Lattice-based group signatures are pivotal in the post-quantum era, with their fully dynamic variants receiving significant attention for aligning well with the dynamic application scenarios of the Internet of Things. However, current dynamic implementations, such as verifier-local revocation and dynamic accumulators, suffer from efficiency drawbacks. Verifier-local revocation incurs additional zero-knowledge proofs to demonstrate that a user has not been revoked, increasing the computational burden on IoT devices. On the flip side, dynamic accumulators, which rely on Merkle trees (or binary trees), require recursive updates to parent and sibling nodes when IoT devices are replaced. To address these limitations, this paper proposes a novel fully dynamic group signature scheme from lattices for IoT, which utilizes the zero-knowledge proof of range membership to support dynamic revocation. We first present a zero-knowledge proof for inequality and interval relations based on a more efficient commitment scheme. Leveraging this protocol, the division of one universal group into multiple groups, and flexible interval variations, we achieve full dynamic functionality, which well accommodates the frequent changes of IoT devices. Subsequently, we provide security proofs for anonymity and traceability using a sequence of games. Finally, we implement the proposed scheme programmatically and analyze its overhead and time cost. The results demonstrate that it outperforms other fully dynamic schemes, particularly in terms of overhead and revocation time, making it more suitable for resource-constrained IoT devices.

With the rapid development of artificial intelligence and new energy technologies, intelligent driving has become a core component of modern transportation infrastructure. The open nature and diverse connectivity of vehicular networks raise two coupled requirements: privacy-preserving collaboration across competing vehicle fleets and freedom from single-point-of-failure risks in the aggregation layer. Existing federated learning solutions that rely on a central aggregator and on lattice-based or Paillier-style homomorphic encryption struggle to meet the real-time budget of V2X communication. This paper proposes a distributed homomorphic-encryption federated learning algorithm for cross-trust-domain collaboration in vehicular networks. The algorithm combines three components: a vehicle clustering and ring-formation scheme based on graph attention networks, a lightweight obfuscated modular decomposition homomorphic encryption scheme, and a ring-topology distributed aggregation protocol. The proposed scheme achieves 87.28% test accuracy on CIFAR-10, exceeding the no-encryption FedAvg baseline of 86.78%, and retains this advantage under a strict non-IID partition with Dirichlet $\alpha \!=\!0.3$. Per-round encryption is 7.6 times faster than CKKS-FL and produces 4.4 times less ciphertext. Against gradient-inversion attacks, the aligned structural similarity index drops to 0.027 on CIFAR-10, an order of magnitude below unprotected FedAvg, and remains an order of magnitude below the baseline even when up to three ring members collude. The network-layer simulation confirms that the ring topology meets the V2X channel coherence time with controlled latency and throughput overhead.

As an important part of the intelligent transportation system, the Internet of Vehicles (IoV) optimizes traffic management, improves road safety, and travel experience by connecting vehicles with the cloud, infrastructure, pedestrians, and other vehicles. However, with the rapid development and widespread application of IoV technology, its network security issues have become increasingly prominent. Its attack threats may not only cause the functional failure of individual vehicles, but also seriously impact the stability of the entire transportation system. With the expansion of scale and increase in complexity, new threats and large-scale communications have put forward higher requirements for the security protection capabilities of IoV. Therefore, the introduction of adaptive active defense mechanisms can further improve the security and adaptability of the system. Because of the diverse defense needs of different levels of IoV systems, this paper constructs an intelligent agent attack and defense game scenario, and proposes an optimal defense strategy for moving targets based on an incomplete information Bayesian Markov game and Q-learning algorithm from the perspectives of adaptive active defense, reducing resource overhead, and intelligent defense scheduling. This strategy targets different types of attack problems existing at the vehicle side, pipe side, and cloud side of the IoV system. Taking into account four complex attack scenarios, namely scanning attack, false message injection attack, denial of service attack (DoS), and malware attack, this strategy establishes an incomplete information Bayesian Markov game model and adopts the reinforcement learning Q-learning algorithm to dynamically adjust the game environment strategy by continuously interacting with the environment to obtain reward values. This strategy seeks the most suitable defense elements and the optimal mutation time of defense elements for network attacks at all levels of the IoV, solves the problem of intelligent scheduling of defense resources under active defense systems at different levels, and avoids the limitations of a single defense mechanism. The experimental results show that compared with the existing comparative algorithms, the proposed method exhibits better defense performance and limitations in dealing with multi-level compound attacks on the IoV, and provides a new technical approach for building an intelligent and active IoV security protection system. This paper uses OMNeT++, SUMO, and Veins network joint simulation tools to verify the effectiveness of the above two solutions through four performance indicators: attack success rate, data packet loss rate, communication delay, and system overhead. The simulation results show that the defense mechanism performs well in reducing attack success rate, data packet loss rate, communication delay, and system overhead, and the overall defense effect is significant, especially when dealing with complex and changing attack scenarios.

With the rapid expansion of large-bandwidth grid services in recent years, the efficient resource allocation has become a critical challenge. This article explores an optimized resource allocation strategy that integrates satellite communication technology to ensure efficient and stable grid communication services in high-bandwidth scenarios. The main contributions of this study are as follows: we propose a two-stage prediction-allocation closed-loop framework for high-throughput satellite (HTS)-enabled smart-grid communications. The framework comprises an attention-based traffic-prediction module and a dynamic bandwidth-allocation module, which, respectively, provide accurate forecasts of future node traffic and priority-aware multibeam bandwidth optimization, thereby offering end-to-end decision support for satellite resource scheduling. Experimental results show that the proposed scheme enhances the grid's adaptability to future traffic variations at communication nodes, addresses bandwidth provisioning under uncertain high-bandwidth conditions, improves priority-aligned bandwidth utilization and priority efficiency (PE), and ensures both the stability of large-bandwidth grid communications and the performance of critical services.

In intelligent connected vehicles (ICV) systems, driving users (DUs) share data with service providers (SPs) to receive personalized services. While SPs facilitate these services, they face critical challenges related to data breaches and trust crises. SPs can invest in privacy protection to mitigate these risks. However, without financial incentives, they lack the motivation to allocate resources for such investments. Meanwhile, numerous heterogeneous DUs pose a computational challenge for SPs in determining optimal privacy protection investments. This paper investigates SPs' privacy protection investment strategies for data breach losses under the shared liability mechanism. We construct a time-varying bilevel Stackelberg mean-field game (TB-SMFG) model to capture the dynamic interactions between SPs and DUs. The paper incorporates the average behavior of DUs to mitigate the curse of dimensionality. Specifically, SPs act as the leader, controlling privacy protection investments and the incentive price. DUs act as followers, dynamically adjusting the data-sharing ratio and the trust rating. The dynamic game participants interact through the mean-field term in the state function. By solving this model, we derive the optimal investment strategy and pricing trajectory for SPs in a dynamic environment. The effectiveness and feasibility of the proposed approach are validated through a series of numerical simulation experiments.

The explosive proliferation of large-bandwidth intelligent-grid services has dramatically intensified the demand for high-speed, reliable communications. Although satellite communication systems can deliver extensive coverage and greater capacity, existing resource-allocation strategies still struggle to cope with sudden traffic surges, stringent quality of service (QoS) requirements, and frequent satellite handovers. These shortcomings often result in poor bandwidth utilization and service interruptions, highlighting the need for more adaptive and fine-grained optimization frameworks. To address this issue, this paper investigates an optimized resource allocation scheme that integrates satellite communication technology in high-bandwidth power grid service scenarios, aiming to provide efficient and stable communication services. Building upon the application of low Earth orbit (LEO) satellites in smart grid communications, this paper presents a joint resource allocation scheme that optimizes the association between power grid communication nodes and subchannels, along with the allocation of transmission power, to satisfy the requirements of high-bandwidth power grid services. First, an optimization problem is formulated to maximize the total uplink transmission rate of power grid communication devices. Then, a genetic algorithm (GA) is employed to optimize the satellite-to-grid node association. Subsequently, a two-stage Gale-Shapley (GS) stable matching algorithm, combined with the Lagrangian duality theory, is utilized for the joint allocation of subchannel and power resources. Finally, simulation results validate the performance and convergence of the proposed algorithm. The results demonstrate that, compared to matching and Lagrangian-based algorithms, random channel allocation with average power, and the joint optimization Algorithm without minimum rate constraints, the proposed algorithm achieves a higher total system transmission rate while significantly enhancing the robustness of the power grid communication network. These findings confirm the effectiveness of the proposed method.

In intelligent connected vehicles (ICVs) system, driving users connect to service providers (SPs) to obtain location-based services (LBS). Users transmit large volumes of encrypted sensitive information related to their itineraries to SPs to access value-added services. Attackers may launch chosen-ciphertext attacks (CCA) against SPs by exploiting the malleability of homomorphic encryption. This enables adversaries to infer or steal private key information, thereby threatening the long-term privacy of user data. Furthermore, existing key management technologies in ICVs system predominantly rely on passive defense strategies and suffer from limitations such as single protection mechanisms, delayed updates, and limited adaptability. To address these issues, this paper proposes an adaptive key update security mechanism based on a differential game framework. This mechanism treats the cumulative information leakage of the private key as a contested resource to construct a differential game model. Based on the feedback Nash equilibrium (NE), the mechanism adaptively derives the optimal homomorphic private key update frequency in response to the attack frequency, thereby maximizing the defense benefit. Finally, numerical simulations validate the correctness of the proposed model and demonstrate the effectiveness of the mechanism.

ABSTRACT This paper investigates the task offloading and resource allocation problem in unmanned aerial vehicle (UAV) swarm networks, with the objective of minimizing a weighted sum of task completion latency and energy consumption. Considering the autonomous decision‐making characteristics of individual UAVs in the swarm, each UAV is modeled as an intelligent agent and classified into heterogeneous types according to its computational capability. Based on this modeling framework, a mixed‐integer nonlinear programming (MINLP) problem is formulated to jointly optimize task offloading decisions and UAV transmission power. Owing to the high computational complexity of the original problem, it is decomposed into a transmission power allocation subproblem and a task offloading subproblem, where the optimal transmission power allocation strategy is obtained via a bisection‐based method. Furthermore, to enable efficient and rational task offloading within the UAV swarm, a matching game‐based task offloading algorithm is proposed, and its stability and convergence are theoretically proven. Finally, extensive simulation results and comparisons with multiple baseline schemes demonstrate the effectiveness and superiority of the proposed approach in terms of system latency and energy efficiency.

The rapid development of the low-altitude economy has accelerated the widespread adoption of unmanned aerial vehicle (UAV) systems in various intelligent applications. However, UAV placement is fundamental to low-altitude computing services, while the limited battery and computing resources of UAVs pose a critical challenge for latency-aware offloading and resource allocation. In this paper, we propose a semantic-aware UAV-assisted edge computing architecture that extracts task-related semantic information and offloads it to edge servers, thereby enabling reliable and efficient adaptive adjustment of task offloading communication and computation. Within this network architecture, the UAV placement, task offloading, and computing resource allocation are jointly optimized to minimize the total task execution delay while satisfying the UAV resource constraints. However, the formulated problem is a large-scale mixed-integer nonlinear optimization problem, whose computational complexity increases rapidly with the number of devices. To address this challenge, the original problem is reformulated through successive convex approximation (SCA) and reformulation linearization techniques (RLT). Furthermore, a parallel optimization framework is proposed to decompose the large-scale optimization problem into multiple low-complexity subproblems that can be solved in parallel, thereby accelerating the computation process and reducing the overall computational complexity. Simulation results demonstrate that the proposed method achieves near-optimal performance with significantly lower computational complexity. Compared with existing schemes, the proposed algorithm effectively reduces task execution delay for users and ensures stable and efficient system operation.

In intelligent connected vehicles (ICVs), driver-users connect with the Internet through roadside units (RSUs). The driver-users transmit a large amount of personally sensitive information about their journeys to RSUs to access value-added services. Simultaneously, the attackers infiltrate RSUs at a certain frequency to steal or intercept some connected ICVs users' privacy information, which causes a great threat to the privacy of user data elements. Moreover, most defense technologies in current ICVs systems rely on a passive defense approach, which has limitations such as single strategy and difficult flexible adjustment. Given the above issue, this paper proposes a data element circulation security defense mechanism based on differential game theory. The mechanism conceptualizes the health level of RSUs as a contested resource and establishes a differential game model. Leveraging the Nash equilibrium solution with feedback, the mechanism adaptively determines the optimal packet marking strength (PMS) in accordance with the attack frequency. And the defenders can achieve maximization the effectiveness of their defenses. Ultimately, numerical simulations are conducted to validate the accuracy of the model and demonstrate the effectiveness of the proposed mechanism.

In intelligent connected vehicle applications, tasks, such as path planning and health management involve numerous matrix operations, particularly matrix multiplication. Due to limited resources, these tasks are often outsourced to the edge server. However, outsourcing these tasks involving matrix multiplication might incur potential risks, such as returning incorrect results to expedite processing or even exposing sensitive data during the computation. Privacy-preserving verifiable matrix multiplication schemes address these concerns. However, it is meaningful in practice only if the verification and decoding time is lower than that of local computation. In this article, we propose a privacy-preserving verifiable matrix multiplication for intelligent connected vehicles that further reduces the verification and decoding time. To achieve this, we first reduce the length of the ciphertext of linearly homomorphic encryption (LHE) when encrypting a group of messages. Subsequently, we construct our verifiable matrix multiplication scheme based on the improved LHE. It has a lower critical dimension than the state-of-the-art scheme with a similar security level, since the shorter ciphertext and the simpler LHE algorithm. Performance analysis and experimental results demonstrate that the critical dimensions of our improved scheme are reduced by 23.3%, while the communication cost is reduced by 68.3%, making it particularly suitable for intelligent connected vehicle applications.

Uncrewed aerial vehicles (UAVs) have recognized as a pivotal technology for advancing wireless augmented reality (AR) applications. However, the considerable energy requirements during the rendering process present a formidable challenge, demanding a precise balance between energy efficiency and latency. Additionally, UAV-enabled systems may face significant security risks in untrusted environments. To solve these issues, we present a secure optimization framework for AR applications, where the blockchain is integrated into the system to provide distributed management and control functions. The critical information during the AR rendering process can be recorded in blockchain promptly to enhance security and privacy. In the proposed framework, a joint optimization problem is formulated to achieve the optimal trade-off between energy consumption and content delivery latency, where the rendering decision, resource allocation, and UAV placement are jointly optimized. Due to the tight coupling variables, the optimization problem is non-convex and difficult to be tackled by adopting the traditional method. To this end, we decouple the formulated problem and design a block coordinate descent (BCD)-based optimization algorithm. In the proposed algorithm, we innovatively combine the Lagrangian multiplier iterative (LMI) method and the deep reinforcement learning (DRL) approach to enhance the solving efficiency by implanting the LMI method into the learning environment of DRL. Simulation results demonstrate that the proposed method can perform well for AR applications compared to other baseline solutions and traditional DRL approaches.

In the Internet of Vehicles (IoV), developing accurate road information models is essential for analyzing perception data gathered from multiple vehicles. However, traditional centralized data-sharing methods can compromise the privacy and security of data providers. federated learning (FL) presents a promising solution as a distributed machine learning approach that balances data privacy protection with efficient utilization by keeping data localized and sharing only model updates. Nevertheless, conventional FL strategies often fail to adequately address differences in resource investment and data quality among participating vehicles while aggregating local training results. This oversight can lead to inequitable model aggregation and distribution, reducing the motivation for vehicles to share their data. This article proposes a reputation evaluation-based, fair, and secure FL scheme for the IoV to address these challenges. In this scheme, the aggregation node utilizes fuzzy comprehensive evaluation to assess the training outcomes of participating vehicles and assigns aggregation weights accordingly. It also calculates reputation values for each vehicle using periodic averaging methods. Subsequently, the node implements differentiated global model compression and distribution based on these reputation scores. Experimental results indicate that the proposed scheme performs comparably to established algorithms while effectively evaluating vehicle reputations. It achieves model compression and equitable distribution, demonstrating an ability to identify and counteract malicious client attacks. Consequently, this approach enhances fairness and security in FL systems designed for the IoV.

Low-Earth orbit (LEO) satellite networks can achieve global network coverage without geographical restrictions and are essential to the future communication network. In this article, we study the computing offloading problem in a satellite-terrestrial integrated network for the Internet of Remote Things (IoRT), which aims to reduce the total cost (weighted sum of energy consumption and delay), and jointly offload node selection, offloading ratio, and computational resource allocation to achieve the dynamic management of network resources. First, we propose a hybrid cloud and satellite multilayer multiaccess edge computing (MEC) network architecture that can provide heterogeneous computing resources to terrestrial users. Subsequently, since the problem under consideration is a mixed-integer nonlinear programming problem, we propose a computing offloading algorithm for multiagent reinforcement learning, which is an integration of double deep Q learning (DDQN) and deep deterministic policy gradient (DDPG). The algorithm can learn the optimal policy for actions containing a mixture of discrete and continuous variables. Finally, an optimal computational resource allocation scheme is proposed to improve the task computation efficiency. Simulation results show that the proposed task offloading and resource allocation scheme can achieve reasonable scheduling of computational tasks and optimal allocation of computational resources, reducing the cost of task computation.

Data-intensive smart applications are currently driving the emergence of edge-enabled computing power network (Edge-CPN) by orchestrating the computing powers (CPs) of edge servers, enabling the converged computing and networking at the edge. Besides, the proliferation of these applications and various smart devices (SDs) is arousing great interest in the joint training of a shared global model by massive SDs via federated learning (FL). Due to the heterogeneity and constrained resources of SDs, the FL performance in the Edge-CPN suffers from the straggler effect, reducing the efficiency of global model aggregation. To tackle this challenge and upgrade the training mode into higher degrees of efficiency and intelligence, in this article, we propose the TwinFed, a novel digital twin (DT)-driven FL framework, which fully leverages ubiquitous CPs to configure the DTs for assisting model training of stragglers. An interplay between the end and edge layers is captured into the architecture design via the hierarchical model aggregation. We develop a unified twinning pipeline to achieve the high-fidelity DTs and efficient model training. A two-stage workflow is also introduced to implement TwinFed by flexibly integrating the computing resource orchestration and training process. Finally, we conduct a case study for anomaly detection in smart factory to validate the superiority of TwinFed in testing accuracy and training loss.

Motivated by addressing the problem of knowledge lack in a fusion process and overcoming the defect of insufficient data coverage, this paper proposes a new concept of generalized representation for multi-source medical data fusion to get more generalized fusion results from limited source data and precise medical diagnosis conclusions based on fusion results. The generalized representations of different feature types of data nodes are uniformly defined by their value and knowledge attributes which are reflected by their normalized representations and decision-making effect indexes. Based on the generalized representation of multi-source entity nodes, the classical and quantum inspired collaborative fusion methods based on linking topology are proposed, in which different entity subsets are formed consistently for autonomous fusion according to the horizontal correlative and vertical collaborative relationships between entity nodes. The proposed generalized representation based fusion method can produce reliable fusion results in experimental practice. Based on the reliable fusion results, the corresponding intelligent pneumonia diagnosis reaches a high accuracy of 96.02%.

Dense low-power small cell base station (SBS)-based heterogeneous wireless cellular networks (HetNets) have attracted much attention to achieving high-traffic density and peak rate performance. However, the serious energy consumption problem is still a challenge for HetNets. The use of renewable energy (RE) has been considered as one promising solution for the above problem. This article proposes an energy trading scheme among base stations in RE-based HetNets. All SBSs in HetNets are considered as either the energy demander (SBS-ED) or the energy supplier (SBS-ES) based on their abilities in producing RE, and the macro base station (MBS) works as the energy trading manager to control the trading price. A dynamic evolutionary game-based energy trading model between SBS-ES and SBS-ED is established to achieve cooperative energy trading, and the evolutionary stable strategy (ESS) of the proposed model is analyzed. The pricing mechanism of MBS is also investigated, which can effectually affect the EES performance of the proposed model. It is concluded that the MBS's strategy in the trading price can affect the energy trading strategies of the SBSs. An energy transmission model is proposed, and the minimum energy loss is considered as the goal to obtain the optimal solutions. Numerical results are given to prove the validity and correctness of our proposed method.

This paper considers a two-hop wireless powered relay network consisting of multiple sources, multiple destinations, and one relay. The relay can receive energy from the sources and forward data to the destinations. We focus on the source selection problem during the energy transfer process and the resource allocation problem during the data transmission process. Firstly, the relay can choose among all sources based on the transferred energy from the sources. A credit mechanism is introduced for the relay to achieve optimal selection. Secondly, a Stackelberg differential game-based model is adopted for the resource allocation problem in the data transmission process, using the differential equation to describe the dynamic variation of energy, and the Stackelberg game to describe the relationships between the sources and the relay. In the proposed approach, both sources and relays consider energy consumption and energy revenue. To find the optimal solutions, an adaptive dynamic programming-based algorithm is utilized. The Lyapunov-based stability analysis shows that the system has uniform ultimate boundedness and convergence. Finally, the trained neural networks can achieve optimal resource allocation strategies. Through extensive simulation experiments, the effectiveness of the proposed algorithm is verified.

The rapid increase in reliability, security, timeliness, and efficiency of power communication services puts forward higher requirements for the integrity of power grid fusion communication systems. The individual needs of different services make the switching strategy that can be automatically generated in a grid fusion communication system become the focus of research. In this paper, a Switching Strategy Generation scheme based on Improved Genetic algorithm (S2GIG) is proposed for the Space-Air-Ground smart grid. Under the premise of considering the specific communication status of the network, according to the characteristics of the service, it automatically switches to one or a combination of various power grid communication modes such as the ground cable network, the ground mobile communication network, the low-orbit satellite Internet of Things, and the Beidou satellite short message for transmission, so as to meet the individual needs of the power grid business and improve the service quality of the network. The experimental simulation verifies that the proposed algorithm can improve the transmission efficiency of power grid communication service.