and scalable solution for seamless indoor–outdoor positioning for Internet of Things systems.

With the rapid development of deep learning in wireless communication for signal detection, target recognition, and parameter estimation, the vulnerability of models to adversarial examples poses a critical challenge to system robustness and security. Existing adversarial attacks mainly focus on single-task models, limiting their applicability in multi-task scenarios. To address this issue, we propose a Multi-Teacher Distillation-guided Multi-Task Attack (MTDMA) framework. It integrates three high-performance teacher models—direct sequence spread spectrum (DSSS) detection, modulation recognition, and direction-of-arrival (DOA) estimation—to train a unified student model through hybrid knowledge distillation. By fusing soft targets from multiple teachers, the student jointly learns discriminative features of multiple tasks within a shared representation space. Furthermore, we design a Dual-Representation Momentum Attack (DRMA) that generates perturbations in both amplitude and IQ feature spaces with a momentum mechanism, improving attack transferability and stability. Experimental results demonstrate that MTDMA achieves high attack success rates on three task models and transfers effectively to heterogeneous architectures, outperforming FGSM, BIM, PGD, and MIM. This work extends adversarial attacks from single-task to multi-task settings, providing new insights into robustness evaluation and multi-task adversarial sample generation for wireless communication models.

Automatic modulation recognition is essential for large-scale wireless communications, but traditional methods often ignore unknown signals in open-set conditions, leading to their incorrect classification as known types and thereby compromising system reliability and communication security. To handle this challenge, a novel automatic modulation open set recognition (AMOSR) model based on deep metric learning with embedding enhancement is proposed in this article. First, for each known example, deep neural model is employed to separately extract the original in-phase and quadrature (IQ) signal and its instantaneous features, which are then fused to obtain embedding. Second, random erasing is employed to the original IQ signal and instantaneous features separately to obtain an augmented example, which has similar structure but different semantics with known example, and the embedding of this example is obtained by using first step. Then, the embedding space, in which tuplet loss and margin loss are combined with the embeddings of known and augmented examples, is trained to improve the overall performance of the model. Finally, after training, AMOSR is implemented using the class centers of known classes. Experiments on three automatic modulation datasets show that our model has better average performance than several mainstream methods in the field of computer vision.

Accurate and efficient positioning in complex environments remains a critical challenge where satellite-based systems (e.g., GNSS) suffer from signal attenuation and multipath interference. This paper proposes WK-Pnet, a lightweight positioning framework that utilizes frequency modulation (FM) signals and integrates Wavelet Packet Decomposition (WPD) with knowledge distillation. WK-Pnet first decomposes raw FM IQ signals using WPD to extract fine-grained multi-scale time-frequency features, preserving both spectral and phase information. These features are then fed into a deep neural network for location estimation. To reduce computational complexity, we employ a knowledge distillation strategy that transfers knowledge from a large-capacity ResNeXt-based teacher model—enhanced with a spatial attention mechanism—to a compact student network with significantly fewer parameters and FLOPs. The proposed method is validated on publicly available indoor and outdoor datasets, showing that WK-Pnet achieves comparable positioning accuracy to the teacher model while reducing FLOPs by 95.9%, model parameters by 99.3%, and inference latency by 90.5% on edge devices. Experimental comparisons also reveal that WPD outperforms STFT and EMD in positioning stability and accuracy, especially in outdoor scenarios. WK-Pnet demonstrates strong robustness, low-latency inference, and high accuracy, making it highly suitable for real-time, resource-constrained mobile and IoT applications.

Automatic Modulation Open Set Recognition (AMOSR) has practical significance in detecting unknown classes. However, a challenge arises when unknown samples closely resemble known samples, posing a formidable task for accurate detection. A novel AMOSR framework based on multi-view discriminators’ joint judgment is proposed to handle this challenge. Firstly, utilizing signal domain knowledge, multi-dimensional features are extracted through varied signal time-frequency transforms and encoders, baesd on which multiple discriminators are created. Secondly, Constrained Clustering Prototype Loss and Geodesic Contrastive Loss are introduced to pretrain these discriminators, providing more space for unknown signals. Then, collaborative learning is employed to further fine-tune the aforementioned discriminators, enhancing information sharing between modalities. Furthermore, a set of indicators is constructed, and multi-criteria fusion is performed using the TOPSIS algorithm to evaluate the discrimination capabilities of different classifiers in both closed-set and open-set scenarios. Furthermore, a decision tree is constructed to segregate test signals into known and unknown classes, in which discriminators with higher confidence levels are given precedence. Finally, TOPSIS hierarchical ensemble pruning algorithm that considers diversity and open-set recognition capabilities is adopted to reduce model complexity while maintaining original performance. Extensive experiments conducted on modulation datasets demonstrate the superiority of this framework over state-of-the-art AMOSR results.

Deep neural networks (DNNs) greatly enable the task of automatic modulation classification (AMC) by virtue of their powerful feature extraction capability. However, extensive research has shown that DNNs are highly vulnerable to adversarial attacks, which can lead them to confidently output incorrect results with high confidence scores. Existing adversarial attack methods often focus solely on temporal characteristics of signals while neglecting frequency domain information, resulting in adversarial examples with poor transferability and inadequate performance in the closed-box scenario. An adversarial attack method based on frequency domain feature enhanced and integral gradient (FEIG) for AMC task is proposed in this paper. The approach utilizes techniques such as translation interpolation and Inverse Fast Fourier Transform to enhance the frequency domain information of original examples, thereby constructing enhanced baseline examples. Subsequently, these generated enhanced baseline examples are used as new inputs for gradient integration to obtain adversarial examples. Compared to traditional methods, the generated adversarial examples exhibit stronger transferability. Furthermore, in order to improve the defense performance of the model, an enhanced hybrid adversarial training (EH-AT) framework is proposed in this paper. The original clean example and the adversarial example generated by the proposed attack method are trained with joint loss constraints, which greatly enhances the robustness of the model. Experimental results demonstrate the effectiveness of the FEIG attack method and the EH-AT framework.

Spectrum is the fundamental medium for transmitting information services, including communication, navigation, and detection. Spectrum anomalies can lead to substantial economic losses and even endanger life safety. Anomaly detection constitutes a critical component of spectrum risk management. Through spectrum anomaly detection (SAD), anomalous spectrum usage behaviors, such as malicious user activities, can be identified. Given the significant limitations of current SAD algorithms in terms of accuracy and localization capabilities, this article proposes an approach for detecting spectral anomalies that utilizes knowledge distillation and memory-enhanced autoencoders (AEs). First, the pretrained network with robust feature extraction capabilities is distilled into the teacher network. Subsequently, both an AE and a memory-enhanced AE with an identical structure are trained to predict the teacher network’s normalized outputs on a spectrum devoid of anomalies. Finally, in the case of an anomalous spectrum, difference exist between the normalized outputs of the teacher network and the outputs of different student networks, as well as among the outputs of different student networks, which facilitates the process of anomaly detection. The outcomes of experiments reveal that the proposed algorithm is more effective on both synthetic spectral data sets and real IQ signals, demonstrating its proficiency in accurately detecting and locating anomalies.

Due to the openness of wireless channels, wireless communication is vulnerable to be eavesdropped, which results in confidential information leakage. Physical Layer security (PLS) technology provides a new way to solve this hidden danger of Internet of Things system. However, traditional PLS methods are often restricted by limited communication resources and unknown instantaneous channel state information of eavesdroppers, which makes it challenging to strike a balance between security and reliability in the communication system. Therefore, an adversarial defense embedded waveform design (ADEWD) method for physical layer reliable communication (PLRC) is proposed in this paper. Firstly, we use generative adversarial networks to generate amplitude controllable adversarial perturbation, and then superimpose it with original communication signal to form an adversarial signal. At the same time, we also design a demodulation network based on the modulation type of legitimate users to constrain the amplitude of the generated perturbations, to reduce the bit error rate (BER) loss after demodulation of the adversarial signal. With this waveform design, the adversarial signal not only enables reliable communication between legitimate users, but also utilizes embedded defense traps to prevent eavesdroppers from recognizing legitimate users. The experimental results demonstrate that our ADEWD method for PLRC has stronger defense capability and lower BER in both white-box and black-box scenarios, which reflects the defense robustness and communication reliability of the proposed waveform design method.

Cognitive radio-based Internet of Things (CR-IoTs) provide an efficient spectrum management for IoT networks with massive wireless access and data transmission needs. As one of the key technologies of CR-IoT, spectrum sensing is of great research significance. Motivated by the recent boom on applications of deep learning in wireless communications networks and IoT, several spectrum sensing methods based on deep learning have emerged. However these algorithms train the sensing models with the extracted features of received signals and require a retraining of sensing models when the number of sensing antennas changes. Thus, we develop multiple-antenna spectrum sensing methods based on convolutional neural networks (MASSnet) using the in-phase (I) and quadrature (Q) components of the signals as the input. The three schemes of MASSnet also provide the flexibility to choose between retraining the sensing models or using the obtained models for different sensing antenna configurations. Experiment results demonstrate the superior performance of the proposed methods over existing deep learning-based spectrum sensing methods in terms of probability of detection especially in very low signal-to-noise ratio (SNR) condition. Furthermore, the proposed methods have good generalization ability to new noise distribution, new fading channel, different frequency offsets, and detecting signals with a new modulation even without retraining.

Frequency Modulation (FM) broadcast signals, regarded as opportunistic signals, hold significant potential for indoor and outdoor positioning applications. The existing FM-based positioning methods primarily rely on Received Signal Strength (RSS) for positioning, the accuracy of which needs improvement. In this paper, we introduce FM-Pnet, an end-to-end FM-based positioning method that leverages deep learning. This method utilizes the time-frequency representation of FM signals as network input, enabling automatically learning of deep features for positioning. We also propose two strategies, noise injection and enriching training samples, to enhance the model’s generalization performance over long time spans. We construct datasets for both indoor and outdoor scenarios and conduct extensive experiments to validate the performance of our proposed method. Experimental results demonstrate that FM-Pnet significantly outperforms traditional RSS-based positioning methods in terms of both positioning accuracy and stability.

Automatic modulation recognition (AMR) of radio signal is an important research topic in the area of non-cooperative communication and cognitive radio. Recently deep learning (DL) techniques enable significant progress in AMR. However, the techniques of adversarial machine learning cause the threats of adversarial attacks in DL-based AMR. In this paper, we aim to make AMR model robust, accurate and lightweight, thus propose a multi-distillation mechanism for robust training of DL-based AMR models, namely Adversarial Multi-Distillation (AMD). In the framework of AMD, by knowledge distillation, two powerful teacher models transfer the learned classification knowledge and defense knowledge, respectively, to the student model to form robust training. Our experiments with public dataset RML2016.10a show that the proposed method can significantly improve the defense of AMR models to against adversarial perturbations and keep relatively high classification accuracy, which enables robust decision making with lightweight models under adversarial attacks.

A wireless communications system usually consists of a transmitter which transmits the information and a receiver which recovers the original information from the received distorted signal. Deep learning (DL) has been used to improve the performance of the receiver in complicated channel environments and state-of-the-art (SOTA) performance has been achieved. However, its robustness has not been investigated. In order to evaluate the robustness of DL-based information recovery models under adversarial circumstances, we investigate adversarial attacks on the SOTA DL-based information recovery model, i.e., DeepReceiver. We formulate the problem as an optimization problem with power and peak-to-average power ratio (PAPR) constraints. We design different adversarial attack methods according to the adversary’s knowledge of DeepReceiver’s model and/or testing samples. Extensive experiments show that the DeepReceiver is vulnerable to the designed attack methods in all of the considered scenarios. Even in the scenario of both model and test sample restricted, the adversary can attack the DeepReceiver and increase its bit error rate (BER) above 10%. It can also be found that the DeepReceiver is vulnerable to adversarial perturbations even with very low power and limited PAPR. These results suggest that defense measures should be taken to enhance the robustness of DeepReceiver.

Deep learning has been widely used in automatic modulation classification (AMC) recently. Most of deep learning-based AMC uses a single network model to deal with radio signals with a single input format. In this paper, we propose a hybrid heterogeneous modulation classification architecture named DeepSIG, which integrates Recurrent Neural Network (RNN), Convolutional Neural Network (CNN) and Graph Neural Network (GNN) models in a single framework to process radio signals with heterogeneous input formats, i.e., in-phase (I) and quadrature (Q) sequences, images mapped from IQ signals and graphs converted from IQ signals, to extract and integrate the features from different perspectives. A fusion training mechanism is presented to train DeepSIG. We use three different radio signal datasets for simulations. Results show that our proposed DeepSIG performs the best in terms of classification accuracy compared with the three methods with single input, i.e., sequence, image or graph. The performance gain is larger in few-shot scenarios.

In this paper, we focus on designing an unified deep neural demodulation network for recovering multiple QAM signals, which can adapt to the adaptive QAM modulation signal. We specifically introduce the convolution block, identity block, self-attention block to extract the input complex signal feature, such that the unified demodulation network can jointly decide the modulation type and symbol constellation index. In the meanwhile, our proposed demodulation network can compensate the time and frequency offset induced by the channel and receiver oscillator. We evaluate the symbol error rate (SER) of our proposed demodulation network for BPSK, QPSK, and 8QAM signals in AWGN, Nakagami and Rician channels under different constellation mapping schemes. Simulation results show that, even when the constellation graph overlaps with each other for BPSK, QPSK, and 8QAM signals, for AWGN channel, the SER of our proposed unified demodulation network can approach the theoretical bound. While for the Nakagami and Rician channels, our proposed demodulation network can also provide a competitive SER performance.

Automatic modulation classification (AMC) is an important technology for the monitoring, management, and control of communication systems. In recent years, machine learning approaches are becoming popular to improve the effectiveness of AMC for radio signals. However, the automatic modulation open-set recognition (AMOSR) scheme that aims to identify the known modulation types and recognize the unknown modulation signals is not well studied. Therefore, in this paper, we propose a novel multi-modal marginal prototype framework for radio frequency (RF) signals (MMPRF) to improve AMOSR performance. First, MMPRF addresses the problem of simultaneous recognition of closed and open sets by partitioning the feature space in the way of one versus other and marginal restrictions. Second, we exploit the wireless signal domain knowledge to extract a series of signal-related features to enhance the AMOSR capability. In addition, we propose a GAN-based unknown sample generation strategy to allow the model to understand the unknown world. Finally, we conduct extensive experiments on several publicly available radio modulation data, and experimental results show that our proposed MMPRF outperforms the state-of-the-art AMOSR methods.

Although deep learning methods have made significant advancements across various domains, recent research has shown that carefully crafted adversarial samples can lead to a significant degradation in the performance of deep learning models. Such adversarial examples raise concerns about the reliability and safety of deep learning-based models. Currently, there is a lack of research on the robustness of deep learning based DOA methods against adversarial samples. This letter aims to fill this research gap by leveraging the differentiability of the transformation process from the original signal to the covariance matrix. By utilizing this differentiability, the robustness of the DOA estimation model, which takes the covariance matrix as input, is investigated. Four different white-box attack methods are considered to generate adversarial samples to evaluate the resilience of the model. The experimental results demonstrate that all four methods employed significantly increase the estimation error of the DOA estimation model, posing a serious threat to the model's security.

In cognitive radio networks, providing accurate recognition of the primary user’s signal is great important for designing the spectrum access strategies. Source number estimation has served as a key fundamental technique to facilitate the signal recognition in the mixed received signals scenario. In this paper, we propose a deep learning based source number estimation method under the single-channel conditions. The architecture of the network is first designed. Then, the received complex signals are reconstructed as in-phase and quadrature (IQ) data in order to adapt to the convolutional neural network for extracting the deep features. Moreover, the cost function that is used to train the proposed network is properly designed by exploring the maximum likelihood function. Supervised training is performed to generate a classifier that can identify the number of sources in the mixtures. The proposed deep leaning-based source number estimation method is tested by experiments on simulation signals and actual signals. The results revealed the effectiveness of the proposed method in solving the problem of single-channel source number estimation compared to the conventional methods.

Multiple-input–multiple-output (MIMO) technology is one of the most widely used communication technologies. However, with the increasing number of antennas, the complexity of the MIMO wireless communication receiver becomes higher and higher. On the other hand, the complex communication channels also raise up a great challenge to the reliability of the communication receiver system. With the rapid development and wide application of deep learning, it has been applied in the field of communication to solve some problems that are difficult to solve by the traditional methods, and thereby, improves the reliability of communication systems. Inspired by this idea, this article reviewed the signal processing process of the MIMO receiver system from the perspective of system reliability. Based on deep learning, the signal processing modules of the receiver system are jointly optimized, which changes the information recovery process of the traditional receiver and proposes the intelligent receiver for MIMO communication. In order to verify the system reliability of the intelligent receiver, this article analyzes it from the aspects of antenna numbers and channel conditions. The influence of different implementation methods of the intelligent receiver on the system reliability is also analyzed. Simulation results show that the proposed intelligent receiver for the MIMO wireless communication can recover information with a lower bit error rate and higher reliability compared with the traditional receiver under different conditions and antenna configurations.

Signal modulation recognition plays a critical role in many fields to identify the modulation type of wireless signals. Since the deep learning based models have achieved great success in classification tasks, more deep neural networks are proposed for signal modulation recognition. In this paper, we explore the use of different deep neural networks in both macro network architecture level and micro cell size and layer level to compare and understand their effect on classification performance. We also bring up a feature explainable deep neural network by visualizing the critical features in the deep neural models. We visually show and compare the commonality and differences of hidden layer characteristics extracted by different network structure to explain and analyze the reason why some models can achieve better classification results than the others. We believe it is an effective way to explain how deep neural model based signal classification work. Thus the explanation will help users establish appropriate understand and trust in predictions from deep modulation recognition networks.

Cognitive radio has been regarded as a promising technology to improve spectrum utilization significantly. In this letter, spectrum allocation model is presented firstly, and then spectrum allocation methods based on genetic algorithm (GA), quantum genetic algorithm (QGA), and particle swarm optimization (PSO), are proposed. To decrease the search space we propose a mapping process between the channel assignment matrix and the chromosome of GA, QGA, and the position of the particle of PSO, respectively, based on the characteristics of the channel availability matrix and the interference constraints. Results show that our proposed methods greatly outperform the commonly used color sensitive graph coloring algorithm.

Abstract One of the basic capabilities of cognitive radio is to adapt the radio parameters according to the changing environment and user needs. This paper proposes a new adaptation method which uses particle swarm optimization (PSO) to optimize cognitive radio parameters given a set of objectives. The procedure of the proposed method is presented and multicarrier system is used for simulation analysis. Experimental results show that the proposed method performs far better than genetic algorithm (GA)‐based adaptation method in terms of convergence speed, converged fitness values, and stability. The proposed method can also provide the tradeoffs of the objective functions, and the resulting parameter configuration is consistent with the weights of the objective functions. Copyright © 2008 John Wiley & Sons, Ltd.