This paper investigates covert communication in spectrum-sharing Internet of Things (IoT) device-to-device (D2D) networks assisted by unmanned aerial vehicles equipped with intelligent reflecting surfaces (UAV-IRS). Existing schemes assume known or deterministic eavesdropper locations, which limits their effectiveness in the presence of spatially random adversaries in IoT environments. They also often neglect severe interference in spectrum-sharing environments and lack effective protection mechanisms for cellular links, resulting in degraded covert reliability and cellular performance. To address these challenges, we propose a UAV–IRS-assisted D2D covert communication scheme for spectrum-sharing networks, where a full-duplex base station simultaneously receives cellular signals and generates artificial noise (AN) to combat spatially random eavesdroppers. The spatial uncertainty is characterized by modeling three representative configurations between the eavesdroppers’ surveillance regions and the guard zone of the cellular link, i.e., disjoint, overlapping, and contained. Meanwhile, the average minimum detection error probability is derived as the covertness requirement. A two-layer optimization framework is then designed to jointly optimize D2D and AN transmit powers, UAV location, and IRS phase shifts, while ensuring cellular reliability. Simulation results demonstrate the effectiveness of the proposed scheme in enhancing the covert rate in IoT spectrum-sharing networks.

This paper investigates covert transmission in semantic communication systems, with a particular focus on protecting sensitive information embedded in images. While current research on semantic secure communication emphasizes semantic confidentiality, it often overlooks the covertness of semantic content. To bridge this gap, we propose a semantic covert communication scheme comprising a semantic covert artificial noise module and a deep learning module. The semantic covert artificial noise module ensures that the covertness-oriented artificial noise (CoAN) lies within the null space of the legitimate user’s channel, so that it perturbs only the eavesdropper while preserving the legitimate user’s ability to recover semantic information. The deep learning module is designed to distinguish sensitive content from the background in an image. Furthermore, the semantic covert communication neural network is trained using a customized information-bottleneck–inspired loss function that simultaneously enhances the semantic expressiveness of the representation observed by the legitimate user and improves the semantic representational capability of the CoAN at the eavesdropper, thereby inducing the eavesdropper to misinterpret the transmitted information as background. Simulation results demonstrate that the proposed scheme conceals sensitive information from the eavesdropper’s perspective while maintaining high-fidelity semantic recovery for the legitimate user.

This paper investigates backdoor attacks in image oriented semantic communications. The threat of backdoor at tacks on symbol reconstruction in semantic communication (Sem Com) systems has received limited attention. Existing research on backdoor attacks targeting SemCom symbol reconstruction primarily focuses on input-level triggers, which are impractical in scenarios with strict input constraints. In this paper, we propose a novel channel-triggered backdoor attack (CT-BA) framework that exploits inherent wireless channel characteristics as activation triggers. Our key innovation involves utilizing fundamental channel statistics parameters, specifically channel gain with different fading distributions or channel noise with different power, as potential triggers. This approach enhances stealth by eliminating explicit input manipulation, provides flexibility through trigger selection from diverse channel conditions, and enables automatic activation via natural channel variations without adversary intervention. We extensively evaluate CT-BA across four joint source-channel coding (JSCC) communication system architectures and three benchmark datasets. Simulation results demonstrate that our attack achieves near-perfect attack success rate (ASR) while maintaining effective stealth. Finally, we discuss potential defense mechanisms against such attacks.

In this paper, we investigate fairness-aware intelligent reflecting surface (IRS)-assisted multiple-user uplink communication systems. Current IRS-assisted communication methods do not consider the fairness issues of communication performance for multiple users. We formulate the IRS phase shift and receive beamforming design as an α-fair utility maximization problem by considering different fairness metrics, including zero fairness and proportional fairness, followed by an iterative algorithm to solve this problem. To overcome the high computational complexity of the iterative algorithm, we introduce a deep learning-based framework, which leverages convolutional neural networks with residual and attention mechanisms for real-time phase shift prediction. Extensive numerical simulations demonstrate that the proposed method achieves performance comparable to the traditional approach while significantly reducing computational overhead, making it a promising solution for future fairness-aware IRS-assisted communication systems.

This article investigates the energy-efficient computation offloading problem in a meta computing environment, integrating multiple-input multiple-output (MIMO) technologies and statistical channel state information (CSI). Existing resource allocation approaches always depend on instantaneous CSI, which is impractical for highly dynamic meta computing scenarios due to rapid CSI fluctuations. To address this issue, we propose a novel computation offloading scheme that jointly optimizes transmit power and computation resource allocation for multi-antenna users by leveraging statistical CSI. Specifically, a closed-form expression for the ergodic capacity and a tight bound for the MIMO diversity transmission scheme are derived. These theoretical results decouple the complex joint optimization into two tractable subproblems, where the ergodic capacity expression is utilized for global computation resource assignment via a Kuhn–Munkres (KM) algorithm, while the bound significantly simplifies the transmit power optimization problem. Experimental results demonstrate that the proposed scheme achieves substantial energy consumption reductions compared to conventional methods, highlighting the effectiveness of utilizing statistical CSI in meta computing resource allocation.

This article proposes an energy-efficient covert offloading scheme for blockchain-enabled Internet of Things (IoT), allowing sensors to upload tasks undetected by adversaries while ensuring satisfaction in paid computation offloading. Covert communication conceals the existence of transmitted signals or links. However, existing schemes primarily rely on artificial noise (AN) or wireless channel uncertainty, resulting in low covert rates for IoT offloading scenarios. Additionally, blockchain-enabled IoT, being value-oriented, necessitates consideration of sensors’ satisfaction during covert offloading. To tackle these challenges, the proposed scheme combines the adversary’s channel estimation errors with AN to enhance the covert rate, while also matching sensors’ satisfaction with the computation resources of mobile edge servers. Notably, a closed-form expression of the average minimum error detection probability is derived to maximize the effective covert rate. Furthermore, an integrated algorithm combining the Kuhn-Munkres (KM) algorithm with two bubble sort algorithms is designed to minimize energy consumption. Both analytical and simulation results demonstrate that the proposed scheme significantly reduces energy consumption compared to existing solutions.

While many existing movie recommendation systems have been integrated to personalize entertainment in FAVs, they face challenges in addressing the diverse and dynamic preferences of multiple passengers. To tackle these issues, this article introduces the multigate mixture of experts for multiinterest model (MEMI) specifically designed for FAVs. The proposed model employs a Transformer-based multi-interest extractor within a multigate mixture of experts (MMoE) structure to capture a range of person interests while managing network complexity. Additionally, a novel peak interest alignment (PIA) loss function is introduced to improve consistency between the training and inference phases, ensuring more accurate recommendations. Experimental evaluations using the Movielens dataset demonstrate that the proposed model significantly outperforms existing systems, providing more personalized and effective movie recommendations.

The convergence of aUtonomous aerial vehicles (UAVs) and federated learning (FL) has emerged as a promising paradigm to facilitate artificial intelligence (AI) services with enhanced privacy preservation. However, notwithstanding the inherent advantages of FL in terms of privacy protection, attackers can still exploit inference attacks to deduce raw data of UAVs. The existing studies predominantly assume FL servers (hereafter servers)to be fully rational and have access to all privacy preference information of UAVs (i.e., information symmetry scenario), in the design of privacy-preserving incentive schemes. To tackle these challenges, we propose a privacy-preserving incentive scheme for UAV-aided FL in the presence of information asymmetry while considering the serverexhibits bounded rationality. Specifically, a practical UAV-aided FL framework is first introduced to enable AI model training between UAVs and the server with bounded rationality. In addition, based on differential privacy, we quantify the privacy level of UAVs and subsequently analyze its impact on the aggregation accuracy of the server. This scenario entails two conflicting objectives: the server aims for higher-quality local models to achieve better aggregation accuracy, while UAVs prioritize injecting more noise into their local models to enhance privacy protection. To reconcile the conflicting objectives, we develop an incentive mechanism based on contract theory to optimize the server’s aggregation accuracy in the presence of information asymmetry. Furthermore, we employ prospect theory (PT) to the above contract to capture biases in the server’s subjective decision-making process. Besides, we deduce closed-form solutions for optimal contracts under PT and expected utility theory (EUT), where participants are assumed to be fully rational. Finally, simulation results validate the superiority of our proposed scheme in motivating UAVs to share high-quality local models and improving the aggregation accuracy of the server.

The rise of the Metaverse has ushered in a new era of social networking, offering users deeply engaging spaces to connect and participate in social activities. However, rendering these virtual environments is resource-intensive. With many users accessing simultaneously and requiring diverse Metaverse services, optimizing Metaverse resources to deliver the best quality-of-experience (QoE) for users is a significant challenge. In this paper, we propose a cooperative virtual reality (VR) rendering and dynamic resource leasing mechanism to address this issue. Specifically, we first introduce a cooperative VR scene pre-rendering framework between users and Planets (i.e., edge servers hosting users), and establish a new user QoE metric named EdgeVRQoE which considers both rendering delay and visual quality. We formulate the multidimensional rendering resources (e.g., GPU, CPU, and outbound bandwidth) leasing problem between Planets and users as a double-layer decision problem, and devise a hybrid action multi-agent reinforcement learning-based dynamic resource auction mechanism to efficiently allocate limited resources of Planets in a distributed and adaptive manner. Extensive simulations demonstrate that our proposed scheme outperforms the representatives in user QoE and resource utilization efficiency. Particularly, the proposed scheme shows at least an 18-fold improvement in QoE over other schemes, demonstrating its capability in providing immersive Metaverse experiences.

This paper explores the fairness issues on physical layer security (PLS) in intelligent reflecting surface (IRS) assisted multiple-user uplink systems. Due to unknown instantaneous eavesdropper channel state information (CSI), it is not possible to acquire exact secrecy rate of a PLS system. In this paper, we introduce secrecy outage probability (SOP) as a security metric, instead of secrecy rate as used in most existing works, and formulate a minimization problem of maximum (min-max) SOP among multiple users. To solve this problem, we propose two independent approaches: one is alternating optimization (AO) and the other is a deep learning based (DL) scheme. The AO scheme decouples the problem into two sub-problems to alternately optimize phase shift matrix and receiver beamforming vectors, which gives a near-optimal performance but with a high complexity. The DL scheme, on the other hand, works based on neural networks through offline training, which is used for online generation of phase shift matrix and receiver beamforming vectors with a lower complexity. As traditional self-supervised neural networks cannot achieve a good solution to the max-min problem, we design a multiple-stage booster (MSB) framework to solve this problem. Simulations demonstrated that SOP is improved significantly with the proposed schemes compared to benchmark schemes. In particular, the AO scheme outperforms the DL-based approach slightly at the cost of a relatively high computational complexity.

Aerial intelligent reflecting surface (AIRS), integrating unmanned aerial vehicle (UAV) with IRS, has emerged as a promising paradigm to improve the transmission quality and security in emergency communication, space-air-ground-integrated network and mobile edge computing, etc. However, the size of a single AIRS is constrained by the limited energy and payload capacity of the UAV, as well as the path loss of the air-to-ground reflective link, which makes the gain from a single AIRS finite. To address these problems, we propose a hybrid aerial IRS-assisted cooperative secure transmission system, where an aerial active IRS and an aerial simultaneously transmitting and reflecting IRS (STAR-IRS) are employed to achieve reflection amplification and 360-degree ubiquitous coverage, respectively. Additionally, the cooperative beamforming gain generated by the secondary reflection between the hybrid AIRSs can further improve communication quality. Specifically, an optimization problem is proposed with the objective of maximizing the sum secrecy rate by jointly optimizing the transmitting beamforming and the reflection coefficients of each AIRS. We first reformulated the original non-convex problem by fractional programming method, and a three-layer alternating optimization algorithm is introduced to address the proposed problem with the successive convex approximation (SCA) as well as penalty convex-concave procedure (PCCP) techniques. Finally, extensive simulations are conducted to demonstrate that the proposed scheme substantially improves the sum secrecy rate compared to other baseline schemes.

With the development of low-altitude network, cross-domain collaboration between unmanned aerial vehicle (UAV) and ground station is becoming increasingly important. To ensure the legitimacy of identity, it is imperative to design a cross-domain authentication scheme. However, current cross-domain authentication schemes still face challenges such as low computational efficiency and high storage overhead, and still have some security vulnerabilities. This paper proposes a cross-domain authentication and key agreement scheme for UAVs, which primarily employs a combination of symmetric encryption and hash functions. By holding cross-domain tokens, UAVs can directly engage in cross-domain authentication, ultimately achieving key agreement with ground station. We conduct security and performance analyses, demonstrating the feasibility of our scheme in resource-constrained low-altitude network.

In the evolving landscape of sequential recommendation systems, this paper propels the frontier forward with the introduction of the knowledge graph enhanced multi-task learning (KGML) model. At its core, KGML harnesses the capability of big data analytics, enabling a nuanced understanding of both the immediate and enduring interests of users. This is achieved through the integration of item knowledge graphs and multitask learning, which are meticulously enriched with big data insights, thereby ensuring a comprehensive representation of item attributes and interconnections. Such a method not only elevates the model’s precision in tailoring recommendations for less popular items with limited data but also effectively counters the "Matthew Effect", where visibility becomes disproportionately skewed towards already popular items. Through rigorous validation across three public datasets, the KGML model demonstrates that the proposed approach significantly enhances the accuracy of sequential recommendations.

This paper introduces defensive adversarial attacks aimed at enhancing the security of semantic communication systems by confusing potential eavesdroppers. Existing research predominantly focuses on enhancing the accuracy of semantic communications while neglecting the security vulnerabilities posed by eavesdroppers. In this study, from the standpoint of physical layer security, defensive adversarial attacks are employed to introduce artificial noise into semantic communications, effectively concealing real information. This artificial noise is generated by deep neural networks to mislead eavesdroppers into perceiving the content of images as unrelated information, with little probability of disrupting normal semantic communications. Experimental results demonstrate that the proposed model can selectively mislead the decoding efforts of eavesdroppers, while ensuring uninterrupted decoding by legitimate receivers.

This paper delves into the realm of intelligent and cooperative computing offloading within satellite-assisted Inter-net of Vehicles (Sat-IoVs). More specifically, it focuses on enabling efficient computing offloading for highly mobile vehicle users by formulating and executing an offloading path selection and multidimensional resource management (OPS-MDRM) optimization problem at a central controller. Given the complex amalgamation of continuous and discrete action spaces, along with various timescales inherent to the OPS-MDRM problem, we introduce a two-timescale framework. In this framework, we present a hierarchical hybrid policy optimization (HHPO) based on-policy algorithm to effectively tackle the aforementioned problem. Our comparative analysis against three traditional resource allocation methods underscores the outstanding performance achieved by the HHPO-based approach in the Sat-IoV networks.

Differential-privacy federated learning (DP-FL) has emerged as a promising approach to mitigate the inherent risks associated with traditional FL architectures, which are susceptible to inferential attacks due to the continuous sharing and updating of model parameters. However, existing DP-FL frameworks typically assume that the perturbations introduced by clients remain constant throughout the FL process, overlooking the dynamic influence of these perturbations on model performance across different communication rounds. In this paper, we present a long-term privacy-preserving FL framework designed to address issues of optimal incentive design, considering the dynamic influence of perturbations on model performance. Specifically, we first analyze the effect of local perturbations on the model’s convergence performance during various communication rounds, elucidating the balance between learning performance and privacy loss. Then, to harmonize learning performance with privacy loss, we develop a long-term privacy-preserving incentive scheme, where the interactions between clients and the FL server throughout the FL process are modeled as a multi-stage privacy-preserving game. Furthermore, we utilize contract theory to derive the equilibrium of this game. Finally, simulations show that our scheme can incentivize clients to contribute high-quality models, thereby enhancing the accuracy of the global model, as compared to benchmarks.

Crowd computing in large-scale cellular networks typically involves a significant number of participants, leading to high spectrum interference, reduced communication efficiency, and low user trustworthiness. To overcome these challenges, this paper proposes trustworthy and spectrum-efficient crowd computing scheme in massive multiple-input multiple-output (MIMO) networks based on deep neural networks (DNNs). Existing machine learning-aided multiple antenna technologies usually ignore the pilot contamination, which reduces spectrum efficiency. Here, we leverage the DNN to devise detection and precoding algorithms by inputting an imperfect channel state information (CSI) big data and provides detection and precoding matrices as outputs, where the online-to-offline learning framework offloads the training task to servers to reduce the overhead of base station (BS). With the well-trained DNNs, the BS can generate the detection and precoding matrices with low computation overheads. Especially, minimum-mean-square-error (MMSE) triggers between received signals and sources considering channel estimation error are seen as labels to improve spectrum efficiency. Besides, a multi-factor trust model is designed to enhance user authentication security. The simulations and numerical analysis show the proposed scheme can provide higher spectral efficiency, compared to conventional methods.

The fusion of federated learning (FL) and unmanned aerial vehicles (UAVs) garnered significant attention as a propitious paradigm, enabling the provision of ubiquitous Artificial Intelligence (AI) services in a privacy-preserving manner. However, despite the intrinsic superiority of FL in safeguarding privacy, an attacker could utilize differential attacks to infer the original data of UAVs. To address the aforementioned challenges, we design a privacy-preserving incentive scheme for UAV-aided FL. In particular, a UAV-aided FL framework is first proposed to facilitate AI model training between UAVs and the server. Then, we quantify the privacy level of UAVs based on differential privacy and analyze its influence on the aggregation accuracy of the server. This scenario involves a complex trade-off between two conflicting objectives. On the one hand, the server desires to obtain higher quality local models for superior aggregation accuracy. On the other hand, UAVs prefer to add more noise to their local models for better privacy protection. Besides, by employing contract theory, we propose an incentive mechanism to optimize the server's aggregation accuracy under information asymmetry. Finally, simulation results validate the superiority and feasibility of our proposed scheme.

The rapid development of fog-assisted vehicular crowdsensing systems (FVCSs) enables real-time vehicular data sharing, but redundant and similar data in report results in unnecessary costs. However, previous studies only focus on duplicate reports and neglect deduplication of similar data. Besides, transmitting crowdsensing data in Internet of Vehicles (IoV) exposes vulnerabilities to offline brute-force and fake report attacks. In this article, we present P2SimiDedup, a scheme for secure deduplication of similar crowdsensing reports. Specifically, we develop cryptographic primitives and introduce an improved generalized deduplication technique (GreedyGD) to achieve secure deduplication over similar crowdsensing data. Then, we construct a two-level deduplication framework that can perform secure and efficient similar-based deduplication at fog nodes and cloud server. Besides, P2SimiDedup can ensure that only data requesters can decrypt and recover crowdsensing data. The security analysis and evaluation results demonstrate that P2SimiDedup can achieve privacy-preserving deduplication for similar crowdsensing reports with moderate computational, communication, and storage costs.

This paper investigates an intelligent reflecting surface (IRS) aided green multiple-user downlink communication system. In contrast to the existing works that deploy the IRS in a fixed location, the location of the IRS is taken as an optimization variable to minimize the total transmit power by jointly optimizing the location of the IRS, transmit beamformers at the base station (BS), and IRS phase shifts. We point out a critical conclusion that before and after IRS deployment, the channel state information (CSI) of all the communication terminals is different, so an offline-online hybrid-CSI optimization framework is proposed to solve the problem. In the offline stage, we optimize the IRS location with only the statistical CSI (S-CSI) so the ergodic quality of service (QoS) constraints have to be considered, and universal lower bounds associated only with the location variable are derived to decouple all variables. In the online stage, all the instantaneous-CSI (I-CSI) are available. To solve this non-convex problem, an alternating optimization framework is developed. We propose a Riemannian Manifold (RM) algorithm to optimize the IRS phase shifts. Simulation results validate that the proposed algorithm is convergent and effective, and show that the location deployment of IRS is crucial for green communication.

Crowdsourcing has gained many developments and wide applications in our daily life. Traditional centralized crowdsourcing systems suffer from high management costs and low efficiency. The recent advance in the blockchain technology has enabled the construction of decentralized crowdsourcing systems, which can overcome the limitations of centralized systems and make crowdsourcing solution reuse possible. However, such systems also bring new security and privacy challenges. For instance, transactions on blockchain are publicly visible which can lead to privacy leakage of crowdsourcing users. Moreover, unfair exchange is a critical issue on these platforms. In this paper, we propose a privacy-preserving and fair crowdsourcing framework with fine-grained reuse based on blockchain to meet the security requirements for decentralized crowdsourcing. Specifically, we construct one-address-only (OAO) authentication to ensure the uniqueness of the participant’s address in the crowdsourcing process. Additionally, We design a submit-then-open method with commitments to resist the “free-riding" and “false-reporting" attacks. Thus, fair exchange between entities can be guaranteed. To ensure data confidentiality and fine-grained solution item reuse, we employ pairing-based cryptography to generate an encryption key and ensure flexible authorization reuse. We also adopt stealth authorization techniques to ensure privacy-preserving access authorization during the reuse phase. Finally, security analysis and implementation results have shown that the proposed framework can effectively achieve privacy-preserving and fair crowdsourcing as well as fine-grained crowdsourcing reuse. Specifically, the gas consumption in the reuse phase is reduced by approximately 49% to 81% compared to the normal operation, which significantly improves the efficiency of blockchain applications.

Application (App)-assisted mobile edge crowd- sensing is a promising paradigm, in which Apps are in charge of tagging the location of the sensing tasks as point-of-interest (PoI) to assist the platform in recruiting users to participate in the sensing tasks. However, there exist potential security, incentive, and privacy threats for App-assisted mobile edge crowdsensing (AMECS) due to the presence of malicious Apps, the low-quality shared sensing data, and the vulnerability of wireless communication. Therefore, we propose a differential privacy-based incentive (DPI) scheme for AMECS to provide secure and efficient crowdsensing services while protecting users’ privacy. Specifically, we first propose an App quality management mechanism to correlate the behavior of each App with its quality and then select reliable Apps based on quality thresholds to assist the platform in recruiting users. With the designed mechanism, we further present an auction game-based incentive mechanism to encourage Apps to mark the location of the sensing tasks as PoI. To protect the privacy of users, a privacy-preserving sensing data sharing algorithm is devised based on differential privacy. Further, given the difficulty of obtaining accurate network parameters in practice, a reinforcement learning-based incentive mechanism is designed to encourage users to participate in sensing tasks. Finally, simulation results and security analysis demonstrate that the proposed scheme can effectively improve the utilities of users, ensure the security of the crowdsensing process, and protect the privacy of users.

Skeleton-based action recognition task has been widely studied in recent years. Currently, the most popular researches use graph convolutional network (GCN) to solve this task by modeling human joints data as spatio-temporal graph. However, a large number of long-term temporal motion relationships cannot be effectively captured by GCN. Thus, recurrent neural network (RNN) is introduced to solve this defect. In this work, we propose a model namely graph convolutional network with long time memory (GCN-LTM). Specifically, there are two task streams in our proposed model: GCN stream and RNN stream, respectively. The GCN stream aims to capture the spatial motion relationships as well as the RNN stream focuses on extracting the long-term temporal patterns. In addition, we introduce the contrastive learning strategy to better facilitate feature learning between these two streams. The multiple ablation experiments have verified the feasibility of our proposed model. Numerous experiments show that the proposed model is superior to the current state-of-the-art method under two large-scale datasets including NTU-RGBD and NTU-RGBD-120.

This paper investigates the power efficiency of physical layer security (PLS) in intelligent reflecting surface (IRS)-assisted multi-user uplink channels. Existing research works usually focus on enhancing secrecy performance, and neglect measures to improve power efficiency. In this paper, the optimization problem is formulated to minimize the sum radio frequency (RF) power of multiple users in the uplink channel subject to secrecy outage probability constraint. This problem is solved by an alternating optimization (AO) algorithm that includes three optimization sub-problems, i.e., phase shift matrix, receiving matrix, and RF power optimization. Furthermore, to reduce the complexity of the proposed AO algorithm, a deep learning (DL)-based approach is proposed to optimize the sophisticated phase shift matrix optimization process. Simulation results demonstrate that the proposed scheme can significantly reduce the average RF power, and the DL-based scheme achieves similar performance as AO algorithm while reducing the time complexity significantly.

Cross-silo federated learning, as a distributed learning paradigm, allows clients to collaboratively train an artificial intelligence (AI) model and jointly share the model ownership without local data transfer or exposure. However, the valuable AI models are facing fatal intellectual property (IP) infringement threats when offering AI services. Existing researches on IP protection mainly focus on the centralized models (i.e., single ownership), but leave federated models (i.e., shared ownership) unexplored. In this paper, we propose IPSF, a novel shared IP protection framework with all-round verification for multiple owners under cross-silo federated learning. Specifically, instead of embedding private watermarks individually, we adopt joint watermarks and soft labels as a conjoint fingerprint, and present a watermark generative adversarial network (WM-GAN) mechanism to fuse private watermarks and facilitate the integrated verification. We also design a diversity-and similarity-oriented assessment mechanism to support mutual evaluation between private and joint watermarks. Through the designed assessment mechanism, the correlation and variability between private and joint watermarks are dynamically maintained to ensure the stability of WM-GAN and the fairness among users in verification. Extensive experiments validates that our IPSF achieves desirable fidelity and high robustness under attacks.

To prevent implicit privacy disclosure in sharing gradients among data owners (DOs) under federated learning (FL), differential privacy (DP) and its variants have become a common practice to offer formal privacy guarantees with low overheads. However, individual DOs generally tend to inject larger DP noises for stronger privacy provisions (which entails severe degradation of model utility), while the curator (i.e., aggregation server) aims to minimize the overall effect of added random noises for satisfactory model performance. To address this conflicting goal, we propose a novel dynamic privacy pricing (DyPP) game which allows DOs to sell individual privacy (by lowering the scale of locally added DP noise) for differentiated economic compensations (offered by the curator), thereby enhancing FL model utility. Considering multi-dimensional information asymmetry among players (e.g., DO's data distribution and privacy preference, and curator's maximum affordable payment) as well as their varying private information in distinct FL tasks, it is hard to directly attain the Nash equilibrium of the mixed-strategy DyPP game. Alternatively, we devise a fast reinforcement learning algorithm with two layers to quickly learn the optimal mixed noise-saving strategy of DOs and the optimal mixed pricing strategy of the curator without prior knowledge of players' private information. Experiments on real datasets validate the feasibility and effectiveness of the proposed scheme in terms of faster convergence speed and enhanced FL model utility with lower payment costs.

The emergence of the Metaverse has brought forth a new era of social networks, offering immersive virtual spaces for users to engage in social activities. However, the resource-intensive nature of rendering avatars and virtual scenes places considerable strain on end devices. To improve the Quality of Experience (QoE) for users, the utilization of edge servers’ resources becomes crucial. Moreover, accommodating the diverse QoE requirements and time dynamics of users (e.g., user join/departure, and social activities) escalates the complexity of resource allocation. In this paper, we propose an auction-based dynamic resource allocation algorithm to efficiently and economically allocate various limited resources (e.g., CPU, GPU, RAM, and VRAM) of edge servers to social user groups in a rapid and decentralized manner. First, with heterogeneous and dynamic varying resources at each Planet (i.e., edge server to host Metaverse users), we design an optimal Planet access scheme to help social user groups to determine which Planet to connect. Second, considering the dynamic nature of social applications, e.g., users dynamically join and depart the network with dynamic requirements on resources, we present a multi-round auction game between social user groups and edge servers to compete for the dynamic multi-dimensional resources before each scheduled time period. By using the above mechanisms, our scheme optimizes the dynamic resource utilization by considering the social feature of Metaverse. Using extensive simulations, we demonstrate that the proposed algorithm dynamically and effectively allocates resources for social Metaverse activities, outperforming conventional allocation approaches.

The paper explores the issue of multiple-user fairness of intelligent reflecting surface (IRS)-assisted physical layer security (PLS) in the digital twin (DT)-enabler power grid. Previous research works have focused on achieving secrecy rate fairness through beamforming or phase shift optimization. However, in the DT-enabler power grid, the secrecy rate is not available as the instantaneous channel state information (CSI) of the passive eavesdropper is unknown. To address these challenges, we apply an expression for secrecy outage probability, measured based on the statistical CSI of the eavesdropper for the scenario where multiple DT users are present. Using zero-forcing (ZF) precoding at the transmitter, we formulate the problem of achieving fairness in secrecy outage probability, and then solve it by optimizing the phase shift matrices. Simulation results demonstrate that the proposed methods can achieve higher fairness among users in comparison to existing IRS-assisted PLS schemes.

Connected and Autonomous Vehicles (CAVs) provide significant societal benefits but pose serious security risks due to their high connectivity and openness. Traditional security measures like cryptography and intrusion detection systems (IDSs) are reactive and passive, posing significant challenges to securing CAVs. We propose a proactive and collaborative threat-sharing framework to tackle the above challenges and enhance CAV security through vehicular honeypots. The proposed framework leverages federated learning, which allows CAVs to share threat information decentralized while preserving their privacy. Additionally, we design an optimal incentive mechanism that considers three private information of CAVs, including deployment, training, and communication costs. Specifically, we leverage the self-disclosure property of the contract theory, which can effectively address information asymmetry and incentive mismatches between CAVs and the IDS server, motivating CAVs to participate in threat sharing. Finally, through a series of simu- lation experiments, we validate the feasibility of the contract and evaluate the effectiveness of our proposed incentive mechanism.

This article investigates computation offloading in blockchain-empowered Internet of Things (IoT), where the task data uploading link from sensors to a base station (BS) is protected by intelligent reflecting surface (IRS)-assisted physical-layer security (PLS). After receiving task data, the BS allocates computational resources provided by mobile-edge computing (MEC) servers to help sensors perform tasks. Existing blockchain-based computation offloading schemes usually focus on network performance improvements, such as energy consumption minimization (ECM) or latency minimization, and neglect the Gas fee for computation offloading, resulting in the dissatisfaction of high Gas providers. Also, the secrecy rate during the data uploading process cannot be measured by a steady value because of the time-varying characteristics of IRS-based wireless channels, thereby computational resources allocation with a secrecy rate measured before data uploading is inappropriate. In this article, we design a Gas-oriented computation offloading scheme that guarantees a low degree of dissatisfaction of sensors, while reducing energy consumption. Also, we deduce the ergodic secrecy rate of IRS-assisted PLS transmission that can represent the global secrecy performance to allocate computational resources. The simulations show that the proposed scheme has lower energy consumption compared to existing schemes and ensures that the node paying higher Gas gets stronger computational resources.

By interacting, synchronizing, and cooperating with its physical counterpart in real time, digital twin (DT) is promised to promote an intelligent, predictive, and optimized modern city. Via interconnecting massive physical entities and their virtual twins with inter-twin and intra-twin communications, the Internet of DTs (IoDT) enables free data exchange, dynamic mission cooperation, and efficient information aggregation for composite insights across vast physical/virtual entities. However, as IoDT incorporates various cutting-edge technologies to spawn the new ecology, severe known/unknown security flaws, and privacy invasions of IoDT hinder its wide deployment. Besides, the intrinsic characteristics of IoDT, such as decentralized structure, information-centric routing, and semantic communications, entail critical challenges for security service provisioning in IoDT. To this end, this article presents an in-depth review of the IoDT with respect to system architecture, enabling technologies, and security/privacy issues. Specifically, we first explore a novel distributed IoDT architecture with cyber–physical interactions and discuss its key characteristics and communication modes. Afterward, we investigate the taxonomy of security and privacy threats in IoDT, discuss the key research challenges, and review the state-of-the-art defense approaches. Finally, we point out the new trends and open research directions related to IoDT.

Abstract Vehicular platoons are among the most advanced driving assistance systems that may generate considerable fuel savings and increase traffic efficiency. However, communication between vehicles in a platoon is always vulnerable to eavesdropping due to the broadcast nature of wireless channels. To address this issue, we investigate security issues from the physical layer perspective for multi-hop vehicular platooning. The dynamic shielding secured transmission scheme is proposed to guarantee the confidential transmission of private information. Specifically, the neighboring vehicles can alternately act as friendly shielders, which transmit jamming signals to interfere with the eavesdroppers without knowing the channel state information, and thus the secrecy capacity would increase. The mathematical derivation of secrecy capacity is obtained for performance analysis. Meanwhile, the numerical results verify the properties, efficiency, and adaptability of the proposed scheme.

Privacy-preserving authentication is designed to protect vehicular ad-hoc networks (VANETs) from illegitimate users and fake messages while maintaining the privacy of legitimate users’ identities. However, existing authentication schemes have disadvantages such as non-transparent certificate issuance and revocation, high identity authentication and certificate revocation overhead. In this paper, we propose an efficient privacy-preserving authentication scheme with on-chain certificate management (PACM) in VANETs, where the service manager (SM) of each domain serves as a node of the blockchain to build a distributed system. Specifically, based on elliptic curve cryptography (ECC) and exclusive-OR operations, we achieve secure and lightweight mutual authentication between vehicles and roadside units (RSUs) by regularly updated pseudonyms. Then, we adopt the blockchain to record the issuance and revocation of all certificates, which makes SM’s activities transparent. Moreover, we introduce the counting garbled bloom filter (CGBF) to enable fast query and revocation of certificates. Besides, we design a non-forgeable and non-repudiable billing mechanism based on the hash chain technology. Security analysis and experimental results show that PACM achieves stronger security with less overhead.

Cloud outsourcing provides flexible storage and computation services for data users in a low cost, but it brings many security threats as the cloud server may not be fully trusted. Previous secure outsourcing solutions mostly assume that the server is honest-but-curious while the adversary model of a malicious server that may return incorrect results is rarely explored. Moreover, with the increasing popularity of verifiable computations, existing verification schemes are yet not efficient and cannot cater to different scenarios in practice. In this paper, we propose a blockchain-based verifiable search framework in the adversarial cloud outsourcing context. When outsourcing the encrypted data to the cloud or Interplanetary File System (IPFS), we also store the encrypted data index in a decentralized blockchain (i.e., Ethereum in this paper) which is public and cannot be modified. Once a user is authorized, he/she can flexibly obtain the query results and efficiently check the query integrity via the pre-deployed smart contract, without the need of the data owner being online. Moreover, for user's privacy protection, we construct a stealth authorization scheme to deliver the access authorization without any identity disclosure. Finally, theoretical analysis and performance evaluation validate the security and efficiency of our proposed framework.

This article investigates physical layer security (PLS) in reconfigurable intelligent surface (RIS)-assisted multiple-input multiple-output multiple-antenna-eavesdropper (MIMOME) channels. Existing researches ignore the problem that secrecy rate can not be calculated if the eavesdropper’s instantaneous channel state information (CSI) is unknown. Furthermore, without the secrecy rate expression, beamforming and phase shifter optimization with the purpose of PLS enhancement is not available. To address these problems, we first give the expression of secrecy outage probability for any beamforming vector and phase shifter matrix as the RIS-assisted PLS metric, which is measured based on the eavesdropper’s statistical CSI. Then, with the aid of the expression, we formulate the minimization problem of secrecy outage probability that is solved via alternately optimizing beamforming vectors and phase shift matrices. In the case of single-antenna transmitter or single-antenna legitimate receiver, the proposed alternating optimization (AO) scheme can be simplified to reduce computational complexity. Finally, it is demonstrated that the secrecy outage probability is significantly reduced with the proposed methods compared to current RIS-assisted PLS systems.

In this paper, we propose a multi-roadside unit (RSU) assisted mobile crowd computing framework for intelligently connected vehicle (ICV) networks, where vehicles within RSUs’ coverage act as workers to provide their computation and communication resources for computing resource limited vehicle user equipments (VUEs). Physical (PHY) layer security is used to secure computation task offloading and results feedback in time-varying vehicular channels. Artificial noise (AN) assisted adaptive wiretap coding is adopted to enhance the security of offloading links. With PHY security, the intended receiver can decode secret message while eavesdropper cannot. A modified exhaustive two-dimensional (2D) search algorithm is proposed to optimize transmission rate and secrecy rate in an effective secrecy throughput maximization problem, and a multi-agent twin delayed deep deterministic policy gradient algorithm (MATD3) is utilized to assign VUEs’ tasks without a central controller, where a reward function is defined according to the computing costs, including execution time, energy consumption, and price paid for computing. Finally, simulations verify the effectiveness of the proposed framework.

The maturity in information technology and new energy technologies enables participants to generate, buy, and sell energy in energy trading systems. Although applying blockchain technology to energy trading has solved some drawbacks in traditional centralized energy systems, the openness and transparency characteristics make the trading records stored on the blockchain vulnerable to data-mining attacks that may cause indispensable privacy leakage. Due to high efficiency and low overhead, noise-addition is an appropriate solution for privacy preservation. Nonetheless, recent research on noise-addition needs to generate massive accounts, which brings a certain amount of waste and inconvenience for later regulation and management. To avoid the aforementioned issues, this paper proposes a consortium blockchain-enabled scheme to ensure the privacy of data stored on the blockchain and resist linking attacks initiated by data mining algorithms. Our scheme utilizes a dynamic partition algorithm to leverage an account mapping algorithm and a virtual token algorithm. Specifically, the account mapping algorithm utilizes a dynamic account allocation method to hide the trading distribution of active users. Furthermore, the virtual token algorithm applies Laplace noise to hide the actual energy consumption of inactive users and curb excessive accounts generation. Finally, we formally demonstrate the privacy and effectiveness of our proposed scheme in security analysis and experiment evaluations.

In this article, we propose a physical layer security (PLS) scheme to secure vehicle-to-vehicle (V2V) communications, where vehicles share spectrum resources with cellular users via underlaying device-to-device (D2D) technologies. Existing PLS-assisted V2V communication schemes neglect the issue of the unknown eavesdropper's instantaneous channel state information (CSI), and the management of spectrum resource reuse-incurred interference should be enhanced. Here, we design an artificial noise (AN)-assisted beamforming scheme to protect the V2V communications that does not require the eavesdropper's instantaneous CSI. Especially, the spectrum resource reuse-incurred interference can be eliminated by detection methods. Also, we deduce the expression of secrecy outage probability as the security metric, and minimize the secrecy outage probability via power allocation. The simulations demonstrate that the proposed scheme can reduce the secrecy outage probability compared to conventional methods.

In this paper, we propose a secure computationoffloading scheme (SCOS) in intelligently connected vehicle (ICV) networks, aiming to minimize overall latency of computing via offloading part of computational tasks to nearby servers in small cell base stations (SBSs), while securing the information delivered during offloading and feedback phases via physical layer security. Existing computation offloading schemes usually neglected time-varying characteristics of channels and their corresponding secrecy rates, resulting in an inappropriate task partition ratio and a large secrecy outage probability. To address these issues, we utilize an ergodic secrecy rate to determine how many tasks are offloaded to the edge, where ergodic secrecy rate represents the average secrecy rate over all realizations in a time-varying wireless channel. Adaptive wiretap code rates are proposed with a secrecy outage constraint to match time-varying wireless channels. In addition, the proposed secure beamforming and artificial noise (AN) schemes can improve the ergodic secrecy rates of uplink and downlink channels even without eavesdropper channel state information (CSI). Numerical results demonstrate that the proposed schemes have a shorter system delay than the strategies neglecting time-varying characteristics.

This paper proposes a brain-robot interface (BRI)-based control strategy in combination with the simultaneous localization and mapping (SLAM) to achieve the navigation and control of a mobile robot in uncertain environments. The BRI is based on steady state visually evoked potentials, utilizing the multivariate synchronization index classification algorithm to analyze the human electroencephalograph (EEG) signals in such a manner that human intentions can be recognized and motion commands can be produced for the brain controlled robot. The entire system is semi-autonomous since the navigation of mobile robot is commanded by the BRI, and the low-level motion of the mobile robot is autonomous with a designed kinematic controller. By utilizing vanishing points and door plates as the environmental features, a global metric map of the environment has been built by a sequential SLAM algorithm. The main contribution of this paper is the combination of an artificial potential field (APF) and the brain signals, which builds up the relationship between the strength of EEG signals and the intensity of the potential field. Through the proposed EEG-APF method, motion commands that would plan an obstacle-free trajectory in un-structured environments, can be obtained. The entire system has been tested with eight volunteer subjects, and all subjects are able to successfully fulfill manipulating mobile robot in the experiments.

In this paper, we propose a joint spatial division and multiplexing (JSDM) beamforming based on a neighbor scheme for frequency division duplex (FDD) massive multi-input multi-output (MIMO) systems. The neighbor-based JSDM (N-JSDM) can fully utilize signal space, leading to higher spectral efficiency over the conventional JSDMs. The reason is that for the neighbor scheme, neighbors and non-neighbors are classified adaptively by the angles of departure (AoD), and the prebeamformer is designed to mitigate the non-neighbors' interference by the statistical channel state information. The effective channel matrix after the prebeamformer then becomes a band matrix, from which the downlink training length (DTL) and the channel feedback length are much smaller than the number of antennas. Moreover, an optimal prebeamformer which is proved to be able to achieve the same system capacity as the full CSI system is proposed, followed by a suboptimal prebeamformer with constrained DTL, and a DFT-based prebeamformer. On the other hand, the neighbors' interference is mitigated using the banded channel state information. Simulation results validate the good performance of the proposed N-JSDM.

To achieve the better navigation performance of a mobile robot in the unknown environments, a novel human-robot hybrid system incorporating a motor-imagery (MI)-based brain teleoperation control is presented in this paper, where a deep-learning-based active perception is developed in the simultaneous localization and mapping (SLAM) framework. Using the deep-learning-based object recognition in the red-green-blue-depth (RGB-D) data acquisition process, the designed SLAM approach can select the valid feature points effectively, and the speed of displacement tracking can be improved by combining the oriented FAST and rotated BRIEF (ORB) SLAM algorithm with the optical flow method. The global trajectory map can also be mended using graph-based nonlinear error optimization. In addition, to build the connection between human intentions and the robot control commands flexibly in the developed mobile robot, a common spatial pattern (CSP)-based support vector machine (SVM) classification algorithm is proposed so that the control commands can be obtained directly from the human electroencephalograph (EEG) signals, which are preanalyzed and classified using the phenomena of event-related synchronization/desynchronization (ERS/ERD). Experiments involving several operators have verified the effectiveness of the proposed framework in the actual unstructured environments.

Artificial noise (AN) can be used to confuse eavesdroppers in a physical layer security system. One of the main issues concerned in AN schemes is how to improve secrecy capacities. Most existing AN schemes were proposed based on an assumption that the number of transmit antennas t is larger than that of receiver antennas r, such that they can utilize all r eigen-subchannels of a multiple-output multiple-input (MIMO) system to send messages, and use remaining t - r null spaces for transmitting AN signals. These AN signals null out legitimate receivers and degrade eavesdropper channels. However, transmitting messages in all eigen-subchannels is not always a good strategy. In particular, when the number of transmit antennas is constrained or even smaller than those of receivers, the secrecy capacities of legitimate receivers will be impaired significantly if using all eigen-subchannels for message transmission. To improve secrecy capacity, we propose an AN scheme where messages are encoded in s (which is a variable) strongest eigen-subchannels based on ordered eigenvalues of Wishart matrices, while AN signals are generated in remaining t - s spaces. We derive the average secrecy capacity of a single-user MIMO wiretap channel in the presence of an eavesdropper with multiple antennas. We show that the numerical results are in a good agreement with simulation results. The secrecy capacity of the proposed AN scheme can be improved by approximately 20% ~ 40% if compared with existing AN schemes.

Abstract In this paper, physical layer security is investigated in the dual‐hop amplify‐and‐forward cognitive relay network with one secondary source, one secondary destination, multiple cognitive relays under the presence of multiple primary receivers and eavesdroppers which can be either primary receivers or secondary receivers. The power allocation method at secondary source and artificial noise at relays are utilized to secure the secondary source‐destination transmission. Two optimization problems, namely, to maximize the received signal‐to‐interference‐and‐noise ratio of secondary source in the lightly‐loaded relay cluster situation, and to minimize the relay cluster total power in the fully‐loaded relay cluster situation, are formulated. In addition, these two problems should ensure both physical‐layer security and lower interference temperature. The semi‐definite relaxation technique is used to solve the considered optimization problems. In the second problems, we further optimize the performance with the help of the bisection method. Complexity analysis shows that our proposed method is efficient and also can be solved in polynomial time. Theoretical analysis and the Monte‐Carlo simulation results validate the proposed method. Copyright © 2016 John Wiley & Sons, Ltd.

Chang$^{\prime}$E-2 (CE-2) lunar orbiter is the second robotic orbiter in the Chinese Lunar Exploration Program. The charge-coupled-device (CCD) camera equipped on the CE-2 orbiter acquired stereo images with a resolution of less than 10 m and global coverage. High-precision topographic mapping with CE-2 CCD stereo imagery is of great importance for scientific research, as well as for the landing preparation and surface operation of the incoming Chang$^{\prime}$E-3 lunar rover. Uncertainties in both the interior orientation (IO) model and exterior orientation (EO) parameters of the CE-2 CCD camera can affect mapping accuracy. In this paper, a self-calibration bundle adjustment method is proposed to eliminate these effects by adding several parameters into the IO model and fitting EO parameters using a third-order polynomial. The additional IO parameters and the EO polynomial coefficients are solved as unknowns along with ground points in the adjustment process. A series of strategies is adopted to ensure the robustness and reliability of the solution. Experimental results using images from two adjacent tracks indicated that this method effectively reduced the inconsistencies in the image space from approximately 20 pixels to subpixel. Topographic profiles generated using unadjusted and adjusted CE-2 data were compared with Lunar Orbiter Laser Altimeter data. These comparisons indicated that the local topographies generated after bundle adjustments, which reduced elevation differences by 9–10 m, were more consistent with LOLA data.