This paper explores a near-field integrated sensing and communication (ISAC) system enabled by a beyond diagonal reconfigurable intelligent surface (BD-RIS), where the target and users are located at the reflective and refractive space, respectively. In contrast to most existing RIS-aided ISAC systems that employ a single-connected architecture, we investigate group-and fully-connected BD-RIS architectures. Moreover, depending on the availability of prior information about the extended target, we propose a parametric scattering model (PSM) and an unstructured channel model (UCM). For PSM and UCM, we derive the Cramér-Rao bound (CRB) of target positions and the response matrix, respectively. Both CRB minimization problems are investigated: 1) For CRB minimization in PSM, we develop the alternative optimization (AO) algorithm that leverages semidefinite relaxation (SDR) and penalty-based manifold optimization (PBMO). 2) For CRB minimization in UCM, we prove that the optimal beamformers reside on a complex matrix sphere manifold and utilize the PBMO to jointly optimize. Finally, the numerical results show that: 1) The developed schemes consistently outperform the benchmark counterparts in sensing performance, irrespective of whether prior information is available; 2) The additional prior information can effectively improve the sensing accuracy of the extended target.

Social recommendation systems leverage both user-user (u-u) social relations and user-item (u-i) collaborative interactions to improve recommendation quality. Despite their effectiveness, existing models often struggle with task-irrelevant information and misalignment between social and collaborative signals and the downstream recommendation task, leading to suboptimal performance. To address these limitations, we propose a novel framework for Effective and Consistent Information Extraction for Social Recommendation (ECSR). Our approach focuses on two key modules: (1) a task-irrelevant information discarding module that filters out noisy signals from both social relations and user-item interactions, and (2) a task-relevant information alignment module that captures both shared and view-specific task-relevant information, ensuring alignment with the recommendation objective. By integrating them into a unified form, our method extracts minimal and sufficient statistics, which significantly enhance the model's ability to predict user preferences. We validate ECSR on three real-world social recommendation datasets, demonstrating that it consistently outperforms state-of-the-art baselines.

Autonomous aerial vehicle (AAV)-assisted mobile edge computing systems have high mobility and can be deployed in various rugged terrain and emergency scenarios for communication and monitoring. However, the malicious use of AAV swarms poses a potential threat to key areas. Therefore, accurate positioning of AAV swarms is crucial for the security of high-value civilian facilities and equipment. This article investigates angle estimation of coherent signals from AAV swarms in bistatic multiple-input multiple-output radar under nonuniform noise. The nonuniform noise powers are iteratively estimated based on the structural characteristics of the covariance matrix and subsequently removed from the observations. Transmission-reception diversity smoothing is then applied to the signal subspace, obtained through higher order singular value decomposition, to recover the rank deficiency. Furthermore, a block sparse reconstruction method is proposed, utilizing the reweighted smoothed$\ell _{0}$-norm, to obtain angle estimates. This method automatically pairs the direction-of-arrivals and direction-of-departures of AAVs. Experimental results demonstrate the superiority of our approach over existing solutions.

This paper presents a robust sparse direct position determination (DPD) method for multiple emitters using distributed sensor arrays in the presence of unknown gain-phase errors. The proposed method tackles the problem under a block sparse Bayesian learning (BSBL) framework, which incorporates perturbed steering vector factorization to separate the position parameter from the gain-phase errors, making dictionary completely known without learning. This paper devises a customized hyperparameter update rule for the proposed DPD model within the foundation of the BSBL-EM method, allowing for varying block parameters instead of constraining them to be consistent. The position estimates of emitters are determined by calculating the mean value of the posterior distribution of the reconstructed waveforms. Simulations demonstrate the superior performance of the developed BSBL direct localization method over its state-of-the-art rivals, which exhibits enhanced localization accuracy and robustness against gain-phase errors.

Graph-based social recommendation improves the prediction accuracy of recommendation by leveraging high-order neighboring information contained in social relations. However, most of them ignore the problem that social relations can be noisy for recommendation. Several studies attempt to tackle this problem by performing social graph denoising, but they suffer from 1) adaptability issues for other graph-based social recommendation models and 2) insufficiency issues for user social representation learning. To address the limitations, we propose a model-agnostic graph denoising module (denoted as MADM) which works as a plug-and-play module to provide refined social structure for base models. Meanwhile, to propel user social representations to be minimal and sufficient for recommendation, MADM further employs mutual information maximization (MIM) between user social representations and the interaction graph and realizes two ways of MIM: contrastive learning and forward predictive learning. We provide theoretical insights and guarantees from the perspectives of Information Theory and Multi-view Learning to explain its rationality. Extensive experiments on three real-world datasets demonstrate the effectiveness of MADM.

The unmanned aerial vehicle (UAV) array equipped with sensors is widely used for target localization owing to its superior maneuverability. Unfortunately, limited by the current manufacturing technology, sensor arrays usually exhibit inconsistent gain and phase responses across channels, i.e., gain–phase errors, which can seriously affect the target localization accuracy. Herein, we consider that the gain–phase consistency of all array channels is not been precalibrated. For accurate target localization, we develop a system architecture for bistatic multiple-input–multiple-output (MIMO) radar equipped with a UAV array at the receiver part to realize angle estimation. First, the UAV array is controlled to move near the transmitter to receive the transmitted signals directly. Therefore, the gain–phase consistency of the transmitter can be calibrated by using the data after matched filtering and combining the known relative position information of the transmitter and receiver. Second, we control UVAs away from the transmitter to form a bistatic MIMO radar and use the synthetic aperture technique introduced by the UAV array motion to convert the receive array into partially calibrated. Meanwhile, the array manifold matrices with unknown model errors can be obtained by parallel factor decomposition. Finally, the angle estimates, gain–phase errors, and position errors are estimated by the element-wise division of the steering vectors without iteration. Moreover, our method is insensitive to sensor position errors of the original UAV array while determining angles. Simulation results demonstrate that the proposed method can obtain accurate angle estimates under the aforementioned model errors.

Direct position determination (DPD) has received much attention in emitter localization, owing to its better accuracy than conventional two-step positioning. Most of the existing DPD algorithms are developed for circular signals (CS) by using uniform linear arrays (ULAs). However, these algorithms may ignore other characters of the signals, e.g., noncircularity. The use of ULAs limits the accuracy of source localization and the number of sources that can be estimated. In this article, a weighted$\ell_{0}$-norm sparse reconstruction algorithm for noncircular signals (NCS) is developed for DPD with a designed sparse array in motion. First, a sparse array configuration named extended nested array (ENA) is devised for NCS, which consists of three subarrays. Theoretical analysis proves that the designed array can obtain higher degrees of freedom (DOFs) effectively, and reduce the mutual coupling effects between antennas. Then, a weighted$\ell_{0}$-norm sparse reconstruction algorithm is developed to improve the accuracy of DPD. Finally, simulation results are provided to demonstrate the superiority of the proposed algorithm with the designed sparse array. Our scheme can provide better localization performance than the state-of-the-art methods.

Efficient communication in massive multiple-input-multiple-output (MIMO) systems relies on accurate channel estimation to optimize signal transmission efficiency, reliability, and minimize interference and power consumption. However, the presence of nonuniform array gain-phase perturbations among antenna elements poses practical challenges, degrading the precision of estimation. In response, this article introduces a parameterized joint angle and delay estimation (JADE) method tailored for multipath channel estimation in fully uncalibrated arrays within massive MIMO systems. Our innovative spatial and frequency-based co-smoothing method is proposed to construct a rank-recovered data covariance matrix, enhancing the system’s ability to distinguish coherent multipath signals. The JADE method employs a 1-D angular spectrum and delay spectrum search under the principle of rank reduction, providing a closed-form solution for array gain-phase perturbation estimates. The deterministic Cramér-Rao lower bound for the proposed model is derived. Numerical simulations affirm the method’s superior performance. In conclusion, our approach addresses the demand for precise channel estimation in low-signal-to-noise ratio scenarios, particularly benefiting Internet of Things (IoT) applications.

In this paper, we investigate the auxiliary vehicle positioning system and localization method for intelligent transportation systems, as a supplement to the Global Navigation Satellite System which is prone to large positioning deviations and even failures in occluded scenes such as urban canyons and tunnels. The time modulated antenna arrays are first introduced into the positioning system, avoiding mutual coupling between antennas and greatly reducing the hardware cost of the vehicle terminal. The auxiliary positioning framework for the smart vehicle is advocated in conjunction with existing radio frequency signals. To take full advantage of the multiple auxiliary sources around the road net, Doppler shifts embedded into the received signals are unearthed, and a smoothed block sparse reconstruction is developed for directly locating the vehicle, providing significant enhancements of degrees of freedom and the localization accuracy. Additionally, the proposed direct localization method is robust to multichannel gain and phase mismatch in practice, and the array perturbations can be estimated and compensated without any calibration source. Extensive simulation results corroborate that the proposed system and method achieves superior localization accuracy (approximately 0.22 m error at SNR of 10 dB), outperforming its state-of-the-art counterparts.

Abstract A sparse reconstruction algorithm for the localisation of mixed near‐field and far‐field sources (MFNS) based on four‐order statistics is proposed in this study. First, utilising the structural characteristics of a uniform symmetric linear array, a fourth‐order cumulant (FOC) matrix is constructed, which decouples the angular information from the range parameters. Based on the sparse representation framework, a weighted l 1 ‐norm minimisation algorithm is developed to obtain the direction of arrivals (DOAs) of the MFNS. However, the existing selection strategy of the tuning factor is not adaptive to different observation scenarios. So a closed‐form expression of the tuning factor based on the FOC estimation error is presented. Then, another FOC matrix is constructed, which includes both the DOA and range information of the MFNS. With the DOA estimates, the two‐dimensional spatial dictionary can be reduced into a one‐dimensional dictionary, which only depends on the range parameters. Using the similar sparse reconstruction method, the range estimates of the MFNS can be obtained, and the types of the sources can be distinguished according to their range parameters. According to numerical simulations, the estimation performance of the proposed algorithm approaches the CRB in the high signal‐to‐noise ratio region, which successfully circumvents the saturation problem due to the fixed tuning factor.

In this paper, a novel efficient algorithm is developed for direction of arrival (DOA) estimation of coherent signals under the direction-dependent mutual coupling (DDMC) based on weighted subspace fitting. DOAs are determined by applying the least square fitting between signal space and the modified array manifold at first. Subsequently, we put forward an approach to calculate the DDMC matrices and the complex fading coefficients by utilizing the estimated DOAs. Without any iterations, the proposed algorithm can identify the angular information of the coherent signals in a single step in the presence of DDMC. Numerical simulation results show the effectiveness of the proposed algorithm.

In this paper, an efficient multi-beam transmission scheme that uses symbol-level precoding based on frequency diverse array (FDA) is proposed to enhance the physical layer security (PLS). Unlike the usual maximization of secrecy rate, we assume that the position information of passive eavesdropper (Eve) is not available at transmitter, which is a more realistic assumption. We use a minimum transmission message power criterion to design the precoder, subject to constraint on received signals at symbol level for per legitimate user (LU). This guarantees the valid reception of LUs to obtain the corresponding symbols under transmission messages power minimization. Then, after accurate calculation of the transmission message power, the remaining power can be allocated to artificial noise (AN), which deteriorates the quality of received signals at other regions. Numerical simulations show the validity and effectiveness of the proposed scheme.

Directional modulation (DM), as an emerging promising physical layer security (PLS) transmission technique for wireless communications, has attracted much attention over the past decade. It endows transmitters the ability to directly transmit the confidential messages to legitimate receivers along with pre‐specified directions while distorting signal waveform signatures projected along all other spatial directions to guarantee the security of information transmission. Traditional DM designs are based on the assumption that eavesdroppers (Eves) and legitimate users (LUs) are in different directions. Nevertheless, it is not always the scenario in practical applications, as it is possible that Eves and LUs are in the same directions or even at the same positions, which results in that signals received by Eves will be approximately the same as LUs’. To address this problem, the chaos scrambling (CS) technique is employed in this paper. A DM technique based on CS and artificial noise (AN) is proposed for PLS enhancement. The symbol error rate, secrecy rate, and robustness of the proposed CS‐AN‐aided scheme are analysed and simulated. Simulation results show the effectiveness of the proposed method and the PLS can also be guaranteed even if Eves are aligning with the desired directions or very close to the LUs.

In this paper, a frequency diverse array-based directional modulation with artificial noise synthesis scheme is proposed to enhance the physical layer security of wireless communications. We aim to optimize the secrecy performance by jointly optimizing the frequency offsets, the beamforming vector, and the artificial-noise projection matrix (ANPM). Specifically, we address the physical layer security problems for known locations of proximal eavesdropper (Eve) and legitimate user (LU). The beamforming vector and frequency offsets are designed to preserve the signal power at LU. The ANPM is calculated to minimize the effect of AN on LU. Furthermore, we extend our approach to the case of multi-LUs with unknown Eve locations. Being different from the case of a single LU, the frequency offsets across array antennas are optimized to equally allocate transmitted power to each LU. The numerical results show that the proposed method can provide a higher secrecy performance than conventional DM methods. In the case of multi-LUs with unknown Eve locations, the proposed method can provide a high secrecy capacity while achieving almost equal achievable capacity to each LU.