Unmanned aerial vehicle (UAV)-assisted non-orthogonal multiple access (NOMA) satellite Internet of Things (IoT) networks face severe physical-layer security challenges in the presence of eavesdroppers. To address this issue, this paper proposes a secure graph neural network (Sec-GNN) based multi-dimensional anti-eavesdropping optimization method. The approach models the node-spatial relationships among the UAV, legitimate users, and eavesdroppers as a graph structure. By leveraging the message-passing mechanism of graph neural networks—sequently performing message generation, message aggregation, and node update—it dynamically integrates network topology information and node interaction features. This enables end-to-end joint optimization of the UAV’s three-dimensional position, beamforming vectors, and multi-user power allocation strategies. The method does not rely on explicit channel state information and directly generates near-optimal resource allocation schemes based on node location information and observable signal features. Experimental results demonstrate the superiority of Sec-GNN across various scenarios, and ablation studies confirm that partial optimization leads to significant performance degradation, thereby verifying the necessity of multi-dimensional joint design. The proposed framework provides an efficient and scalable solution for secure resource management in dynamic space-air-ground integrated networks.

The proliferation of model-sharing platforms has intensified the need for data-free backdoor detection, as deployed models are often accessed without accompanying clean validation data. This constraint renders traditional, data-dependent detection methods ineffective. However, existing strategies to evade data-free detection are inadequate; they frequently fail to circumvent the multi-faceted discriminative criteria of modern detectors that analyze model output behavior, and they often compromise the backdoor model's primary task performance, thus failing to balance attack stealth with functionality. To evade these detections, this paper introduces a Stealthy Backdoor Attack (SBdA) based on label smoothing. Our method dynamically adjusts the training labels for backdoor samples by leveraging the feature similarity between each class and the attacker's target class. This optimization shapes the backdoored model's output distributions to closely mimic those of a benign model, thereby evading detection mechanisms that rely on outlier characterization and posterior distribution analysis. Extensive experiments demonstrate that SBdA maintains a high attack success rate (exceeding 91% under all tested conditions, with 75% of models surpassing 96%) while significantly reducing its detectability. On CIFAR-10, the average outlier score for our models was 0.554-merely 0.050 higher than benign models-compared to a 25.517 deviation for conventional attacks. On GTSRB, SBdA reduced the outlier score gap by 82.86% compared to the average-label technique. Furthermore, by carefully calibrating the posterior distribution, SBdA effectively avoids detection by posterior matrix-based methods across all four tested datasets.

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.

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.

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.

Abstract Blind watermark extraction in discrete cosine transform (DCT) domain has a wide application prospect as well as a challenging subject. The imperceptibility of watermark signal makes watermark extraction a weak signal reception issue in essence. For DCT coefficients of host image generally disobey Gaussian distribution, at which the performance of linear correlated reception is no longer optimal. Aiming at this, a novel blind watermark extraction scheme combining the uncorrelated reception with adaptive bistable stochastic resonance (ABSR) technique is proposed. First, by block DCT transformation for host image, an additive watermark embedding algorithm is introduced, in which the watermarked image can be converted to one dimensional time domain weak signal (binary watermark image) reception under additive Laplacian noise (selected DCT coefficients). On this basis, through the key technology research on quantitatively cooperative resonance relationship under Laplacian noise, the ABSR system can be implemented by bistable system parameters self‐adaptive adjustment, in which the ABSR system output signal will be enhanced rather than be weakened by random noise. Finally, the ABSR‐based watermark extraction scheme is investigated, and both the visual effect, bit error ratio performance and robustness of proposed scheme are testified to be superior to that of traditional uncorrelated extraction.

In this paper, we propose a covert wireless uplink transmission strategy in space-air-ground integrated vehicular networks, where the source vehicle transmits its own message over the channel that being used by the host communication system, to avoid being detected by the warden. It is obvious that the data transmission efficiency of the covert communication system is limited due to the co-channel interference. To improve the data transmission efficiency, we consider that the covert communication system adopts improper Gaussian signaling (IGS). We formulate a joint transmit power and IGS factor optimization problem to minimize the outage probability of the covert communication system. The minimum error detection probability of the warden is first analyzed with noise uncertainty, which is used to measure the system covertness. Under the constraints of the quality of service (QoS) of host communication system and the covertness requirement, the optimal transmit power is first derived with proper Gaussian signaling (PGS) scheme. Then, with the approximate outage probability derived under IGS scheme, the optimization problem is solved by jointly designing the transmit power and IGS factor. Finally, we provide extensive numerical results to validate the proposed covert transmission strategy, and demonstrate that the IGS scheme is beneficial in improving the data transmission efficiency in terms of outage probability compared to PGS scheme.

Advances in adversarial attack and defense technologies will enhance the reliability of deep learning (DL) systems spirally. Most existing adversarial attack methods make overly ideal assumptions, which creates the illusion that the DL system can be attacked simply and has restricted the further improvement on DL systems. To perform practical adversarial attacks, a detection tolerant black-box adversarial-attack (DTBA) method against DL-based automatic modulation classification (AMC) is presented in this article. In the DTBA method, the local DL model as a substitution of the remote target DL model is trained first. The training dataset is generated by an attacker, labeled by the target model, and augmented by Jacobian transformation. Then, the conventional gradient attack method is utilized to generate adversarial attack examples toward the local DL model. Moreover, before launching attack to the target model, the local model estimates the misclassification probability of the perturbed examples in advance and deletes those invalid adversarial examples. Compared with related attack methods of different criteria on public datasets, the DTBA method can reduce the attack cost while increasing the rate of successful attack. Adversarial attack transferability of the proposed method on the target model has increased by more than 20%. The DTBA method will be suitable for launching flexible and effective black-box adversarial attacks against DL-based AMC systems.

In this paper, we propose an improper Gaussian signaling (IGS) empowered covert communication strategy in unmanned air vehicle (UAV) assisted communication systems. The ground user (Alice) covertly transmits confidential messages to the UAV amounted based station (Bob) by superimposing over an overt channel, which is licensed to an existing communication system. To alleviate the inner-system interference caused by the superimposed waveforms, we design the IGS as the waveform of the covert communication system and the proper Gaussian signaling (PGS) as the waveform of the existing communication system. We derive closed-form expressions for the outage prob-ability of the overt and covert transmission links, respectively. Moreover, we formulate a joint transmit power and IGS factor optimization problem to maximize the outage performance of the covert communication system constraining the covertness requirement and quality of service of the existing communication system. The optimal solution is derived by leveraging the mono-tonic properties of objective function and constraints. Finally, simulation results are provided to verify the effectiveness of the proposed covert communication strategy.

With unprecedented progress on Internet of Things (IoT), spectrum scarcity becomes even severe with the explosive growth of wireless smart devices. To deal with spectrum scarcity issues, cognitive radio (CR)-enabled IoT has been emerged as a promising solution, which allows IoT devices reusing the underutilized spectrum bands. In this article, we investigate spectrum sensing in CR-IoT, in which full-duplex CR-IoT node can perform spectrum sensing and data transmission concurrently for reducing sensing delay. Two sensing methods are proposed based on multiple high-order cumulants for excavating rich information of the non-Gaussian transmitted signals. Specifically, for the scenarios with a single sensing antenna, we first propose a multiple high-order cumulants-based sensing method (MCS) derived from the likelihood ratio test, which is assumed to be near optimum. The test statistics are derived, respectively, in two cases, i.e., the case only performing sensing and the case performing sensing and transmission simultaneously. Interestingly, the derived two test statistics have same expression, while the corresponding sensing thresholds are different from each other. For the scenarios with multiple sensing antennas, we propose a multiantenna-assisted multiple high-order cumulants-based sensing method (MMCS), which can provide a tradeoff between the computational complexity and sensing performance. We conduct the hypothesis test with Hotelling's T2-statistic and derive the corresponding sensing threshold. Theoretical performance evaluated by detection probability and computational complexity of the proposed methods are analyzed. Additionally, extensive simulations are provided, which show both the proposed methods can counter the adverse effects of noise uncertainty, and MCS has superiority over MMCS in terms of sensing accuracy.

Covert wireless communications aim to hide the existence of transmission behavior from watchful adversaries to enhance security. In this paper, we propose a spectrum mask based covert communication strategy in Internet of Things (IoT) networks, where the overt channels are leveraged to enhance the covertness. Specifically, the legitimate IoT transmitter superimposes its own message on the overt channel to avoid being detected by the warden. We assume that proper Gaussian signaling (PGS) is adopted at the overt channel, and improper Gaussian signaling (IGS) is adopted at the legitimate transmitter to improve the covert transmission performance. To maximize the covert rate of the legitimate IoT system, a joint transmit power and IGS factor optimization problem is formulated under the constraints of covertness requirement. The metric of minimum error detection probability, that represents the worst-case for the legitimate transmitter, is utilized to measure the covertness. By exploiting the piece-wise monotonic properties of the objective function and the constraints, we derive the optimal transmit power and IGS factor pairs in both the IGS and PGS schemes. Finally, extensive numerical results are presented to demonstrate that the IGS scheme can improve the covert rate compared to the PGS scheme under a given covertness constraint.

Massive malicious jamming devices and advanced jamming techniques have been emerged along with the development of wireless communication networks. In order to effectively eliminate the harmful interference, a new scheme known as the cognitive frequency hopping (CHF) is proposed recently, which can evaluate the occupancy of frequency hopping slots and adjust the parameters dynamically according to the spectrum sensing results. Although the existing literature only shows that CFH systems can achieve reliable data transmission, the factors affecting the reliability are rarely analyzed. Therefore, we analyze the reliability performance of the CHF systems in this article, and define a new metric named anti-interference tolerance to measure the reliability performance of the CHF systems. Moreover, we derive the analytic expression of anti-interference tolerance by analyzing the effect of false alarm probability, missed detection probability, and communication link convergence delay. Simulation results validate the effectiveness of our analyses for measuring the reliability capacity of the CHF systems. To satisfy the demands of different communication scenarios, the CFH systems can adjust relevant parameters in the light of our theoretical derivation.

In this letter, we consider the varying detection environments to address the problem of detecting small targets within sea clutter. We first extract three simple yet practically discriminative features from the returned signals in the time and frequency domains and then fuse them into a 3-D feature space. Based on the constructed space, we then adopt and elegantly modify the support vector machine to design a learning-based detector that enfolds the false alarm rate (FAR). Most importantly, our proposed detector can flexibly control the FAR by simply adjusting two introduced parameters, which facilitates to regulate detector's sensitivity to the outliers incurred by the sea spikes and to fairly evaluate the performance of different detection algorithms. Experimental results demonstrate that our proposed detector significantly improves the detection probability over several existing classical detectors in both low signal to clutter ratio (up to 58%) and low FAR (up to 40%) cases.

This study presents a new machine learning (support vector machine (SVM))‐based cooperative spectrum sensing (CSS) model, which utilises the methods of user grouping, to reduce cooperation overhead and effectively improve detection performance. Cognitive radio users were properly grouped before the cooperative sensing process using energy data samples and an SVM model. The resulting user group which participates in cooperative sensing procedures is safe, less redundant, or the optimised user group. Three grouping algorithms are presented in this study. The first grouping algorithm divides normal and abnormal users (malicious and severely fading users) into two groups. The second grouping algorithm distinguishes redundant and non‐redundant users. The third grouping algorithm establishes an optimisation model with the objective of minimising average correlation within subsets. All users are then divided into a specific number of optimised groups, only one of which is required for cooperative sensing in each time. The performances of the three algorithms were quantified in terms of the average training time, classification speed and classification accuracy. Experimental results showed the proposed algorithms achieved their intended function and outperformed a conventional machine learning‐based CSS model (proposed by Karaputugala et al. ) in terms of security, energy consumption, and sensing efficiency.

In this paper, we propose a high order cumulants based spectrum sensing and power recognition (CSR) detector for hybrid interweave-underlay spectrum access, where the primary system is with multiple transmit power levels. Specifically, to detect the idle spectrum when primary user (PU) is absent, high order cumulants based spectrum sensing is performed in interweave model. When PU is detected, the working model is switched to underlay model, where detection of PU's transmit power level is performed to allow secondary user (SU) to adjust its power for fully exploring spectrum access opportunities without harmful interference to PU. Given a certain order and a certain lag, the test statistics of the proposed detector is derived by leveraging general likelihood ratio test. Since cumulants higher than second order are zero for Gaussian distributions, the proposed CSR detector can extract non-Gaussian signal from Gaussian noise even when the noise is colored. Additionally, the proposed detector does not require any prior knowledge about the noise variance, thus it is robust to noise uncertainty. Closedform results for threshold expression are derived, and numerical results are provided to evaluate the proposed detector.

To boost the performance of binary pulse amplitude modulation at low signal‐to‐noise ratio, the parameter‐tuned stochastic resonance (SR) is introduced into digital communications system. In this study, an analytical framework is developed for evaluating the system performance by approximating the probability density function of the Ornstein–Uhlebeck noise based on the central limit theorem. Expression for the bit error rate of the bistable SR system with coherent receiver is derived. Theoretical and numerical results are presented to verify the analysis that the noise can improve the performance of the SR system with non‐coherent receiver. Also, it is shown that the performance of the bistable SR system with coherent receiver is superior to that with non‐coherent receiver, and the background noise is not favourable to signal processing in the coherent receiver.

Cooperative spectrum sensing, proposed to improve the performance of spectrum sensing in cognitive radio systems where there are multiple secondary users who can cooperatively detect the presence of one primary user, is receiving significant attention. However, few cooperative sensing algorithms take the correlation among the received primary user signals into account. A feasibly efficient cooperative spectrum sensing scheme based on Cholesky decomposition of the correlation matrix of the received signals is proposed. The ratio of the maximum eigenvalue to the minimum eigenvalue of the matrix obtained by Cholesky decomposition is used to construct the test statistic. Analytical approximations for the false alarm probability and decision threshold are derived using a moment matching method. The new scheme is in the category of blind cooperative spectrum sensing schemes requiring neither information about the primary user signal nor the channel nor the noise power. The new scheme can work better than the existing eigenvalue‐based cooperative spectrum sensing methods in some conditions, and it has lower complexity.

Green communications is an inevitable trend for future communication network design, especially for a cognitive radio network. Power allocation strategies are of crucial importance for green cognitive radio networks. However, energy-efficient power allocation strategies in green cognitive radio networks have not been fully studied. Energy efficiency maximization problems are analyzed in delay-insensitive cognitive radio, delay-sensitive cognitive radio, and simultaneously delay-insensitive and delay-sensitive cognitive radio, where a secondary user coexists with a primary user and the channels are fading. Using fractional programming and convex optimization techniques, energy-efficient optimal power allocation strategies are proposed subject to constraints on the average interference power, along with the peak/average transmit power. It is shown that the secondary user can achieve energy efficiency gains under the average transmit power constraint, in contrast to the peak transmit power constraint. Simulation results show that the fading of the channel between the primary user transmitter and the secondary user receiver and the fading of the channel between the secondary user transmitter and the primary user receiver are favorable to the secondary user with respect to the energy efficiency maximization of the secondary user, whereas the fading of the channel between the secondary user transmitter and the secondary user receiver is unfavorable to the secondary user.

As spectrum sensing detects the presence of primary user (PU) signal, an efficient and reliable spectrum sensing scheme plays a critical role in cognitive radio (CR). In this paper, a novel spectrum sensing method via stochastic resonance (SR) under low signal-to-noise ratio (SNR) is presented. It is shown that the introducing of suitable additional noise can enhance sequential energy detector (SED). Theoretical analysis and the optimal SR noise probability distribution function (PDF) are given. Simulation results show that in comparison with traditional sequential energy detection, our method delivers considerable reduction on the average sample number (ASN), while maintaining a comparable detection performance under low SNR.

In this paper, we investigate the performance of adaptive modulation in spectrum sharing cognitive relay networks, where the secondary user (SU) transmits the message by one best relay with amplify and forward (AF) relaying scheme. Moreover, to guarantee the primary user (PU)'s transmission quality, the interferences caused by the SU transmitter and the selected relay are less than a predetermined interference threshold. The capacity of adaptive modulation with opportunistic AF relaying is firstly derived under independent Rayleigh fading channels. Then, with adaptive L-QAM modulation and fixed switching threshold, the outage probability, average spectral efficiency, and the average error bit rate (BER) of the SU are derived. Finally, simulation results validate our analysis and show that adaptive modulation can improve the SU's performance significantly.