This letter proposes the Dynamic Margin Meta Metric Learning (DM-MML) framework to address unknown class interference and sample scarce challenges in few-shot open set specific emitter identification. In DM-MML, the dual-loop meta-training phase adopts a hierarchical margin strategy with triple-optimization, where in the inner loop, strong margins enhance feature separability for source domain emitters and the outer loop applies weak margins to calibrate target domain distributions. This strategy simultaneously preserves open decision boundaries for potential unknown class samples. During meta testing, a collaborative decision mechanism combines distance metrics and confidence scores, enabling precise known-class identification while maintaining open-space rejection capacity for unknown classes. Experiments on the WiSig dataset demonstrate superior performance over existing benchmarks under few-shot conditions.

Accurate radio environment maps (REMs) can enhance the performance of wireless networks and optimize spectrum utilization efficiency. However, in rugged terrain environments, radio propagation is significantly affected by terrain variations, resulting in spatial heterogeneity in received signal strength (RSS) and impairing the accuracy of REM construction. To address these challenges, a Gaussian Process Regression method incorporating terrain (GPRT) is proposed to exploit both spatial and terrain correlation properties. In GPRT, a specialized kernel function is designed to integrate digital elevation data into the Gaussian process framework, capturing anisotropic spatial correlation and terrain effects. In addition, an Adaptive Moment Estimation (Adam) optimization algorithm is utilized for efficient hyperparameter tuning, enhancing convergence speed and parameter accuracy. Simulations with varying numbers of emitters and field experiments in real-world terrain demonstrate the superiority and effectiveness of the proposed GPRT over competing methods in terms of robustness and accuracy. Specifically, GPRT outperformed the best comparative approaches by 20% to 33% in simulations and by up to 20% in the field experiment.

Specific Emitter Identification (SEI) is crucial for ensuring the security of physical layer communication. However, signal characteristics can be affected by various factors such as environmental and equipment variations. An effective SEI system must continuously learn and adapt to these changes to maintain accurate signal recognition. This study proposes an advanced domain incremental learning (DIL) framework for SEI, named Adaptive Domain-Incremental Learning with Knowledge Replay and Domain Alignment (ADIRA). ADIRA employs knowledge replay and distillation strategies, along with adaptive coefficients, to balance the model’s performance in recognizing signals across both new and old domains. To address the variations in signal data feature distributions across different domains, we introduce a domain alignment strategy based on adversarial training. This approach integrates embedding distillation loss with supervised contrastive loss, significantly enhancing the model’s adaptability to domain changes. Experimental results on two benchmark datasets demonstrate that ADIRA achieves performance only 0.42% and 1.71% lower than joint training, with replay samples constituting just 1.1% and 1.5% of the training set, effectively mitigating catastrophic forgetting.

Abstract Radio frequency fingerprint identification (RFFI) is a widely used technique for authenticating equipment. It identifies transmitters by extracting hardware defects found in the RF front end. Recent research has focused on the impact of transmitters and wireless channels on radio frequency fingerprint (RFF). Most work is based on the same receiver assumption, while the influence of the receiver on RFF remains unresolved. This paper focuses on the impact of receiver hardware characteristics on RFF and proposes a few‐shot cross‐receiver RFFI method based on feature separation. Data augmentation with noise addition and simulated channels addresses sparse sample issues and enhances the model's robustness to channel variations. Simultaneously, feature separation is realized by reducing the correlation between transmitter and receiver features through classification loss and similarity loss. We evaluate the proposed approaches using a large‐scale WiFi dataset. It is shown that when a trained transmitter classifier is deployed on new receivers with only 30 samples per trained transmitter, the average identification accuracy of the proposed method is 83.6%. This accuracy is 9.45% higher than the baseline method without considering transmitter hardware influence. After fine‐tuning, the average identification accuracy can reach 98.25%.

Abstract Aiming at the problems of weak robustness of single feature and limited recognition range in modulation recognition, this paper proposes a modulation recognition algorithm based on cascaded feature fusion and multi‐classifier combination, in which time‐frequency map and instantaneous amplitude spectral density features are extracted from low intermediate frequency (IF) signal, constellation map and high‐order cumulant features are extracted from zero IF signal, and a three‐level recognition algorithm is designed through the decision fusion of decision tree, convolutional neural network, and support vector machine. In order to verify the performance of the algorithm, a modulation recognition system is designed and built based on USRP2974. The low IF and zero IF modulation signals are received through two channels, and the RF signals are received and processed in real time with the help of the built‐in CPU of the receiver. The recognition of 13 kinds of analog modulation and digital modulation signals is realized. Under the condition of wireless reception, the recognition rate of this system is more than 93% at 5 to 10 dB.

This article investigates secure transmission in an energy harvesting (EH) sensor network in the presence of noncolluding passive eavesdroppers, where the EH sensors communicate with a controller via a buffer-aided relay (sink node). Each sensor has a battery to accumulate the energy harvested from the radio-frequency (RF) signals transmitted by the relay, and purely relies on the harvested energy to perform transmission. Besides, equipping a data buffer at the relay for security improvement, each time slot can be allocated dynamically to the first hop or the second hop, but it also introduce a packet delay. To enhance the security, an adaptive link selection with user scheduling and cooperative jamming (ALS-USaCJ) scheme is proposed. For evaluating the secrecy performance, the coupled evolution of the batteries and data buffer should be dealt with. Specifically, discretizing the energy status of the batteries, the couple evolution is modeled as a Markov chain so that its steady-state distribution can be obtained. On this basis, the analytical expressions of the communication outage probability (COP), secrecy outage probability (SOP), secrecy throughput (ST), and average packet delay (APD) are derived. Finally, theoretical analysis and simulation results demonstrate that equipping a data buffer at the relay can significantly improve the secrecy performance, and the larger the buffer size, the better the performance, but the greater the APD. Moreover, the link selection factor in the proposed scheme, which is related to specific network configuration, can be properly set to achieve the optimal secrecy performance.

This article investigates secure communication between a source and a destination via multiple radio frequency (RF) energy harvesting (EH) relays, in which the RF-EH relays are untrusted and apply the amplify-and-forward policy. On the one hand, to prevent the untrusted relays from eavesdropping the confidential information, the destination-aided jamming is employed, in which the destination emits jamming signal to interfere the relays while the source transmits information signal. On the other hand, the power splitting (PS) policy is adopted at the relays to harvest energy and process information, in which every relay has a finite energy storage to accumulate the energy harvested from the source’s information signal and destination’s jamming signal. To achieve energy-efficient and fully distributed implementation, we propose an energy-aware distributed beamforming (EADB) scheme, in which each relay only needs local information to decide whether to assist the source-destination communication. To evaluate the secrecy performance of the EADB scheme, the charging and discharging behaviors of the relays’ energy storage are first tracked using the Markov chain. On this basis, we derive analytical expressions for the hybrid outage probability (HOP) and secure energy efficiency (SEE) of the considered network. Finally, numerical results show that the EADB scheme has a better secrecy performance than the existing scheme. In addition, despite the curiousness of the untrusted relays, deploying more relays and increasing their energy storage capacity can enhance the security of the considered network.

In this paper, we investigate the secrecy performance of full-duplex multi-antenna spectrum-sharing wiretap networks in which a jamming signal is simultaneously transmitted by the full-duplex secondary receiver (Bob) based on the zero forcing beamforming (ZFB) algorithm. For the security enhancement, we propose the two antenna reception schemes: 1) random selection combining (RSC) where Bob selects LB antennas at random to combine the received signals and 2) generalized selection combining (GSC) where Bob selects LB strongest antennas to combine the received signals. We derive the exact closed-form expressions for the secrecy outage probability of full-duplex multi-antenna spectrum-sharing wiretap networks with ZFB algorithm. In order to explore a new design of the proposed schemes, we provide tractable asymptotic approximations for the secrecy outage probability in high signal-to-noise ratio regime under two distinct scenarios. From the analysis, we demonstrate that: 1) when the main channel is much better than the eavesdropper's channel, GSC/ZFB scheme achieves full diversity NB, while RSC/ZFB scheme only achieves partial diversity LB and 2) GSC/ZFB scheme achieves better secrecy performance than RSC/ZFB with different antenna numbers at Bob.

In this paper, we analyze the secrecy performance of multi-antenna cognitive wiretap network, where the secondary transmitter (Alice) communicates with the secondary receiver (Bob) in the presence of an eavesdropper (Eve). Specifically, we investigate the cases of maximal-ratio combining (MRC) with half-duplex (HD) and selection combining/Zero forcing beamforming (SC/ZFB) scheme with full-duplex (FD) operation, respectively. Assuming the Rayleigh fading, closed-form expressions for the secrecy outage probability of cognitive wiretap channels with MRC and SC/ZFB are derived. Furthermore, we provide simple asymptotic approximations for the secrecy outage probability and find that both schemes achieve full diversity. In addition, simulation results reveal that MRC outperforms SC/ZBF in the low interference threshold regime, while the opposite holds in the high interference threshold regime.

In this paper we analyze the effect of co-channel interference on the performance of cooperative decode-and-forward (DF) automatic-repeat-request (ARQ) scheme under Rayleigh fading channels. We assume that the relay node is affected by one interferer and the destination node is affected by another one. The outage probability of the cooperative DF ARQ scheme is derived based on the maximal ratio combining (MRC) technique. The closed-form expression clearly shows the outage probability has much to do with the maximum number of ARQ transmissions. Finally, Monte Carlo simulation results verify our theoretical analysis.