The rising popularity of Internet of things (IoTs) has spurred technological advancements in mobile internet and interconnected systems. While offering flexible connectivity and intelligent applications across various domains, IoT service providers must gather vast amounts of sensitive data from users, which nonetheless concomitantly raises concerns about privacy breaches. Federated learning (FL) has emerged as a promising decentralized training paradigm to tackle this challenge. This work focuses on enhancing the aggregation efficiency of distributed local models by introducing over-the-air computation into the FL framework. Due to radio resource scarcity in large-scale networks, only a subset of users can participate in each training round. This highlights the need for effective user scheduling and model transmission strategies to optimize communication efficiency and inference accuracy. To address this, we propose an integrated approach to user scheduling and receive beam steering, subject to constraints on the number of selected users and transmit power. Leveraging the difference-of-convex technique, we decompose the primal non-convex optimization problem into two sub-problems, yielding an iterative solution. While effective, the computational load of the iterative method hampers its practical implementation. To overcome this, we further propose a low-complexity user scheduling policy based on characteristic analysis of the wireless channel to directly determine the user subset without iteration. Extensive experiments validate the superiority of the proposed method in terms of aggregation error and learning performance over existing approaches.

AI becomes increasingly vital for telecom industry, as the burgeoning complexity of upcoming mobile communication networks places immense pressure on network operators. While there is a growing consensus that intelligent network self-driving holds the key, it heavily relies on expert experience and knowledge extracted from network data. In an effort to facilitate convenient analytics and utilization of wireless big data, we introduce the concept of knowledge graphs into the field of mobile networks, giving rise to what we term as wireless data knowledge graphs (WDKGs). However, the heterogeneous and dynamic nature of communication networks renders manual WDKG construction both prohibitively costly and error-prone, presenting a fundamental challenge. In this context, we propose an unsupervised data-and-model driven graph structure learning (DMGSL) framework, aimed at automating WDKG refinement and updating. Tackling WDKG heterogeneity involves stratifying the network into homogeneous layers and refining it at a finer granularity. Furthermore, to capture WDKG dynamics effectively, we segment the network into static snapshots based on the coherence time and harness the power of recurrent neural networks to incorporate historical information. Extensive experiments conducted on the established WDKG demonstrate the superiority of the DMGSL over the baselines, particularly in terms of node classification accuracy.

Knowledge discovered from telecom data can facilitate proactive understanding of network dynamics and user behaviors, which in turn empowers service providers to optimize cellular traffic scheduling and resource allocation. Nevertheless, the telecom industry still heavily relies on manual expert intervention. Existing studies have been focused on exhaustively exploring the spatial-temporal correlations. However, they often overlook the underlying characteristics of cellular traffic, which are shaped by the sporadic and bursty nature of telecom services. Additionally, concept drift creates substantial obstacles to maintaining satisfactory accuracy in continuous cellular forecasting tasks. To resolve these problems, we put forward an online cellular traffic prediction method grounded in Multi-Grained Spatial-Temporal feature Complementarity (MGSTC). The proposed method is devised to achieve high-precision predictions in practical continuous forecasting scenarios. Concretely, MGSTC segments historical data into chunks and employs the coarse-grained temporal attention to offer a trend reference for the prediction horizon. Subsequently, fine-grained spatial attention is utilized to capture detailed correlations among network elements, which enables localized refinement of the established trend. The complementarity of these multi-grained spatial-temporal features facilitates the efficient transmission of valuable information. To accommodate continuous forecasting needs, we implement an online learning strategy that can detect concept drift in real-time and promptly switch to the appropriate parameter update stage. Experiments carried out on four real-world datasets demonstrate that MGSTC outperforms eleven state-of-the-art baselines consistently.

The proliferation of networked devices and the surging demand for ubiquitous intelligence have given rise to the artificial intelligence of things (AIoT). However, the utilization of high-resolution analog-to-digital converters (ADCs) and numerous radio frequency chains significantly raises power consumption. This article explores a cost-effective solution using ternary ADCs (T-ADCs) in massive multiple-input–multiple-output systems for low-power AIoT and specifically addresses channel sensing challenges. The channel is first estimated through a pilot-aided scheme and refined using a joint-pilot-and-data (JPD) approach. To assess the performance limits of this two-threshold ADC system, the analysis includes its hardware-ideal counterpart, the parallel one-bit ADCs and a realistic scenario where noise variance is unknown at the receiver is considered. Analytical findings indicate that the JPD scheme effectively mitigates performance degradation in channel estimation due to coarse quantization effects under mild conditions, without necessitating additional pilot overhead. For deterministic and random channels, we propose modified expectation maximization (EM) and variational inference EM estimators, respectively. Extensive simulations validate the theoretical results and demonstrate the effectiveness of the proposed estimators in terms of mean square error and symbol error rate, which showcases the feasibility of implementing T-ADCs and the associated JPD scheme for greener AIoT smart sensing.

Pervasive and high-accuracy positioning has become increasingly important as a fundamental enabler for intelligent connected devices in mobile networks. Nevertheless, current wireless networks heavily rely on pure model-driven techniques to achieve positioning functionality, often succumbing to performance deterioration due to hardware impairments in practical scenarios. Here, we reformulate the direction finding or angle-of-arrival (AoA) estimation problem as an image recovery task of the spatial spectrum and propose a new model-driven deep neural network (MoD-DNN) framework. The proposed MoD-DNN scheme comprises three modules: a multi-task autoencoder-based beamformer, a coarray spectrum generation module, and a model-driven deep learning-based spatial spectrum reconstruction module. Our technique enables automatic calibration of angular-dependent phase error, thereby enhancing the resilience of direction-finding precision against realistic system non-idealities. We validate the proposed scheme both using numerical simulations and field tests. The results show that the proposed MoD-DNN framework enables effective spectrum calibration and accurate AoA estimation. To the best of our knowledge, this study marks the first successful demonstration of hybrid data-and-model-driven direction finding utilizing readily available commodity 5G gNodeB.

High-accuracy positioning has become a fundamental enabler for intelligent connected devices. Nevertheless, the present wireless networks still rely on model-driven approaches to achieve positioning functionality, which are susceptible to performance degradation in practical scenarios, primarily due to hardware impairments. Integrating artificial intelligence into the positioning framework presents a promising solution to revolutionize the accuracy and robustness of location-based services. In this study, we address this challenge by reformulating the problem of angle-of-arrival (AoA) estimation into image reconstruction of spatial spectrum. To this end, we design a model-driven deep neural network (MoD-DNN), which can automatically calibrate the angular-dependent phase error. The proposed MoD-DNN approach employs an iterative optimization scheme between a convolutional neural network and a sparse conjugate gradient algorithm. Simulation and experimental results are presented to demonstrate the effectiveness of the proposed method in enhancing spectrum calibration and AoA estimation.

In future mobile networks, the demand for real-time, accurate localization of multiple signal sources is paramount, but the facilities are often resource-constrained and the deploying environments are complex. In this context, we present a lightweight deep neural network in this work, which is tailored for multi-source angle-of-arrival (AoA) estimation under low signal-to-noise-ratio (SNR) conditions. The network employs mobile inverted bottleneck convolution (MBConv), known for its enhanced feature extraction capabilities and resilience to noise. By leveraging a scale attention mechanism, we effectively integrate the outputs of each layer without the need for neural architecture search. Trained on multi-channel data under low SNR, the network formulates angle estimation as a multi-label classification task. Experimental results confirm that, the proposed network demonstrates superior accuracy in extreme noise conditions and with limited snapshots, outperforming existing methodologies in multi-source scenarios.

Evolving next-generation mobile networks is designed to provide ubiquitous coverage and networked sensing. With utility of multi-view sensing and multi-node joint transmission, cell-free is a promising technique to realize this prospect. This paper aims to tackle the problem of access point (AP) deployment in cell-free systems to balance the sensing accuracy and user rate. By merging the D-optimality with Euclidean criterion, a novel integrated metric is proposed to be the objective function for both max-sum and maxmin problems, which respectively guarantee the overall and lowest performance in multi-user communication and target tracking scenario. To solve the corresponding high dimensional non-convex multi-objective problem, the Soft actor-critic (SAC) is utilized to avoid risk of local optimal result. Numerical results demonstrate that proposed SAC-based APs deployment method achieves 20% of overall performance and 120% of lowest performance.

Channel-state-information-based localization in 5G networks has been a promising way to obtain highly accurate positions compared to previous communication networks. However, there is no unified and effective platform to support the research on 5G localization algorithms. This paper releases a link-level simulator for 5G localization, which can depict realistic physical behaviors of the 5G positioning signal transmission. Specifically, we first develop a simulation architecture considering more elaborate parameter configuration and physical-layer processing. The architecture supports the link modeling at sub-6GHz and millimeter-wave (mmWave) frequency bands. Subsequently, the critical physical-layer components that determine the localization performance are designed and integrated. In particular, a lightweight new-radio channel model and hardware impairment functions that significantly limit the parameter estimation accuracy are developed. Finally, we present three application cases to evaluate the simulator, i.e. two-dimensional mobile terminal localization, mmWave beam sweeping, and beamforming-based angle estimation. The numerical results in the application cases present the performance diversity of localization algorithms in various impairment conditions.

In imperfect multiple-input multiple-output (MIMO) systems, model-based methods for performance prediction and optimization generally experience degradation in the dynamically changing environment with unknown interference and uncertain channel state information (CSI). To adapt to such challenging settings and better accomplish the network auto-tuning tasks, we propose a generic learnable model-driven framework. We further consider transmit regularized zero-forcing (RZF) precoding as a usage instance to illustrate the proposed framework. The overall process can be divided into three cascaded stages. First, we design a light neural network for refined prediction of sum rate based on coarse model-driven approximations. Then, the CSI uncertainty is estimated on the learned predictor in an iterative manner. In the last step the regularization term in the transmit RZF precoding is optimized. The effectiveness of the generic framework and the derivative method thereof is showcased via simulation results.

In this paper, we design a signal detector for OTFS modulation based on transform-domain maximal ratio combining (TD-MRC). The proposed scheme leverages the Doppler diversity and the circulant banded block diagonal structure of the effective channel matrix, which is computationally efficient compared to the traditional MRC detector due to its matrix-inversion-free nature. Another particularly appealing feature of TD-MRC is that its reliability performance improves as the maximal relative velocity increases. Numerous simulation results are presented to demonstrate the superior performance of the proposed method in comparison with the state-of-the-arts.

Orthogonal time frequency space (OTFS) has recently attracted widespread attention for it leverages frequency dispersion as a source of diversity and mathematically unifies the classical multiple-access schemes. However, as a multi-carrier modulation in nature, OTFS is susceptible to the problem of high peak-to-average power ratio (PAPR), especially when the number of symbols is large in order to obtain a high Doppler resolution at the receiver. In this work, we recast the problem of PAPR reduction as constrained optimization of the precoding matrix. To efficiently solve the underlying nonconvex maximum-norm minimization problem, we propose an iterative algorithm based on block coordinate descent. Simulation results show that the proposed method can significantly mitigate the PAPR without unduly compromising the reliability of data transmission.

Nowadays, wireless communication is rapidly reshaping entire industry sectors. In particular, mobile-edge computing (MEC) as an enabling technology for the Industrial Internet of Things (IIoT) brings a powerful computing/storage infrastructure closer to the mobile terminals and, thereby, significantly lowers the response latency. To reap the benefit of proactive caching at the network edge, precise knowledge on the popularity pattern among the end devices is essential. However: 1) the spatiotemporal variability of content popularity; 2) the data deficiency in privacy-preserving system; 3) the costly manual labels in supervised learning; as well as 4) the not independent and identically distributed (non-i.i.d.) user behaviors pose tough challenges to the acquisition and prediction of content popularities. In this article, we propose an unsupervised and privacy-preserving popularity prediction framework for MEC-enabled IIoT to achieve a high popularity prediction accuracy while addressing the challenges. Specifically, the concepts of local and global popularities are introduced and the time-varying popularity of each user is modeled as a model-free Markov chain. On this basis, we derive and validate the essential relationship between the local and global popularities and then propose an unsupervised recurrent federated learning (URFL) algorithm to predict the distributed popularity while achieving privacy preservation and unsupervised training. Moreover, a federated loss-weighted averaging (FedLWA) scheme for the parameter aggregation is further designed to alleviate the problem of non-i.i.d. user behaviors. Simulations indicate that the proposed framework can enhance the prediction accuracy in terms of a reduced root-mean-squared error by up to 60.5%–68.7% compared to other baseline methods, i.e., recommendation algorithms, centralized learning algorithms, and other distributed learning algorithms. Additionally, manual labeling and violation of users’ data privacy are both avoided.

Mobile edge computing (MEC) is a prominent computing paradigm which expands the application fields of wireless communication. Due to the limitation of the capacities of user equipments and MEC servers, edge caching (EC) optimization is crucial to the effective utilization of the caching resources in MEC-enabled wireless networks. However, the dynamics and complexities of content popularities over space and time as well as the privacy preservation of users pose significant challenges to EC optimization. In this paper, a privacy-preserving distributed deep deterministic policy gradient (P2D3PG) algorithm is proposed to maximize the cache hit rates of devices in the MEC networks. Specifically, we consider the fact that content popularities are dynamic, complicated and unobservable, and formulate the maximization of cache hit rates on devices as distributed problems under the constraints of privacy preservation. In particular, we convert the distributed optimizations into distributed model-free Markov decision process problems and then introduce a privacy-preserving federated learning method for popularity prediction. Subsequently, a P2D3PG algorithm is developed based on distributed reinforcement learning to solve the distributed problems. Simulation results demonstrate the superiority of the proposed approach in improving EC hit rate over the baseline methods while preserving user privacy.

The effective monitoring of low-altitude slow small (LSS) targets represented by unmanned aerial vehicle (UAV) is a great challenge in the field of security in recent years. Most of the existing infrared (IR) small target algorithms focus on high-altitude target detection. However, the low-altitude background is complex and changeable, and high-intensity suspected targets exist widely. Existing methods usually cause high false alarm or failure detection for LSS targets. In this letter, we propose a novel spatiotemporal saliency method for LSS IR targets in image sequences. First, spatial variance saliency mapping and temporal gray saliency mapping are calculated in spatial domain and temporal domain, respectively. Then, the fusion saliency map is obtained by fusing the spatial saliency map and temporal saliency map. Finally, the target is extracted by a simple adaptive threshold segmentation. The proposed method is verified in five low-altitude IR image sequences. Experimental results demonstrate that the proposed method can achieve better detection performance than the existing state-of-the-art methods for LSS targets.

The problem of beam alignment and tracking in high mobility scenarios such as high-speed railway(HSR) becomes extremely challenging, since large overhead cost and significant time delay are introduced for fast time-varying channel estimation. To tackle this challenge, we propose a learning-aided beam prediction scheme for HSR networks, which predicts the beam directions and the channel amplitudes within a period of future time with fine time granularity, using a group of observations. Concretely, we transform the problem of high-dimensional beam prediction into a two-stage task, i.e., a low-dimensional parameter estimation and a cascaded hybrid beamforming operation. In the first stage, the location and speed of a certain terminal are estimated by maximum likelihood criterion, and a data-driven data fusion module is designed to improve the final estimation accuracy and robustness. Then, the probable future beam directions and channel amplitudes are predicted, based on the HSR scenario priors including deterministic trajectory, motion model, and channel model. Furthermore, we incorporate a learnable non-linear mapping module into the overall beam prediction to allow non-linear tracks. Both of the proposed learnable modules are model-based and have a good interpretability. Compared to the existing beam management scheme, the proposed beam prediction has (near) zero overhead cost and time delay. Simulation results verify the effectiveness of the proposed scheme.

Radar-based human activity classification is crucial for applications such as healthcare monitoring, fall detection, and assisted living due to its superior sensing capabilities and privacy protection. Traditional classification methods generally retrieve features from the time-range domain or the time-frequency (TF) domain. Such 2-D representation neglects the underlying dependence between the three radar signal variables of time, range, and Doppler frequency, and cannot fully depict the dynamic human motion features. In this article, we propose a time-range-Doppler radar point clouds (RPCs)-based learning model for human activity classification using a frequency-modulated continuous waveform (FMCW) radar. The human echoes are first transformed into a series of 3-D point cloud cubes integrating the motion signatures in three domains, namely time-range, time-Doppler, and range-Doppler domains. The generated RPC cubes are then fed into a newly developed two-layer convolutional multilinear principal component analysis network (CMPCANet) for feature extraction and motion classification. The CMPCANet comprises a simple network architecture with small training parameters, and can be directly implemented on the 3-D tensor dataset to extract highly discriminative features. Experimental results demonstrate that proposed framework can achieve superior classification accuracy and noise robustness compared to other methods using multidomain information, even with small training samples.

Beam alignment and tracking for millimeter-wave communication networks in highly mobile scenarios, such as high-speed railway, suffer from large overhead cost and time delay loss. To solve this problem, we propose a learning-aided beam management scheme, which divides the high-dimensional beam prediction procedure into two stages, i.e., parameter estimation and hybrid beamforming. The locations and velocities of the mobile terminals are estimated using the maximum likelihood criterion, and a data fusion module is employed to further improve the estimation accuracy and robustness. Then, the next probable beam directions and the corresponding hybrid precoders are derived based on the estimated parameter set. Numerical simulations show that, the proposed method yields significantly lower overhead cost and time delay compared to the existing beam management scheme.

In this paper, we investigate the problem of edge caching (EC) optimization in a multi-user privacy-preserving mobile edge computing (MEC) system. The time-varying content popularity is considered and the primary objective is to maximize the EC hit rate on each caching entity in the distributed network. To this end, we introduce the concept of local and global popularities and cast the time-varying local popularities as model-free Markov chains. Next, an unsupervised recurrent federated learning (URFL) algorithm is proposed to predict the popularities while achieving privacy-preserving goal. The underlying distributed optimization problem is then reformulated as a distributed Markov decision process and solved by the privacy-preserving distributed deep deterministic policy gradient algorithm incorporating the URFL algorithm. Simulation results demonstrate the superiority of the proposed scheme in terms of prediction error and hit rate over the baseline methods.

Orthogonal time frequency space (OTFS) as an emerging modulation technique in the 5G and beyond era exploits full time-frequency diversity and is robust against doubly-selective channels in high mobility scenarios. In this work, we consider an OTFS modulation based automotive joint radar-communication system and focus on the design of low-complexity parameter estimation algorithm for radar targets. It is well known that target parameter estimation in OTFS radar is computationally much more expensive than the orthogonal frequency division multiplex based platform, which hampers low-cost and real-time implementation. In this context, an efficient Bayesian learning scheme is proposed for OTFS automotive radars, which leverages the structural sparsity of radar channel in the delay-Doppler domain. We also reduce the dimension of the measurement matrix by incorporating the prior knowledge on the motion parameter limit of the true targets. Numerical simulation results are presented to demonstrate the superior performance of the proposed method in comparison with the state-of-the-art.

Federated learning (FL) as a promising edge-learning framework can effectively address the latency and privacy issues by featuring distributed learning at the devices and model aggregation in the central server. In order to enable efficient wireless data aggregation, over-the-air computation (AirComp) has recently been proposed and attracted immediate attention. However, fading of wireless channels can produce aggregate distortions in an AirComp-based FL scheme. To combat this effect, the concept of dynamic learning rate (DLR) is proposed in this work. We begin our discussion by considering multiple-input-single-output (MISO) scenario, since the underlying optimization problem is convex and has closed-form solution. We then extend our studies to more general multiple-input-multiple-output (MIMO) case and an iterative method is derived. Extensive simulation results demonstrate the effectiveness of the proposed scheme in reducing the aggregate distortion and guaranteeing the testing accuracy using the MNIST and CIFAR10 datasets. In addition, we present the asymptotic analysis and give a near-optimal receive beamforming design solution in closed form, which is verified by numerical simulations.

Frequency-difference electrical impedance tomography (fdEIT) was originally developed to mitigate the systematic artifacts induced by modeling errors when a baseline dataset is unavailable. Instead of fine anatomical imaging, only coarse anomaly detection has been addressed in current fdEIT research mainly due to its low spatial resolution. On the other hand, there has been not enough study on fdEIT reconstruction algorithm as well. In this article, we propose an efficient and high-spatial-resolution algorithm for simultaneously reconstructing multiple fdEIT frames corresponding to inject currents with multiple frequencies. The electrical impedance tomography reconstruction problem is considered within a hierarchical Bayesian framework, where both intratask spatial clustering and intertask dependency are automatically learned and exploited in an unsupervised manner. The computation is accelerated by adopting a modified marginal likelihood maximization approach. Real-data experiments are conducted to verify the recovery performance of the proposed algorithm.

Mobile edge computing (MEC) has received broad attention as an effective network architecture and a key enabler of the wireless virtual reality (VR) video service which is expected to take a huge share of communication traffic. In this work, we investigate the scenario of multi-tiles-based wireless VR video service with the aid of MEC network, where the primary objective is to minimize the system energy consumption and the latency as well as to arrive at a tradeoff between these two metrics. To this end, we first cast the time-varying view popularity as a model-free Markov chain and use a long short-term memory autoencoder network to predict its dynamics. Then, a mixed strategy, which jointly considers the dynamic caching replacement and the deterministic offloading, is designed to fully utilize the caching and computing resource in the system. The underlying multiobjective optimization problem is reformulated as a partially observable Markov decision process and solved by using a deep deterministic policy gradient algorithm. The effectiveness of the proposed scheme is confirmed by numerical simulations.

In this paper, we consider the reconstruction of three-dimensional (3-D) conductivity distribution using electrical impedance tomography (EIT) technique. A high-resolution and efficient algorithm is developed to solve the EIT inverse problem. The presented algorithm is extended upon a recently proposed novel EIT reconstruction approach based on structure-aware sparse Bayesian learning (SA-SBL). The correlation between proximal layers in the 3-D geometry are incorporated into the structure prior to improve the reconstruction accuracy. In addition, an efficient approach based on approximate message passing is developed to accelerate the large-scale 3-D learning process. To validate the algorithm, numerical experiments using real recorded data are conducted. The visual and quantitative-metric comparisons show that the proposed method outperforms the existing methods in terms of reconstruction accuracy and computational complexity in all test cases. The SA-SBL-based reconstruction approach can preserve the 3-D structure of medical volume, reduce the systematic artifacts, and improve the computational efficiency.

Electrical impedance tomography (EIT) is developed to investigate the internal conductivity changes of an object through a series of boundary electrodes, and has become increasingly attractive in a broad spectrum of applications. However, the design of optimal tomography image reconstruction algorithms has not achieved the adequate level of progress and matureness. In this paper, we propose an efficient and high-resolution EIT image reconstruction method in the framework of sparse Bayesian learning. Significant performance improvement is achieved by imposing structure-aware priors on the learning process to incorporate the prior knowledge that practical conductivity distribution maps exhibit clustered sparsity and intra-cluster continuity. The proposed method not only achieves high-resolution estimation and preserves the shape information even in low signal-to-noise ratio scenarios but also avoids the time-consuming parameter tuning process. The effectiveness of the proposed algorithm is validated through comparisons with state-of-the-art techniques using extensive numerical simulation and phantom experiment results.

Image segments are often used as a constraint in stereo matching. However, both over‐segmentation and under‐segmentation can lead to disparity degradation in some regions. To obtain an accurate disparity map, a modified semi‐global matching (SGM) algorithm is proposed which is based on adaptive window models. Introducing the object notion and build a new global energy function to optimise segments and estimate a disparity map jointly. The effective and efficient block coordinate descent approach is used to optimise the global energy function by merging small segments. The authors’ demonstrate the performance of the proposed algorithm on the KITTI and Middlebury benchmarks. The results show that the authors’ algorithm outperforms many state‐of‐the‐art methods and confirm the effectiveness of approach.

Fast and accurate detection of elderly falls can significantly reduce the rate of morbidity and mortality. In the past decade, extensive research has been performed to achieve real-time fall monitoring solutions. In this paper, we consider the radar-based modality and utilize the family of fractional Fourier transform to enhance the motion Doppler signature of falls. Compare with the conventional time-frequency analysis approaches, the proposed method achieves higher signal energy concentration and thus yields improved fall detection in low signal-to-noise ratio scenarios. Experimental results are used to validate the theoretical analysis and to demonstrate the feasibility of the proposed approach.