This paper introduces a novel secure split federated semantic learning (SFsL) framework to facilitate the maintenance and evolution of digital twin networks (DTNs). Efficiently updating and evolving DTNs generally involves several critical processes: semantic extraction and transmission for physical-to-virtual synchronization, virtual model transformation and verification, and ensuring the security and privacy of physical entity data. While conventional semantic communication frameworks can effectively address semantic extraction and transmission, the complexities of virtual model transformation, verification, and data security demand a more comprehensive approach. To address these challenges, the proposed SFsL framework integrates split federated learning with task-oriented secure semantic communication schemes. In addition, it incorporates a token-based semantic defence method to distinguish between adversarial and authentic semantic data and an asynchronous secure model aggregation mechanism to enhance data-sharing efficiency. The system reliability is then formulated as a stochastic optimization problem, aiming to minimize cost complexity while maintaining high accuracy during periodic model aggregation. Evaluation results, obtained using performance metrics such as privacy loss, experienced loss, accuracy, cost and reliability, demonstrate that the SFsL framework outperforms other commonly adopted security and privacy schemes, offering improved efficiency towards the maintenance and evolution of such dynamic systems. This highlights the capability of SFsL to enable adaptive, efficient and reliable network evolutions when deployed in practical DTNs with dynamic resource constraints.

With the advent of 6G networks, digital twin (DT) systems have become critical for real-time monitoring, decisionmaking, and control across various sectors. A digital twin is a virtual representation of a physical twin (PT), enabling continuous interaction and data exchange. However, real-time communication between the DT and PT incurs variable costs as users change locations, significantly affecting system efficiency and user experience. In mobile environments, minimizing these communication costs is essential for maintaining DT performance, as increased latency and resource demands arise with user mobility. Consequently, an optimal dynamic placement strategy for DTs on edge servers is crucial to reducing communication overhead while ensuring responsiveness. This work introduces an optimization framework leveraging Lyapunov optimization to model and minimize communication costs between the Human digital twin (HDT) and PT, considering task drops during twin migration. The proposed solution dynamically adapts to user movements and network conditions, ensuring efficient real-time interactions with minimal costs. Evaluation results demonstrate the effectiveness of our approach in reducing communication costs and task drops while maintaining data exchange quality and reliability in 6G-enabled DT systems.

Maintaining robust physical-to-virtual twin connectivity remains a critical challenge in the deployment of highfidelity digital twin networks (DTNs). To ensure that each virtual twin (VT) in such DTNs is a true replica of its counterpart physical twin (PT), a reliable and cost-effective connectivity framework is essential, especially considering the potential mobility of PTs in the physical environment. This paper thus presents a new multi-modal PT trajectory prediction-enabled multi-connectivity framework where the counterpart VT model is proactively duplicated along multiple predicted paths to compensate for the effects of prediction errors, thereby avoiding communication loss between such PT-VT pair. While such a proposed scheme can improve the overall reliability of the network, the resulting connectivity solution may suffer from network congestion issues, complicating the overall cost. To address this, we analyze the trade-offs between overall reliability and cost and formulate an optimal VT placement problem aimed at minimizing delay, longterm connectivity cost, and prediction errors. We then propose a deep reinforcement learning-based adaptive VT placement strategy for cost-effective, mobility-aware PT-VT connectivity. Simulation results show the solution effectively meets the reliability and cost requirements of DTNs.

Maintaining continuous quality of experience (QoE) for adaptive video streaming in internet of vehicle (IoV) networks is significantly challenging due to rapid user mobility and highly dynamic network conditions. To address this, we present a network digital twin (NDT)-enhanced QoE optimization scheme tailored for the unique demands of IoV environments. Our approach integrates a context-aware gated recurrent unit (GRU) within the NDT to proactively predict short-term bandwidth fluctuations and enable anticipatory bitrate adaptation. We introduce a user-centric QoE model that dynamically incorporates the real-time perceptual QoE factors derived from vehicular user digital twins (VUEDTs). Furthermore, we formulate a joint optimization problem that simultaneously determines optimal video bitrate and resource allocation to maximize user-specific QoE under high-mobility IoV scenario. This problem is modeled as a Markov decision process and solved using a proximal policy optimization learning framework. Simulation results show that our proposed scheme can achieve significantly improved user QoE when compared to other baseline schemes.

The rapid advancement of wireless and cyber-physical systems has driven a growing demand for real-time sensor data in cyber environments. While existing solutions attempt to meet these demands, achieving both high-throughput processing and robust data integrity remains a significant challenge. This paper proposes an end-to-end distributed data pipeline and blockchain-enabled framework that integrates online anomaly detection, scalable data aggregation, and secure verification to ensure reliable cyber evolution. An Apache Kafka-based streaming pipeline ingests high-velocity sensor data and employs a dynamic bucketing strategy that finalizes buckets based on data volume, elapsed time, and network gas costs. Once validated, each bucket’s canonical sensor data representation is hashed and committed on-chain for tamper-evident storage. To enhance efficiency and security, the framework supports both Merkle tree and Verkle tree cryptographic data structures for comparative analysis. Implemented on a private Ethereum-like blockchain, our system efficiently handles large-scale sensor ingestion while enabling per-record verification. By integrating real-time anomaly correction, cryptographic proof mechanisms, and on-chain commitments, our solution delivers trustworthy, verifiable sensor streams tailored to the low-latency and high-reliability needs of next-generation systems.

Human digital twin (HDT) is an innovative concept that creates digital representations of humans to support human-centric services by enabling real-time data synchronization between the physical twin (PT) and the virtual twin (VT). This PT-VT synchronization process is inherently data-intensive, requiring efficient resource management, especially in resource-constrained environments. Semantic communication has emerged as a promising alternative to traditional data-driven methods. However, single-encoder models may struggle to meet the dynamic requirements of HDT applications, particularly when resources are limited. To address these challenges, this paper introduces a multi-encoder semantic communication model that dynamically allocates resources—such as bandwidth and computational frequency—on specific application demands. The framework is formulated as a mixed-integer nonlinear programming (MINLP) problem and is solved using a genetic algorithm (GA) based approach. Simulation results demonstrate that the proposed model optimizes synchronization while effectively managing power consumption, outperforming traditional single-encoder models in terms of accuracy, latency, and resource efficiency. This dynamic multi-encoder approach offers a scalable and adaptable solution for future HDT applications.

Artificial intelligence-generated content (AIGC) has been proposed as a solution to meet the requirements of ultra-reliable, secure, and privacy-preserving connectivity in human digital twin (HDT) networks. In such an AIGC-enhanced HDT, contents representing the true statuses of physical twins are generated in the virtual environment for the immediate update and evolution of the corresponding virtual twins (VTs). However, adopting a distributed AIGC in HDT presents several challenges, including the need for personalized VTs, data privacy concerns, and insufficient contextual understanding. This paper introduces a multi-layer federated semantic learning framework to address these challenges, incorporating batch learning to meet the training requirements for semantic-channel encoders and decoders. Furthermore, we introduce a novel user association framework to maximize the overall system performance under shard formation constraints. We then formulate a long-term joint optimization problem for user selection over finite learning periods. A novel Lyapunov-based online optimization strategy was proposed to mitigate the impact of time-varying and unpredictable training conditions. Additionally, we introduce a multi-arm bandit-based method and a context-centric user selection approach to solve the optimization problem. The results demonstrate that the proposed user association framework addresses the limitations of existing approaches, thereby improving the overall performance of the multi-shard AIGC-enhanced HDT.

With the rapid growth of internet of things (IoT) devices, edge computing has emerged as a crucial technology for delivering low-latency and resource-efficient services. However, the surge in edge computing capabilities poses challenges in efficiently managing result caching and deduplication to effectively utilize storage and processing resources. This study introduces a novel approach that harnesses an adaptive enhanced initiation genetic algorithm (AEIGA) for result deduplication/relocation in collaborative and distributed edge computing networks, with the goal of enhancing task offloading and result delivery. Our work proposes an optimization framework that integrates deduplication/relocation strategies to minimize redundancy, improve latency, and optimize storage across edge servers. The proposed AEIGA addresses the NP-hard nature of the formulated optimization problem by efficiently exploring the vast solution space to find optimal or near-optimal solutions. Simulation results demonstrate significant improvements of the proposed AEIGA in various network performance metrics. Our findings highlight the effectiveness of employing AEIGA for result deduplication/relocation in edge computing, offering a scalable solution to meet the escalating demands of IoT networks.

Secure data sharing is imperative in human digital twin (HDT) systems due to the continuous communication requirements among physical and virtual twins, making data security and privacy essential concerns. Previous works have emphasized the significance of blockchain technology in mitigating security challenges within digital twin systems. Nevertheless, existing blockchain-based solutions often fall short of meeting the specific latency and throughput demands of HDT systems, primarily attributed to the complicated consensus process of conventional blockchain solutions. As a result, this paper introduces a novel reputation-enhanced shard-based Byzantine fault-tolerant scheme designed for zero-trust HDT systems. We propose a parallel validation-based reputation-enhanced practical Byzantine fault tolerance consensus framework to address the need for improved throughput and reduced latency during data-sharing processes. This framework incorporates a priority-based block-appending process to prevent forking attacks, ensuring that critical aspects of the blockchain-enabled framework, such as security and decentralization, remain uncompromised. Moreover, we formalize the communication process among validators and their computation resource allocation as a Markov decision process. We then adopt the branching duelling Q-network approach to address the challenge posed by the large dimensions of the action space in our formulated problem. The results demonstrate that the proposed framework significantly enhances authentication, authorization, and validation processes in HDT through increased throughput and reduced latency, providing a robust solution for secure and efficient data sharing in HDT systems.

Due to the potential of revolutionizing a variety of human-centric services, human digital twin (HDT) is envisioned to become an important part of our daily life. The HDT applications need to frequently collect and process data obtained from individuals and their environment, analyzing each physical twin while updating its corresponding virtual twin, which will consume a large amount of computing, storage and sensing resources cumulatively. Meanwhile, running HDT applications, such as emotion recognition, naturally contains the executions of several dependent tasks. Considering the resource limitations of mobile terminals, we enable dependent task offloading to mitigate terminal load and reduce the latency of HDT applications. Specifically, this paper proposes an energy and cost-aware offloading algorithm for dependent tasks with edge-cloud collaboration to empower HDT applications. We show that the problem of dependent task offloading under constraints of service cost and terminal energy consumption is NP-hard. The complexity of task interdependency makes the offloading decision under dual constraints even more challenging. The proposed offloading algorithm firstly generates task paths based on task interdependency and computation load, deriving the initial solution. Then, task reassignment and CPU frequency scaling methods are utilized to further optimize the obtained solution. Simulation results illustrate that our approach can achieve better performance in terms of makespan and service success ratio compared to the existing approaches.

Data timeliness, privacy, and security are key enablers for data-sharing systems to support time-sensitive and mission-critical systems and applications. While blockchain-enabled data sharing frameworks can offer reliable security and privacy when properly implemented, the timeliness of data and the related latency are important issues that can limit the adoption of blockchain in large-scale mission-critical applications. This paper thus carried out a performance analysis of the blockchain-enabled data-sharing framework from latency and data age perspectives to investigate the suitability of blockchain technology in data sharing systems. To achieve this, the uniqueness of such systems such as transactions validation latency, transaction generation rate, waiting time, blockchain-appending rate, and overall communication latency were jointly studied. The communication latency was characterized following the spatiotemporal modeling approach. We further adopted the practical Byzantine fault tolerance (PBFT) consensus protocol due to its well discussed suitability in large-scale data sharing applications and captured the validation stages of such a PBFT scheme using the Erlang distribution of order$k$. Simulations results show that various influential system parameters must be carefully considered when adopting blockchain technology in time-sensitive data sharing applications. This will guide the adoption of blockchain technology in various data sharing applications and systems.

Ensuring reliable update and evolution of a virtual twin in human digital twin (HDT) systems depends on any connectivity scheme implemented between such a virtual twin and its physical counterpart. The adopted connectivity scheme must consider HDT-specific requirements including privacy, security, accuracy and the overall connectivity cost. This paper presents a new, secure, privacy-preserving and efficient human-to-virtual twin connectivity scheme for HDT by integrating three key techniques: differential privacy, federated multi-task learning and blockchain. Specifically, we adopt federated multi-task learning, a personalized learning method capable of providing higher accuracy, to capture the impact of heterogeneous environments. Next, we propose a new validation process based on the quality of trained models during the federated multi-task learning process to guarantee accurate and authorized model evolution in the virtual environment. The proposed framework accelerates the learning process without sacrificing accuracy, privacy and communication costs which, we believe, are non-negotiable requirements of HDT networks. Finally, we compare the proposed connectivity scheme with related solutions and show that the proposed scheme can enhance security, privacy and accuracy while reducing the overall connectivity cost.

The human digital twin (HDT) is a new paradigm that possesses the ability to revolutionize the current healthcare systems. With HDT, ensuring an efficient connectivity scheme between each human-virtual twin pair remains a significant problem. As the concept of HDT is new, conventional connectivity schemes cannot meet the unique requirements of HDT in terms of reliability, security and privacy. This paper thus proposes an edge-assisted connectivity scheme for HDT and adopts an integrated blockchain and federated learning techniques to ensure security and privacy. To minimize long-term average connectivity cost, we formulated the connectivity problem as a Markov decision process and adopted the deep deterministic policy gradient (DDPG) algorithm to learn the optimal connectivity policy in terms of connectivity cost. The obtained results were then compared with the conventional deep Q-network-based solution. The results show that the proposed DDPG-based connectivity solution is feasible to perform the connectivity process better by optimally allocating system resources, thus reducing the overall connectivity cost, while ensuring data security and privacy.

Blockchain and cloud-edge computing techniques are promising technologies for next-generation, secure, and privacy-preserving data-sharing systems. By integrating these technologies, the data demands of many data users, such as research institutes, hospitals, manufacturers, etc., can be met. Despite this promising integration, it is yet to be understood how the vulnerability and uncertainties of wireless communication links between the data producers, blockchain systems, cloud-edge computing-based platforms, and data users, as well as unstable validation parameters can affect the overall performance of such blockchain-enabled data-sharing systems. In this article, we considered a collaborative data-sharing scheme, where multiple data providers and data users collaborate to accomplish data-sharing tasks through the proposed blockchain and cloud-edge computing schemes. We considered the spatial distribution of data providers and data users to follow the independent homogeneous Poisson point process, while the transactions generation rate at each node was also modeled using an independent Bernoulli process. We then obtained analyses for some selected performance metrics of interest and evaluate the performance of the system. The obtained results showed that the proposed analyses can be useful in investigating the performance of any blockchain-enabled data-sharing system. This will aid successful deployments of effective data-sharing systems.

Interference modeling in cognitive radio network is important to ensure adequate coverage in the network. A reliable interference model, however, depends on accurately characterizing the distribution of users. In this paper, the dependence between primary and secondary networks is examined in order to capture more system parameters related to system characterization. Hence, two cases are considered - primary user (PU) interference control and PU with secondary user (SU) interference control mechanisms. Under PU interference control, distributions of PUs follow the Matern hard core process while the distribution of SUs follow the Poisson hole process (PHP). However, under PU with SU interference control, the distribution of active SUs follow a modified PHP. Bound and approximate expressions were derived for coverage probability at both primary and secondary networks, while simple yet accurate expressions were obtained to depict the number of simultaneous active users supported for the two cases. The tight closeness between the bound and the approximate expressions shows the reliability of the presented theoretical analysis. Furthermore, the bipolar network model assumption was relaxed while the case of independence assumption among users was also considered. Numerical result showed close tightness when the bipolar network model assumption was relaxed while the independence assumption was shown to overestimate users' coverage probability.

Integration of mobility into system characterization can complicate system analysis, as characterizing interference in mobile wireless networks is more difficult, especially when the assumption of independence of users' distributions is relaxed. However, in some practical systems, quite a number of users may be mobile, making the integration of mobility important. In this paper, mobile users' coverage in cognitive radio networks is investigated under the assumption that primary users are distributed following the Matern hard core process, while secondary users are distributed following the extended Poisson hole process. A relaying based mobility technique was proposed to sustain coverage at any tagged receiver when such receiver is located outside the coverage of its corresponding tagged transmitter. Analyses of outage probability and average throughput were derived in both primary and secondary networks and verified through simulations. The outcomes of numerical results show that the proposed mobility scheme is capable of improving coverage of mobile users in cognitive networks.

Malicious users (MUs) have the tendency to disrupt the activities of honest users in the network if not properly controlled. In a massive cognitive radio network (CRN) with priority queues, malicious secondary users (SUs) can manipulate their priority queue requirements and mislead legitimate SUs to vacate the channels. In this paper, a game theoretic based signal detection approach is proposed to control the presence of MUs in CRN. If the received signal strength is less than the predefined threshold for primary transmissions in the presence of interference and noise, such a user is marked to be malicious and its payoff table is updated. Through the mixed strategy Nash equilibrium method, the payoff table of each user can be updated to aid removal of MUs from the network. The outcome of the simulation results shows that such an approach can reduce the impact of malicious activities in the massive CRN where SUs are expected to be low-power energy-efficient devices.