Wireless federated analytics face two critical challenges: data privacy and communication efficiency, since the local data may contain sensitive information and the users may be equipped with limited communication capability. Existing methods often adopt a direct combination of privacy-preservation schemes and compression mechanisms but overlook the privacy amplification effect from errors introduced in compression and wireless communication. With such consideration, a Differentially Private Quadrature Amplitude Modulation (DP-QAM) scheme, which leverages privacy amplification from both compression and noisy wireless channels, is proposed. The privacy guarantee is established in terms of the emergingf-DP, and the trade-off between privacy, communication cost, and accuracy in terms of mean square error (MSE) is characterized in the fundamental use cases of distributed mean estimation and frequency estimation, which outperforms the state-of-the-art methods. Moreover, the advantage of the proposed method over the classic Gaussian mechanism is further demonstrated from a rate-distortion perspective. Finally, extensive simulation results validate the effectiveness of the proposed mechanism.

Joint source-channel coding (JSCC) offers a promising avenue for enhancing transmission efficiency by jointly incorporating source and channel statistics into the system design. A key advancement in this area is the deep joint source and channel coding (DeepJSCC) technique that designs a direct mapping of input signals to channel symbols parameterized by a neural network, which can be trained for arbitrary channel models and semantic quality metrics. This paper advances the DeepJSCC framework toward a semantics-aligned, high-fidelity transmission approach, called semantics-guided diffusion DeepJSCC (SGD-JSCC). Existing schemes that integrate diffusion models (DMs) with JSCC face challenges in transforming random generation into accurate reconstruction and adapting to varying channel conditions. SGD-JSCC incorporates two key innovations: (1) utilizing some inherent information that contributes to the semantics of an image, such as text description or edge map, to guide the diffusion denoising process; and (2) enabling seamless adaptability to varying channel conditions with the help of a semantics-guided DM for channel denoising. The DM is guided by diverse semantic information and integrates seamlessly with DeepJSCC. In a slow fading channel, SGD-JSCC dynamically adapts to the instantaneous channel state information (CSI) directly estimated from the channel output, thereby eliminating the need for additional pilot transmissions for channel estimation. In a fast fading channel, we introduce a training-free denoising strategy, allowing SGD-JSCC to effectively adjust to fluctuations in channel gains. Numerical results demonstrate that, guided by semantic information and leveraging the powerful DM, our method outperforms existing DeepJSCC schemes, delivering satisfactory reconstruction performance even at extremely poor channel conditions. The proposed scheme highlights the potential of incorporating diffusion models in future communication systems. The code and pretrained checkpoints will be publicly available at https://github.com/MauroZMJ/SGDJSCC, allowing integration of this scheme with existing DeepJSCC models, without the need for retraining from scratch.

Existing score-based adversarial attacks mainly focus on crafting $top$-1 adversarial examples against classifiers with single-label classification. Their attack success rate and query efficiency are often less than satisfactory, particularly under small perturbation requirements; moreover, the vulnerability of classifiers with multi-label learning is yet to be studied. In this paper, we propose a comprehensive surrogate free score-based attack, named \b geometric \b score-based \b black-box \b attack (GSBA$^K$), to craft adversarial examples in an aggressive $top$-$K$ setting for both untargeted and targeted attacks, where the goal is to change the $top$-$K$ predictions of the target classifier. We introduce novel gradient-based methods to find a good initial boundary point to attack. Our iterative method employs novel gradient estimation techniques, particularly effective in $top$-$K$ setting, on the decision boundary to effectively exploit the geometry of the decision boundary. Additionally, GSBA$^K$ can be used to attack against classifiers with $top$-$K$ multi-label learning. Extensive experiential results on ImageNet and PASCAL VOC datasets validate the effectiveness of GSBA$^K$ in crafting $top$-$K$ adversarial examples.

The prevalent distributed machine learning paradigm faces two critical challenges: communication efficiency and data privacy. SIGNSGD provides a simple-to-implement approach with improved communication efficiency by requiring workers to share only the signs of the gradients. However, it fails to converge in the presence of data heterogeneity, and a simple fix is to add Gaussian noise before taking the signs, which leads to the Noisy SIGNSGD algorithm that enjoys competitive performance while significantly reducing the communication overhead. Existing results suggest that Noisy SIGNSGD with additive Gaussian noise has the same privacy guarantee as classic DP-SGD due to the post-processing property of differential privacy, and logistic noise may be a good alternative to Gaussian noise when combined with the sign-based compressor. Nonetheless, discarding the magnitudes in Noisy SIGNSGD leads to information loss, which may intuitively amplify privacy. In this paper, we make this intuition rigorous and quantify the privacy amplification of the sign-based compressor. Particularly, we analytically show that Gaussian noise leads to a smaller estimation error than logistic noise when combined with the sign-based compressor and may be more suitable for distributed learning with heterogeneous data. Then, we further establish the convergence of Noisy SIGNSGD. Finally, extensive experiments are conducted to validate the theoretical results.

Generalization from a specific scenario to the whole network is of particular importance in wireless artificial intelligence, which usually calls for scaling up in training data and brings unaffordable costs of both data collection and model training. Motivated by the fact that local similarity among wireless scenarios probably means certain similarity in wireless channel characteristics, collecting data from representative scenarios may be an efficient way to scale down the training dataset. In this paper, we propose a proactive approach for scenario diversity assessment so as to choose the most representative scenarios for training data collection. Specifically, this approach first utilizes publicly available coarse environmental maps to extract the geometric features, then combines these with electromagnetic signal propagation models to synthesize electromagnetic characteristic distributions associated with the scenario. By employing Wasserstein distance to quantify the similarity between electromagnetic characteristic distributions across different scenarios, scenario-granularity clustering is achieved for selecting representative scenarios to construct a global dataset with sufficient information for all scenarios. Experimental results demonstrate that the proposed approach can achieve a Pearson correlation coefficient above 0.8 w.r.t. cross-scenario generalization performance, and it further improves the final multi-scenario generalization, surpassing existing methods while requiring no pre-measurement data.

Semantic communication (SemCom) is emerging as a key technology for future sixth-generation (6G) systems. Unlike traditional bit-level communication (BitCom), SemCom directly optimizes performance at the semantic level, leading to superior communication efficiency. Nevertheless, the task-oriented nature of SemCom renders it challenging to completely replace BitCom. Consequently, it is desired to consider a semantic-bit coexisting communication system, where a base station (BS) serves SemCom users (sem-users) and BitCom users (bit-users) simultaneously. Such a system faces severe and heterogeneous inter-user interference. In this context, this paper provides a new semantic-bit coexisting communication framework and proposes a spatial beamforming scheme to accommodate both types of users. Specifically, we consider maximizing the semantic rate for semantic users while ensuring the quality-of-service (QoS) requirements for bit-users. Due to the intractability of obtaining the exact closed-form expression of the semantic rate, a data driven method is first applied to attain an approximated expression via data fitting. With the resulting complex transcendental function, majorization minimization (MM) is adopted to convert the original formulated problem into a multiple-ratio problem, which allows fractional programming (FP) to be used to further transform the problem into an inhomogeneous quadratically constrained quadratic programs (QCQP) problem. Solving the problem leads to a semi-closed form solution with undetermined Lagrangian factors that can be updated by a fixed point algorithm. This method is referred to as the MM-FP algorithm. Additionally, inspired by the semi-closed form solution, we also propose a low-complexity version of the MM-FP algorithm, called the low-complexity MM-FP (LP-MM-FP), which alleviates the need for iterative optimization of beamforming vectors. Extensive simulation results demonstrate that the proposed MM-FP algorithm outperforms conventional beamforming algorithms such as zero-forcing (ZF), maximum ratio transmission (MRT), and weighted minimum mean-square error (WMMSE). Moreover, the proposed LP-MMFP algorithm achieves comparable performance with the WMMSE algorithm but with lower computational complexity.

Federated learning (FL) paradigm faces critical challenges in communication efficiency and fault tolerance. Recently, the federated probabilistic mask training (FedPM) proposes to learn a binary pruning mask instead of model parameters, which alleviates the communication overhead issue thanks to the binary nature of pruning masks. However, its robustness against malicious participants remains unexplored. This work proposes federated weighted probabilistic mask aggregation (FedWPMA), which utilizes the maximum likelihood estimation for binary masks and adapts a weighted aggregation strategy to mitigate the impact of adversarial clients that may share falsified pruning masks. A warm-up strategy is further proposed and incorporated to facilitate the training process. Extensive experimental results validate the effectiveness of the proposed method.

The ever-increasing scale and more stringent latency requirements of mobile computing tasks have driven the recent development of distributed edge computing. In distributed edge computing, a large-scale computing task is partitioned into multiple small subtasks and executed in parallel on multiple edge nodes (ENs) to reduce computation delay. In early works of this area, the computation results of the subtasks are often transmitted back in a non-cooperative manner, which may lead to suboptimal downlink communication delay. Replicated edge computing can alleviate this issue by replicating the computing task over multiple ENs to enable cooperative transmission in the downlink. However, this will inevitably entail multi-fold increase of computation costs. To bridge the gap between the conventional distributed edge computing and the replicated edge computing, a novel partial replication based distributed edge computing scheme is proposed in this work. In particular, by judiciously determining the portion of task to be replicated at the ENs, the proposed scheme can harvest cooperative transmission gains while avoiding excessive computational replication costs. Accordingly, a partial replication based delay minimization problem is formulated. By leveraging the generic alternating optimization framework, this problem can be divided into two subproblems of power allocation and task partitioning. Through analysis, a semi-closed form solution is derived for the former non-convex subproblem, while the latter subproblem turns out to be linear. Simulation results are presented to corroborate the effectiveness of the proposed scheme.

Communication overhead has become one of the major bottlenecks in the distributed training of modern deep neural networks. With such consideration, various quantization-based stochastic gradient descent (SGD) solvers have been proposed and widely adopted, among which SignSGD with majority vote shows a promising direction because of its communication efficiency and robustness against Byzantine attackers. However, SignSGD fails to converge in the presence of data heterogeneity, which is commonly observed in the emerging federated learning (FL) paradigm. In this article, a sufficient condition for the convergence of the sign-based gradient descent method is derived, based on which a novel magnitude-driven stochastic-sign-based gradient compressor is proposed to address the non-convergence issue of SignSGD. The convergence of the proposed method is established in the presence of arbitrary data heterogeneity. The Byzantine resilience of sign-based gradient descent methods is quantified, and the error-feedback mechanism is further incorporated to boost the learning performance. Experimental results on the MNIST dataset, the CIFAR-10 dataset, and the Tiny-ImageNet dataset corroborate the effectiveness of the proposed methods.

Millimeter-wave (mmWave) technology is increasingly recognized as a pivotal technology of the sixth-generation communication networks due to the large amounts of available spectrum at high frequencies. However, the huge overhead associated with beam training imposes a significant challenge in mmWave communications, particularly in urban environments with high background noise. To reduce this high overhead, we propose a novel solution for robust continuous-time beam tracking with liquid neural network, which dynamically adjust the narrow mmWave beams to ensure real-time beam alignment with mobile users. Through extensive simulations, we validate the effectiveness of our proposed method and demonstrate its superiority over existing state-of-the-art deep-learning-based approaches. Specifically, our scheme achieves at most 46.9% higher normalized spectral efficiency than the baselines when the user is moving at 5 m/s, demonstrating the potential of liquid neural networks to enhance mmWave mobile communication performance.

This paper introduces an intelligent reflecting surface (IRS)-aided integrated sensing and communications (ISAC) framework for joint signal demodulation and localization in multiuser millimeter-wave (mmWave) massive multiple-input multiple-output (MIMO) systems. A time block is divided into uplink estimation stage and downlink transmission stage. In the uplink estimation stage, a joint active beamforming at the BS and passive beamforming at the IRSs are designed based on estimated angle of arrival (AoA) information. In the downlink transmission stage, a joint signal demodulation and location sensing algorithm at the users is proposed by exploiting the statistical properties of the received signals. Numerical results demonstrate that the proposed ISAC framework can achieve centimeter-level localization accuracy while maintaining communication performance compared to communication-only systems with perfect channel state information (CSI).

Recently , over-the-air computation (OAC) has shown potential in realizing computation tasks over wireless transmission. Through proper transmit and receive beamforming design, multiple-input multiple-output (MIMO)-based OAC systems can even realize partial functions of neural networks (NNs). In this paper, we propose an OAC-NN with reconfigurable intelligent surface (RIS)-aided MIMO, in which the NN computation task can be realized through updating the RIS reflection matrix. In the proposed structure, the communication system can complete the overall simple NN-based tasks only through multiple rounds of transmissions without introducing any additional computing resources. Numerical results reflect the effectiveness of the proposed scheme and the tradeoff between communication costs and computing performance.

The expansion of the Internet of Things (IoT) has led to a significant surge in data flow over edge networks, posing substantial challenges to data mining and management. While federated edge learning (FEEL) effectively accomplishes global integration and local training based on the decentralized data sets, its deployment across expansive IoT networks introduces additional challenges. The primary issues stem from managing the interaction between the communication load and learning effectiveness. The communication loads driven by recurrent data exchanges between the user equipment (UE) and central servers exacerbate network congestion and latency issues. Moreover, the learning efficacy is undermined due to the typically nonindependent and identically distributed (non-IID) characteristics of real-world IoT data. In this article, a novel communication-efficient hierarchical FEEL framework is proposed to tackle these challenges. Specifically, UEs are adaptively clustered according to their link conditions, geographic locations, and data distributions. Small base stations (SBSs) collect local model updates from the UEs in their clusters and communicate with a macro base station (MBS) for the global model aggregation. To jointly maximize the communication gain (in terms of reducing latency) and the learning gain (in terms of improving accuracy), a clustering and resource allocation optimization problem is formulated, and a cross entropy-based method with low computational complexity is proposed. Numerical experiments validate that the proposed hierarchical FEEL system achieves fast convergence and significantly improves the system efficiency for various learning tasks and the system settings.

Channel estimation in hybrid massive multiple input multiple output (MIMO) orthogonal frequency division multiplexing (OFDM) systems is challenging as only low-dimensional channel measurement can be obtained at the receiver while the true channel is high-dimensional. In the compressed sensing (CS) based channel estimation algorithms, a well-designed measurement matrix will contribute to better estimation performance. Therefore, a temporal correlation-based self-adaptive measurement matrix design (TC-SAMMD) method is proposed, and the corresponding frame structure and channel estimation technique are developed in this paper. Specifically, pilots are transmitted twice during each block, and the process of channel estimation contains three steps: time-varying channel estimation by Kalman filtering, self-adaptive measurement matrix design, and two pilots-based channel estimation via the simultaneous orthogonal matching pursuit (SOMP) algorithm. Simulation results demonstrate that the proposed TC-SAMMD method outperforms conventional measurement matrix design approaches in terms of the channel estimation error, thanks to the “adaptiveness" to the channel’s characteristics in consecutive blocks. Besides, it is shown that the TC-SAMMD strategy with a full-ranged innovation noise can effectively mitigate the performance degradation caused by the angle shifts.

This paper delves into the potential of intelligent reflecting surfaces (IRSs) in enabling integrated sensing and communication (ISAC) in multi-user multi-path scenarios. We introduce a three-dimensional (3D) multi-user ISAC framework with distributed IRSs, which offers simultaneous signal demodulation, channel estimation, and localization. The transmission is divided into a user access stage and a downlink transmission stage. In the first stage, we propose an algorithm for simultaneous uplink signal demodulation and angles of arrival (AoA) estimation at the semi-passive IRS. Moreover, a joint active and passive beamforming scheme inspired by radar-communication, is proposed to enhance both communication and localization performance in the downlink stage, while eliminating the need for distinct localization reference signals. Numerical results demonstrate that the proposed ISAC framework achieves centimeter-level localization accuracy while maintaining comparable communication performance to communication-only systems, thus validating its effectiveness.

Federated edge learning(FEEL) is a popular distributed learning framework for privacy-preserving at the edge, in which densely distributed edge devices periodically exchange model-updates with the server to complete the global model training. Due to limited bandwidth and uncertain wireless environment, FEEL may impose heavy burden to the current communication system. In addition, under the common FEEL framework, the server needs to wait for the slowest device to complete the update uploading before starting the aggregation process, leading to the straggler issue that causes prolonged communication time. In this paper, we propose to accelerate FEEL from two aspects: i.e., 1) performing data compression on the edge devices and 2) setting a deadline on the edge server to exclude the straggler devices. However, undesired gradient compression errors and transmission outage are introduced by the aforementioned operations respectively, affecting the convergence of FEEL as well. In view of these practical issues, we formulate a training time minimization problem, with the compression ratio and deadline to be optimized. To this end, an asymptotically unbiased aggregation scheme is first proposed to ensure zero optimality gap after convergence, and the impact of compression error and transmission outage on the overall training time are quantified through convergence analysis. Then, the formulated problem is solved in an alternating manner, based on which, the noveljoint compression and deadline optimization(JCDO) algorithm is derived. Numerical experiments for different use cases in FEEL including image classification and autonomous driving show that the proposed method is nearly 30X faster than the vanilla FedSGD algorithm, and outperforms the state-of-the-art schemes.

Federated learning allows collaborative clients to solve a machine-learning problem while preserving data privacy. Recent studies have tackled various challenges in federated learning, but the joint optimization of communication overhead, learning reliability, and deployment efficiency is still an open problem. To this end, we propose a new scheme named federated learning via plurality vote (FedVote). In each communication round of FedVote, clients transmit binary or ternary weights to the server with low communication overhead. The model parameters are aggregated via weighted voting to enhance the resilience against Byzantine attacks. When deployed for inference, the model with binary or ternary weights is resource-friendly to edge devices. Our results demonstrate that the proposed method can reduce quantization error and converges faster compared to the methods directly quantizing the model updates.

Coded edge computing is envisioned as a promising solution to cope with the ever-increasing large-scale and computation-intensive mobile applications. Besides alleviating the computation straggling issue, task encoding in coded edge computing is also beneficial to the transmission of computation results. Nonetheless, existing pioneering works in this direction mainly take an information-theoretical perspective and assume the ideal scenarios of high signal-to-noise ratio. To the best of our knowledge, the issue of power control still remains largely unexplored for coded edge computing. In this work, two novel power control schemes are developed for coded edge computing in dynamic wireless environments, which apply to the repetition encoded task computing and the general linearly encoded task computing, respectively. However, the corresponding optimization problems turn out to be non-convex and highly non-trivial. To this end, by exploiting the underlying structural property, a novel partition-based iterative optimization method is developed to obtain the closed-form expression of the optimal dynamic power control strategy for repetition encoded task computing. For the case of more general linearly encoded task computing, the corresponding problem is transformed into a sum-of-ratio problem and then solved iteratively. Simulations are conducted to corroborate the effectiveness of the proposed schemes.

Distributed edge computing has been advocated as a key enabling technology to tackle large-scale intelligence applications, which is however hampered by the straggling effect. To overcome straggling, coded edge computing emerges as a promising solution by creating judiciously designed redundant computations using coding theory. Nonetheless, existing transmission schemes for coded edge computing that make edge nodes (ENs) transmit independently are often sub-optimal, as the computation results are correlated due to coding redundancy. This entails a pressing need for more effective transmission for coded edge computing. With this consideration, a noveltask-decoding assisted cooperative transmissionscheme is proposed in this work to facilitate cooperative transmission in general coded edge computing settings. Specifically, by exploiting the structural relation among the encoded sub-tasks, a task-decoding mechanism is developed to enable ENs to reconstruct computation results ofallother ENs, so that they can cooperatively transmit withanyother EN by forming a virtual multi-antenna system. To characterize the delay performance of the proposed scheme, an analytic bound with closed-form expression is derived first, followed by a more accurate algorithmic bound for scenarios with a relatively small recovery threshold. Simulations are conducted to validate the effectiveness of the proposed scheme.

In this paper, we investigate integrated sensing and communication (ISAC) in high-mobility systems with the aid of an intelligent reflecting surface (IRS). To exploit the benefits of Delay-Doppler (DD) spread caused by high mobility, orthogonal time frequency space (OTFS)-based frame structure and transmission framework are proposed. In such a framework, we first design a low-complexity ratio-based sensing algorithm for estimating the velocity of mobile user. Then, we analyze the performance of sensing and communication in terms of achievable mean square error (MSE) and achievable rate, respectively, and reveal the impact of key parameters. Next, with the derived performance expressions, we jointly optimize the phase shift matrix of IRS and the receive combining vector at the base station (BS) to improve the overall performance of integrated sensing and communication. Finally, extensive simulation results confirm the effectiveness of the proposed algorithms in high-mobility systems.

While a practical wireless network has many tiers where end users do not directly communicate with the central server, the users’ devices have limited computation and battery powers, and the serving base station (BS) has a fixed bandwidth. Owing to these practical constraints and system models, this paper leverages model pruning and proposes a pruning-enabled hierarchical federated learning (PHFL) in heterogeneous networks (HetNets). We first derive an upper bound of the convergence rate that clearly demonstrates the impact of the model pruning and wireless communications between the clients and the associated BS. Then we jointly optimize the model pruning ratio, central processing unit (CPU) frequency and transmission power of the clients in order to minimize the controllable terms of the convergence bound under strict delay and energy constraints. However, since the original problem is not convex, we perform successive convex approximation (SCA) and jointly optimize the parameters for the relaxed convex problem. Through extensive simulation, we validate the effectiveness of our proposed PHFL algorithm in terms of test accuracy, wall clock time, energy consumption and bandwidth requirement.

With the rapid development of deep learning, the parameters of modern neural network models, especially in the field of Natural Language Processing (NLP) are extremely huge. When the parameters of the model are larger even than the storage memory of a single device, it is necessary to split the original big learning model into different parts with each part assigned to one device, thus realizing joint model training over different devices (i.e., distributed training). In this paper, we aim to introduce the advanced coding scheme into the distributed parallel framework, which leads to the perfect combination of coding and the underlying calculation of neural networks. The proposed scheme is not only able to avoid the impact of poor computing power or low bandwidth and even dropped devices (stragglers) on system performance but also reduce the communication load between different devices, thereby greatly improving the performance of distributed parallel systems.

We consider a federated data analytics problem in which a server coordinates the collaborative data analysis of multiple users with privacy concerns and limited communication capability. The commonly adopted compression schemes introduce information loss into local data while improving communication efficiency, and it remains an open problem whether such discrete-valued mechanisms provide any privacy protection. In this paper, we study the local differential privacy guarantees of discrete-valued mechanisms with finite output space through the lens of $f$-differential privacy (DP). More specifically, we advance the existing literature by deriving tight $f$-DP guarantees for a variety of discrete-valued mechanisms, including the binomial noise and the binomial mechanisms that are proposed for privacy preservation, and the sign-based methods that are proposed for data compression, in closed-form expressions. We further investigate the amplification in privacy by sparsification and propose a ternary stochastic compressor. By leveraging compression for privacy amplification, we improve the existing methods by removing the dependency of accuracy (in terms of mean square error) on communication cost in the popular use case of distributed mean estimation, therefore breaking the three-way tradeoff between privacy, communication, and accuracy.

In Cooperative Intelligent Transportation Systems (C-ITS) and Connected Automated Vehicles (CAV), accessing multiple users and providing high-precision positioning are both vital. This paper aims to design an efficient deep learning approach to extend current Channel State Information (CSI)-based positioning to Frequency Division Multiple Access (FDMA) mode. In FDMA mode, different users are allocated with different subcarriers, making the user CSI have diverse frequency domain characteristics. The diverse frequency domain characteristics bring huge interference to the neural network for stable position inference, and efficient designs are required to handle this challenge. This paper proposes a novel approach named multi-frequency fusion learning for CSI-based positioning. By first using a shareable method to extract position-related features from CSI on each subcarrier independently and then fusing the obtained features, the designed neural network obtains excellent frequency domain flexibility to cope with the diverse frequency address challenge in FDMA mode. Meanwhile, we provide the feasibility analysis of this learning approach in massive Multiple-Input Multiple-Output (MIMO) systems to ensure its stable application. Based on the architecture of multi-frequency fusion learning, we propose two specific positioning schemes with differentiated designs. One is a Multi-Frequency Ensemble Network (MFENet), which extracts and fuses frequency-independent features to ensure the network is utterly unharmed by the complicated frequency domain characteristics. The other is a Multi-Frequency Cumulative Network (MFCNet), which uses sufficient feature accumulation to achieve high precision positioning. The key performance indices and applications on vehicles are comprehensively compared with popular deep-learning methods. Experiment results show the effectiveness and superiority of the proposed schemes.

This paper proposes a vehicular edge federated learning (VEFL) solution, where an edge server leverages highly mobile connected vehicles’ (CVs’) onboard central processing units (CPUs) and local datasets to train a global model. Convergence analysis reveals that the VEFL training loss depends on the successful receptions of the CVs’ trained models over the intermittent vehicle-to-infrastructure (V2I) wireless links. Owing to high mobility, in the full device participation case (FDPC), the edge server aggregates client model parameters based on a weighted combination according to the CVs’ dataset sizes and sojourn periods, while it selects a subset of CVs in the partial device participation case (PDPC). We then devise joint VEFL and radio access technology (RAT) parameters optimization problems under delay, energy and cost constraints to maximize the probability of successful reception of the locally trained models. Considering that the optimization problem is NP-hard, we decompose it into a VEFL parameter optimization sub-problem, given the estimated worst-case sojourn period, delay and energy expense, and an online RAT parameter optimization sub-problem. Finally, extensive simulations are conducted to validate the effectiveness of the proposed solutions with a practical 5G new radio (5G-NR) RAT under a realistic microscopic mobility model.

Driven by the ever-increasing scale and intensity of the computing tasks arising from various mobile applications, distributed edge computing has fostered wide research interests. It can effectively reduce the task processing delay by partitioning the original large-scale task into several small subtasks and offloading them to multiple edge nodes (ENs) for parallel computing. In edge computing, as the mobile user usually tends to offload computing tasks to closer ENs to save transmit power, the attacker may stealthily infer user location by exploiting this feature. Although there have been some pioneering works on offloading related location privacy, they mainly focused on the scenario where each task can only be offloaded to a single EN, and may not be directly applicable to distributed edge computing. Besides, the privacy issues considered in existing works are mainly based on good heuristics, and there is still a lack of concrete examples of location privacy attacks in edge computing. To the best of our knowledge, the location privacy issue in distributed edge computing still remains largely unexplored in existing literature. With this consideration, a location inference attack based on matrix sequential probability ratio test (MSPRT) is identified in this work. Besides, a countermeasure based on dynamic multi-EN selection is proposed, together with a location privacy-aware and energy-efficient distributed offloading scheme based on the generic Lyapunov optimization framework. Both theoretic analysis and simulations based on real-world channel measurements are employed to validate the feasibility of the identified MPSRT attack and the effectiveness of the proposed defense scheme.

Federated learning (FL) in a mobile edge network faces challenges from both communication and learning per-spectives. The typically non-i.i.d. data can lead to slow convergence and low accuracy. To ease these challenges, frequent communications between user equipments (UEs) and the cen-tral macro base station (MBS) are necessary, aggravating the communication burden. In this paper, a novel hierarchical FL framework is proposed to alleviate the biased convergence of the global model, achieving better communication and computation efficiency. Specifically, the UEs are adaptively clustered and allocated to specific small base stations (SBSs) according to channel conditions, geographic locations, and data distributions. The SBSs are further aggregated to the MBS, forming a hier-archical FL framework. The joint user clustering and wireless resource allocation optimization problem is formulated. To solve this problem, a cross entropy (CE) based method with low computational complexity is proposed. Simulation results validate that the proposed hierarchical FL system can save more than 87 percent training time under the EMNIST Letters dataset, achieving fast convergence and significantly improving the system efficiency.

Federated learning (FL) is a privacy-preserving paradigm where multiple participants jointly solve a machine learning problem without sharing raw data. Unlike traditional distributed learning, a unique characteristic of FL is statistical heterogeneity, namely, data distributions across participants are different from each other. Meanwhile, recent advances in the interpretation of neural networks have seen a wide use of neural tangent kernels (NTKs) for convergence analyses. In this paper, we propose a novel FL paradigm empowered by the NTK framework. The paradigm addresses the challenge of statistical heterogeneity by transmitting update data that are more expressive than those of the conventional FL paradigms. Specifically, sample-wise Jacobian matrices, rather than model weights/gradients, are uploaded by participants. The server then constructs an empirical kernel matrix to update a global model without explicitly performing gradient descent. We further develop a variant with improved communication efficiency and enhanced privacy. Numerical results show that the proposed paradigm can achieve the same accuracy while reducing the number of communication rounds by an order of magnitude compared to federated averaging.

Obtaining accurate channel state information (CSI) is crucial and challenging for multiple-input multiple-output (MIMO) wireless communication systems. The conventional channel estimation method cannot guarantee the accuracy of mobile CSI while requiring high signaling overhead. Through exploring the intrinsic correlation among a set of historical CSI instances randomly obtained in a certain communication environment, channel prediction can significantly increase CSI accuracy and save signaling overhead. In this paper, we propose a novel channel prediction method based on ordinary differential equation (ODE)-recurrent neural network (RNN) for accurate and flexible mobile MIMO channel prediction. Different from existing works using sequential network structures for exploring the numerical correlation between observed data, our proposed method tries to represent the implicit physics process of path responses changing by a specially designed continuous learning network with ODE structure. Due to the targeted design of the learning network, our proposed method fits the mathematics feature of CSI data better and enjoy higher network interpretability. Experimental results show that the proposed learning approach outperforms existing methods, especially for long time interval of the CSI sequence and large channel measurement error.

The distributed alternating direction method of multipliers (ADMM) is an efficient distributed optimization algorithm, which however shows poor convergence in time-varying network topologies. To solve the challenge, we propose TV-ADMM, a novel distributed ADMM algorithm for time-varying communication networks. More specifically, importance weight parameters are introduced in message fusion, with the purpose of mitigating the potential error brought by the network topology dynamics. Based on that, the updating rules are designed with the first-order approximation and a Bregman divergence term, which can reduce the variance caused by the randomness and enhance the robustness. Moreover, we consider two different practical scenarios with time-varying communication network. In Scenario One, the communication between two nodes succeeds with certain probabilities, based on which the importance weight parameters are designed. Scenario Two considers mobile agents, where the communication link is determined by the distance between two agents. We derive the connectivity probability in this scenario and get the corresponding importance weight. Numerical simulations validate the effectiveness of the proposed algorithm in both scenarios, in comparison with the subgradient-based method.

The dramatic growth in computing capability and the inherent mobility of the unmanned aerial vehicles (UAVs) foster the recent surge of interests in incorporating UAVs into edge computing systems to facilitate on-demand deployment and extended coverage. Nonetheless, due to the limited communication capability of the UAVs, single-UAV edge computing systems may still be incompetent when serving remote users. Although the traditional multi-UAV relay network can be a viable solution, it fails to exploit the computing capability of the UAVs. With this consideration, a multi-hop task offloading with on-the-fly computation scheme is proposed in this work to enable a more powerful multi-UAV remote edge computing network. To solve the corresponding joint resource allocation and deployment problem, two efficient algorithms are proposed. One of them can find the global optimal strategy in a special case, while the other can obtain a good local optimal strategy in the general cases. Both algorithms have a complexity only linear in the number of UAVs and admit distributed implementation. In addition to analysis, numerical results are provided to corroborate the effectiveness of the proposed scheme.

The rapid growth in scale and complexity of mobile applications fosters the development of the coded edge computing paradigm. By exploiting the redundancy in the encoded subtasks, coded edge computing enables collaborative transmission of multiple edge nodes and is promising for distributed computing in wireless fading environments. Nonetheless, to the best of our knowledge, due to challenges arising from the selection of the coding parameters, offloading strategy design for coded edge computing in general fading environments still remains open. With this consideration, the coded offloading problem is studied in this work and a delay-optimal coded offloading scheme is proposed. In particular, when the offloaded tasks are encoded by$(k,r)$linear codes, transmission diversity gains can be obtained by performing edge node selection to mitigate fading. However, the corresponding optimization problem turns out to be a highly non-trivial non-linear mixed-integer programming. To this end, through in-depth analysis based on order statistics, it is found that the average processing delay of the offloaded tasks admits a favorable$V$-structure with respect to the coding parameter$r$, under arbitrary fading distribution. This key theoretic result allows us to efficiently solve the original problem using monotonic optimization. Simulations are conducted to validate our analysis and corroborate the effectiveness of the proposed scheme.

In federated learning (FL), reducing the communication overhead is one of the most critical challenges since the parameter server and the mobile devices share the training parameters over wireless links. With such consideration, we adopt the idea of SignSGD in which only the signs of the gradients are exchanged. Moreover, most of the existing works assume Channel State Information (CSI) available at both the mobile devices and the parameter server, and thus the mobile devices can adopt fixed transmission rates dictated by the channel capacity. In this work, only the parameter server side CSI is assumed, and channel capacity with outage is considered. In this case, an essential problem for the mobile devices is to select appropriate local processing and communication parameters (including the transmission rates) to achieve a desired balance between the overall learning performance and their energy consumption. Two optimization problems are formulated and solved, which optimize the learning performance given the energy consumption requirement, and vice versa. Furthermore, considering that the data may be distributed across the mobile devices in a highly uneven fashion in FL, a stochastic sign-based algorithm is proposed. Extensive simulations are performed to demonstrate the effectiveness of the proposed methods.

Driven by the burgeoning development of unmanned aerial vehicle (UAV) technology, the recently advocated multi-UAV edge computing paradigm is anticipated to greatly enhance the coverage and on-demand deployment capability of the edge networks. One of the prominent advantage of this paradigm is to allow the UAVs to participate in the edge computing process by executing some computing tasks at their onboard processors. To this end, a key prerequisite is that the corresponding computing services must be placed onboard beforehand. Nonetheless, unlike its counterpart for conventional ground edge systems, the service placement issue in multi-UAV edge computing systems remains much less explored. To the best of our knowledge, this work is among the first to consider the joint service placement and resource allocation problem for multi-UAV edge computing. Due to the mutual influence between service placement and resource allocation, this problem turns out to be a computationally intractable mixed-integer nonlinear programming. Fortunately, through our analysis, it is found that this problem can be divided into two subproblems that are submodular and convex, respectively. Based on this observation and the general alternative optimization framework, an efficient joint service placement and resource allocation scheme that can find a reasonably good solution with only a linear complexity is proposed. In addition to the analysis, simulations are conducted to validate the effectiveness of the proposed scheme.

The recent advances in wireless sensor networks and sensing techniques enable various time-sensitive applications that require timely exchange of updates between a Base Station (BS) and ground terminals. In practice, the ground terminals may not be able to communicate with the BS directly due to constraints in transmit power and communication capability, and mobile agents are commonly employed to help collect and deliver the updates. In particular, the emerging mobile crowd sensing (MCS) provides an appealing cost-effective paradigm for such employment. However, in this case, the mobile agents are required to share their locations with the ground terminals and the BS, which incurs location privacy concerns and may deter them from participating in the information delivery process. With this consideration, a location privacy-aware payment mechanism, which can stimulate the mobile agents to report their locations with differential privacy levels desired by the BS, is proposed. Furthermore, considering that the BS usually has a limited budget, it is essential to properly select the set of mobile agents to perform the information collection tasks. Therefore, a cost-efficient mobile agent selection algorithm is proposed. Finally, simulation results are presented to demonstrate the effectiveness of the proposed method.

Driven by the dramatic growth in computing capability and the inherent mobility of the unmanned aerial vehicles (UAVs), the recently advocated UAV edge computing paradigm is expected to enhance the coverage and the on-demand deployment capability of existing terrestrial edge computing systems. Nonetheless, due to the limited onboard resource of the UAV, single- UAV edge computing systems may still be incompetent when serving remote users. Although using multiple UAVs to form a traditional relay network is a viable solution to remote edge computing, it fails to exploit the computing capability of the UAVs. This entails a pressing need to develop multi-UAV remote edge computing mechanisms that allow the UAVs to handle part of the computation tasks using their local processors while conducting multi-hop computation task offloading. To achieve the best performance in such cases, the UAVs have to properly split their power budget for communication and computation and also move to suitable service locations. Nonetheless, finding the optimal UAV power allocation and deployment turns out to be an intractable high-dimensional monotonic optimization problem, even for a mild number of UAVs. To overcome this challenge, a more efficient algorithm that has a complexity only linear in the number of UAVs is developed by exploiting the special structure of this problem. In addition to analysis, numerical results are provided to validate the effectiveness of the proposed scheme.

Recent advances in social media and information technology have enabled much faster dissemination of information, while at the same time raise concerns about privacy leakage after various privacy breaches. Therefore, the privacy guarantees of information dissemination protocols have attracted increasing research interests, among which the gossip protocols assume vital importance in various information exchange applications. Very recently, the rigorous framework of differential privacy has been introduced to measure the privacy guarantees of gossip protocols in the simplified complete network scenario. In this work, we extend the study to general networks. First, lower bounds of the differential privacy guarantees are derived for the gossip protocols in general networks in both synchronous and asynchronous settings. The prediction uncertainty of the source node given a uniform prior is also determined. It is found that source anonymity is closely related to some key network structure parameters in the general network setting. Then, we investigate information spreading in wireless networks with unreliable communications, and quantity the tradeoff between differential privacy guarantees and information spreading efficiency. Finally, considering that the attacker may not be present in the beginning of the information dissemination process, the scenario of delayed monitoring is studied and the corresponding differential privacy guarantees are evaluated.

The wireless offloading feature of the recently advocated mobile-edge computing (MEC) imposes a risk of disclosing private user data to eavesdroppers. Physical-layer security approaches that are built on information theoretic methods can be applied to defend eavesdropping in MEC. Nonetheless, directly incorporating existing physical-layer security technique may introduce extra energy and delay costs to the resource-limited mobile device and thus substantially disrupt the users' offloading decisions. To fulfill effective secure offloading in MEC, there is a compelling need to properly optimize existing physical-layer security techniques and develop new offloading schemes accordingly. With this consideration, a novel physical-layer assisted secure offloading scheme is proposed in this work, in which the edge server proactively broadcasts jamming signals to impede eavesdropping and leverages full-duplex communication technique to effectively suppress the self-interference. Finding the optimal jamming signal and the corresponding optimal offloading ratio turns out to be a challenging bilevel optimization problem. The special structure of the secure offloading problem is exploited to develop efficient offloading algorithms. Numerical results are presented to validate the effectiveness of the proposed scheme.

The limited information processing capability and battery life of mobile devices is becoming a bottleneck in delivering more advanced and high-quality services to the customers. To address this problem, the recently advocated mobile-edge computing (MEC) architecture is promising, where the essential idea is to bring the computation resource to the network edge and allow users to wirelessly offload resource demanding computation tasks to the nearby MEC servers for potentially faster execution and lower battery consumption. Nonetheless, the existing understanding of the privacy aspect of MEC is still far from complete. In this work, a user presence inference attack that invades user privacy by exploiting the feature tasks offloaded from users is identified for MEC. Existing privacy-preserving techniques developed for other applications cannot be applied to defeat this attack in MEC, as they may disrupt the optimal task offloading scheduling and cause severe degradation in user experience. With this consideration, a novel privacy-preserving and cost-efficient (PEACE) task offloading scheme that can preserve user privacy while still ensure the best possible user experience is developed in this work based on the generic Lyapunov optimization framework. The effectiveness of the proposed scheme is validated through both analysis and simulations.

As compared to the conventional cloud computing, the wireless offloading feature of the recently advocated mobile-edge computing (MEC) imposes a new risk of disclosing possibly private and sensitive user data to eavesdroppers. Physical-layer security approaches built on information theoretic methods are believed to provide a stronger notion of privacy than cryptography, and therefore, may be more suitable for defending eavesdropping in MEC. Nonetheless, incorporating a physical-layer security technique may fundamentally change the mobile users' offloading decisions. This suggests a compelling need for new judiciously designed offloading schemes that can jointly reap the benefits of both physical-layer security and MEC. With this consideration, a novel physical-layer assisted privacy-preserving offloading scheme is proposed in this work, in which the edge server proactively broadcasts jamming signals to impede eavesdropping and leverages full-duplex communication technique to effectively suppress the self-interference. Finding the optimal jamming power of the edge server and the corresponding optimal offloading ratio of the mobile user turns out to be a challenging bilevel optimization problem. By exploiting the structure of the considered problem, two efficient algorithms are developed for delay optimal and energy optimal privacy-preserving offloading, respectively. Numerical results are presented to validate the effectiveness of the proposed schemes.

The recent advances in sensor technologies and smart devices enable the collaborative collection of a sheer volume of data from multiple information sources. As a promising tool to efficiently extract useful information from such big data, machine learning has been pushed to the forefront and seen great success in a wide range of relevant areas such as computer vision, health care, and financial market analysis. To accommodate the large volume of data, there is a surge of interest in the design of distributed machine learning, among which stochastic gradient descent (SGD) is one of the mostly adopted methods. Nonetheless, distributed machine learning methods may be vulnerable to Byzantine attack, in which the adversary can deliberately share falsified information to disrupt the intended machine learning procedures. In this work, two asynchronous Byzantine tolerant SGD algorithms are proposed, in which the honest collaborative workers are assumed to store the model parameters derived from their own local data and use them as the ground truth. The proposed algorithms can deal with an arbitrary number of Byzantine attackers and are provably convergent. Simulation results based on a real-world dataset are presented to verify the theoretical results and demonstrate the effectiveness of the proposed algorithms.

The rapid uptake of Internet-of-Things (IoT) devices imposes an unprecedented pressure for data communication and processing on the backbone network and the central cloud infrastructure. To overcome this issue, the recently advocated mobile-edge computing (MEC)-enabled IoT is promising. Meanwhile, driven by the growing social awareness of privacy, significant research efforts have been devoted to relevant issues in IoT; however, most of them mainly focus on the conventional cloud-based IoT. In this paper, a new privacy vulnerability caused by the wireless offloading feature of MEC-enabled IoT is identified. To address this vulnerability, an effective privacy-aware offloading scheme is developed based on a newly proposed deep post-decision state (PDS)-learning algorithm. By exploiting extra prior information, the proposed deep PDS-learning algorithm allows the IoT devices to learn a good privacy-aware offloading strategy much faster than the conventional deep Q-network. Theoretic analysis and numerical results are provided to corroborate the correctness and the effectiveness of the proposed algorithm.

To contest the rapidly developing cyber-attacks, numerous collaborative security schemes, in which multiple security entities can exchange their observations and other relevant data to achieve more effective security decisions, are proposed and developed in the literature. However, the security-related information shared among the security entities may contain some sensitive information and such information exchange can raise privacy concerns, especially when these entities belong to different organizations. With such consideration, the interplay between the attacker and the collaborative entities is formulated as Quantitative Information Flow (QIF) games, in which the QIF theory is adapted to measure the collaboration gain and the privacy loss of the entities in the information sharing process. In particular, three games are considered, each corresponding to one possible scenario of interest in practice. Based on the game-theoretic analysis, the expected behaviors of both the attacker and the security entities are obtained. In addition, the simulation results are presented to validate the analysis.

While providing unprecedented convenience to people's daily life, location-based services (LBSs) may cause serious concerns on users' location privacy, when the system is compromised. Although various location privacy protection mechanisms have been developed for LBSs, the ambient physical environment often imposes some fundamental limitations on their performances. As a result, mobile users may experience a spatial diversity in the achievable location privacy when traveling along their routes. However, to the best of our knowledge, an appropriate location privacy metric that can capture the influence of the ambient environment is still missing in the literature. Also, none of the existing location privacy protection methods can properly leverage such spatial diversity. With this consideration, new ambient environment-dependent location privacy metrics are proposed in this paper, together with a stochastic model that can capture their spatial variations along the user's route. Based on this modeling, a new optimal stopping-based LBS access scheme that allows mobile users to fully leverage the spatial diversity and achieve a substantially better performance is developed. The effectiveness of the proposed scheme is corroborated by both numerical results and simulations over real-world road maps.

Recently, mobile-edge computing (MEC) emerges as a promising paradigm to enable computation intensive and delay-sensitive applications at resource limited mobile devices by allowing them to offload their heavy computation tasks to nearby MEC servers through wireless communications. A substantial body of literature is devoted to developing efficient scheduling algorithms that can adapt to the dynamics of both the system and the ambient wireless environments. However, the influence of these task offloading schemes to the mobile users' privacy is largely ignored. In this work, two potential privacy issues induced by the wireless task offloading feature of MEC, location privacy and usage pattern privacy, are identified. To address these two privacy issues, a constrained Markov decision process (CMDP) based privacy-aware task offloading scheduling algorithm is proposed, which allows the mobile device to achieve the best possible delay and energy consumption performance while maintain a pre-specified level of privacy. Numerical results are presented to corroborate the effectiveness of the proposed algorithm.

Deception has been widely considered in literature as an effective means of enhancing security protection when the defender holds some private information about the ongoing rivalry unknown to the attacker. However, most of the existing works on deception assume static environments and thus consider only myopic deception, while practical security games between the defender and the attacker may happen in dynamic scenarios. To better exploit the defender's private information in dynamic environments and improve security performance, a stochastic deception game (SDG) framework is developed in this work to enable the defender to conduct foresighted deception. To solve the proposed SDG, a new iterative algorithm that is provably convergent is developed. A corresponding learning algorithm is developed as well to facilitate the defender in conducting foresighted deception in unknown dynamic environments. Numerical results show that the proposed foresighted deception can offer a substantial performance improvement as compared to the conventional myopic deception.

As information systems become ubiquitous, Intrusion Detection Systems (IDSs) have assumed increasing importance. As a result, substantial amount of research efforts have been devoted to developing various intrusion detection algorithms. However, there is still no single detection algorithm that can catch all possible attacks. On the other hand, it is infeasible for practical IDSs to run all the detection algorithms simultaneously due to resource limitation, leaving potential opportunities for the adversaries to explore. This resource scarcity problem becomes more severe when the system is in an ill state (e.g., partially compromised). Enabling collaboration among multiple IDSs may be a viable way to mitigate this problem. Particularly, IDSs in the healthy state can share some of their idle computational resources to those in ill states, so as to improve the overall intrusion detection performance. Considering this, the collaborative IDS configuration problem is formulated as a two-layer stochastic game (SG) in this work and a new algorithm is proposed to solve this two-layer SG. Simulation results show that the proposed algorithm can provide an effective collaborative configuration scheme, leading to significant detection performance gain. Some performance analysis has also been given, and the conditions under which there is a guaranteed improvement in expected system performance have been derived.