The spread of infectious diseases in networked populations poses significant challenges for public health intervention strategies. Traditional centrality-based and heuristic network dismantling approaches prioritize highly connected nodes but often fail to account for individual mortality risk, limiting their effectiveness in minimizing overall fatalities. While recent advances in machine learning have improved intervention strategies, existing models largely focus on reducing disease transmission rather than directly targeting mortality outcomes. To address this gap, we propose a reinforcement learning-based framework that integrates graph representation learning to identify and remove high-risk nodes, thereby maximizing network fragmentation while minimizing overall deaths. The framework is trained using synthetic networks and evaluated on five synthetic and four real-world datasets, benchmarking its performance against state-of-the-art network dismantling methods [graph dismantling with machine learning (GDM), generalized network dismantling (GND), and graph enhanced reinforcement learning (GERL)]. Experimental results demonstrate that the proposed method consistently outperforms baseline approaches, particularly in scale-free and community-structured networks, where targeted node removal significantly weakens network connectivity and suppresses epidemic spread. Moreover, in real-world networks, the method achieves lower cumulative death rates and higher epidemic thresholds, highlighting its robustness in controlling disease propagation. By incorporating mortality risk into network representation learning, the proposed framework offers a scalable, adaptive, and socially responsible approach to epidemic mitigation, misinformation control, and network resilience enhancement.

Federated learning (FL) is a promising distributed machine learning approach with enhanced data privacy protection. However, wireless communication remains a key bottleneck, directly affecting the efficiency and performance of FL. In this paper, we introduce a device-centric cell-free network to mitigate the negative effects of random fading and limited radio resources on FL. The convergence gap, representing the difference between the FL model’s performance and that of the optimal model, is analyzed to evaluate the impact of communication and computation factors, including inter-device interference, on FL performance. Then, access point (AP)-device association, transmission power, and computation frequency are jointly optimized to minimize the convergence gap. Lyapunov techniques are employed to decouple the long-term optimization into a series of online solvable problems. A deep reinforcement learning-based scheme is proposed to optimize the AP association and transmission power for devices, reducing the computational complexity from a prohibitive level to a real-time feasible quadratic level. Additionally, a closed-form solution for the optimal device computation frequency is derived. Simulation results show that the proposed scheme significantly outperforms the traditional cell-free FL and cellular FL schemes in both model training performance and energy efficiency.

With the rapid development of the Industrial Internet of Things (IIoT), real-time monitoring and high-precision control scenarios impose stringent requirements on the freshness and reliability of short-packet communications. The age of information (AoI) and block error rate (BLER) are selected as the performance measures of freshness and reliability, respectively. The quantitative relationship model of BLER with packet length and signal-to-noise ratio and the mathematical model of average AoI under both retransmission and no-retransmission mechanisms are derived. Based on the above analysis, a multidimensional performance relationship analysis is conducted. It is proved that there exists a strong tradeoff between average AoI and BLER, leading to establish a multi-objective co-optimization problem. With the goal of minimizing average AoI and BLER simultaneously, a multi-objective optimization problem is formulated. The weighted sum method and the Non-dominated Sorting Genetic Algorithm II (NSGA-II) based Pareto optimization method are proposed for solving the problem. Simulation results demonstrate that both algorithms effectively identify optimal trade-off solutions, achieving balanced performance between AoI and BLER in short-packet communications.

The right to deletion mandates that data controllers receiving deletion requests not only erase the specified data but also notify other controllers to do the same. Designing a secure and privacy-preserving protocol for distributing deletion notifications across involved data controllers, a topic not previously addressed in the literature, presents significant challenges. In this paper, we introduce CAREFUL, a secure and privacy-preserving deletion notification distribution protocol—the first to address these challenges. It operates within a centralized architecture consisting of regulatory and service planes. The fundamental principle of CAREFUL is that the regulatory plane creates a cryptographic access control structure, ensuring that data controllers in the service plane only identify the next-hop nodes for notification. With CAREFUL, the circulation history of data pending for deletion, which is a special user’s privacy, can be preserved. Moreover, CAREFUL protects the deletion notification distribution from various malicious attacks over untrusted underlay networks. Experimental results validate CAREFUL’s practicality and its efficiency in terms of resource overhead.

We investigate straggler-relay association and ener-gy consumption minimization for cooperative relay assisted fed-erated learning (FL) over fading channels to tackle the straggler effect and limited device energy. To the best of our knowledge, this is the first work to explore joint computation-communication optimization for cooperative relay assisted FL over fading chan-nels, where some devices act as relays for stragglers. A closed-form expression for the computation frequency is derived to facilitate low-complexity straggler identification. A bandwidth sharing decode-and-forward relay scheme is proposed, which benefits both straggler and relay. The closed-form expressions for the transmission power, which minimizes the computation and communication energy of devices under global time constraints, are derived. A low-complexity joint straggler-relay association and multi-domain resources optimization (JSAMRO) algorithm is proposed. Simulation results show that the proposed JSAMRO algorithm achieves a significant performance gain in terms of device's energy consumption and availability rate over the comparison schemes.

In last decades, joint design of user scheduling and precoding has been investigated for single transmission time interval (TTI). However, this family of single-stage design cannot optimize real-time metrics that are measured in temporal dimension. In this paper, we target on optimizing the experience rate, which is defined as the ratio of a user's data packet size to the total time for completely delivering the user's data. A novel multi-stage dynamic resource programming is formulated as a multi-stage mixed integer nonlinear programming (MINLP) problem. Then, a low-complexity iterative algorithm is proposed for a jointly optimizing user scheduling and precoding. In particular, a second cone programming problem is dedicatedly designed, for providing a high-quality initial point for the iterative algorithm. The simulation results demonstrate that the proposed design endorses enhanced experience rate performance, with fast convergence behavior.

In this paper, we investigate multi-criteria client selection and energy consumption minimization for hierarchical federated learning (HFL) to deal with clients' heterogeneity and limited energy. To the best of our knowledge, this is the first work to investigate multi-criteria client selection for HFL. A fuzzy logic assisted client selection (FLACS) scheme is proposed, where multiple criteria are taken into account, including the distance, clients' battery capacity and computational resource. The FLACS scheme enables a significant performance gain in terms of the clients' average normalized suitability over the previous schemes. A joint communication and learning factors optimization (JCLFO) algorithm is proposed to minimize the system energy consumption. Thanks to the derived closed-form expressions for the optimal aggregation intervals, computation frequency and transmission power, the JCLFO algorithm can achieve the optimal performance in terms of the system energy consumption and converge within only 3 iterations, with a significant complexity reduction over exhaustive search.

In this paper, the handover problem is investigated for a cellular vehicle-to-everything (C-V2X) based vehicle platoon. We propose a cooperative multipoint transmission (CoMP)-based seamless handover between roadside units (RSUs) and resource allocation (C-SHRA) algorithm, where a CoMP-based seamless handover is proposed to avoid ping-pong handover for RSU-to-platoon (R2P) link, and a cooperative spectrum-sharing based resource allocation scheme is designed to mitigate both intra-platoon and inter-RSU interferences for intra-platoon vehicle-to-vehicle (V2V) links. A joint block length and transmission power optimization (JBLTPO) algorithm is proposed to maximize the effective throughput of the R2P link in the CoMP mode. Numerical results show that the proposed C-SHRA algorithm is verified for effective communication performance enhancement, and the proposed JBLTPO algorithm achieves the optimal performance in terms of effective throughput with a significant complexity reduction over exhaustive search.

The universal presence of networks makes them an important conduit to study interactions in complex natural and artificial systems. While maintaining their integrity is crucial, in many cases, we are also interested in disconnecting them for disease prevention and control, failure containment, crime disruption, etc. With an array of methods exploring node importance, localized execution, and measurement of fragmentation, the choices we have can be disorienting. In this work, we propose a general framework, DisNet, in order to investigate the choice of node centralities and how distributed information gathering and decision making can help us achieve the balance between efficacy and cost of doing so. The framework was evaluated using computer simulation of network dissolution for the full process of weakening, breaking, and shattering. Measurements of focus include the structural losses such as increased effective diameter, homogenization of node degrees, and Shannon diversity of resultant network fragments.

In this paper, we propose a novel dynamic manager selection scheme for vehicles platooning systems and investigate the corresponding resource allocation in the finite blocklength regime, to meet the ultra-reliable and low-latency communication (URLLC) requirements of safety-related data. To the best of our knowledge, this is the first work to investigate the impact of finite blocklength on communications of vehicles platooning systems. By taking into account the factors of communication and the changes in the platoon structure, the proposed dynamic platoon manager selection scheme enables a significant performance enhancement over the conventional fixed manager scheme where the head vehicle in the platoon acts as a manager. Based on the proposed platoon manager selection scheme, a joint resource allocation and coding rate optimization algorithm is proposed to minimize the intra-platoon groupcast latency. Thanks to the closed-form expression of the optimal coding rate and transmission power derived, the proposed optimization algorithm achieves optimal performance in terms of the intra-platoon groupcast latency and converges within only 3 iterations, with a significant complexity reduction over exhaustive search.