Malware propagation in Internet of Underwater Things (IoUT) can disrupt stable communications among wireless devices. Timely control over its spread is beneficial for the stable operation of IoUT. Notably, the instability of the underwater environment causes the propagation effects of malware to vary continuously. Traditional control methods cannot quickly adapt to these abrupt changes. In recent years, the rapid development of reinforcement learning (RL) has significantly advanced control schemes. However, previous RL methods relied on long-term interactions to obtain a large amount of interaction data in order to form effective strategy. Given the particularity of underwater communication media, data collection for RL in IoUT is challenging. Therefore, improving sample efficiency has become a critical issue that current RL methods need to address urgently. The algorithm of Attention-Model-Based Multiagent Policy Optimization (AMBMPO) is proposed to achieve efficient use of data samples in this study. First, the algorithm employs an explicit prediction model to reduce the dependence on precise model. Secondly, an attention mechanism network is designed to capture high-dimensional state sequences, thereby reducing the compound errors during policy training. Finally, the proposed method is validated for optimal control problems and compared with verified benchmarks. The experimental results show that, compared with existing advanced RL algorithms, AMBMPO demonstrates significant advantages in sample efficiency and stability. This work effectively controls the spread of malware in underwater systems through an interactive evolution based approach. It provides a new implementation approach for ensuring the safety of underwater systems in deep-sea exploration and environmental monitoring applications.

To address the dynamic control challenges of malware propagation in smart agricultural UAV-WSN (Unmanned Aerial Vehicle-Wireless Sensor Network) systems, this paper proposes a novel hybrid control modeling framework based on fractional-order calculus. A 14-dimensional dual-layer SEIQR-SEIQRL fractional-order propagation model with time-delay effects is established to characterize cross-layer infection mechanisms between UAVs and sensor nodes, as well as system memory properties. By innovatively integrating direct treatment and dynamic quarantine strategies, and further incorporating node energy constraints, this study formulates an optimal control problem. To overcome the dependency of traditional numerical solutions on system models, a reinforcement learning (RL)-based decision framework is designed, which integrates the Proximal Policy Optimization (PPO), Soft Actor-Critic (SAC), Twin Delayed Deep Deterministic policy gradient (TD3) and Multi-Agent Soft Actor-Critic (MASAC) algorithms. Experimental results validate the memory advantages of the proposed fractional-order theory, the effectiveness of hybrid control strategies, and the adaptability of RL-based solvers. This study provides a dynamically defensive architecture with memory-aware capabilities, offering both theoretical innovation and engineering applicability for IoT security in smart agriculture, thereby supporting reliable data acquisition in sustainable precision agriculture.

Autonomous Underwater Vehicle (AUV)-assisted Underwater Wireless Sensor Networks (UWSNs) offer transformative potential for data collection, disaster prediction and environmental monitoring. However, these networks remain critically vulnerable to malware attacks due to inherent communication challenges and limited energy resources. To rigorously capture the spatiotemporal dynamics of malware propagation, this paper introduces a novel two-layer heterogeneous model rooted in Partial Differential Equations (PDEs). To suppress malware propagation under limited bandwidth, this paper proposes a spatiotemporal optimal control strategy integrating saturation functions. The existence of the optimal pair is demonstrated by the minimizing sequence method, and the first-order necessary condition for spatiotemporal control is derived based on convex perturbation methods. Further, extensive numerical experiments validate the effectiveness of the proposed two-layer heterogeneous model and spatiotemporal control strategy. Finally, the applications of the proposed model and control strategy in real-world scenarios are discussed.

Internet of Vehicles (IoV) is gradually becoming popular, but it also brings more opportunities for malware intrusion. The intrusion of malware into IoV will cause a series of security issues and increase the incidence of road accidents. Therefore, the suppressing measures to combat the spread of malware in IoV will be fundamental and urgent. To address this critical issue, this paper proposes a fractional-order IoV (FIOV) to investigate malware propagation patterns in Road Side Unit (RSU) and Vehicles. To accurately reflect the actual spread of malware, the traffic density, the channel fading and the actual connectivity are considered in mathematical model. Then, the model-based optimal treatment and quarantine control strategy is derived by optimal control theory. Additionally, a novel model-free FIOV multi-agent soft actor-critic (FIOV-MASAC) approach is first proposed to suppress the malware propagation in IoV. Simulation experiments demonstrate that the proposed FIOV-MASAC approach exhibits better learning ability compared to other reinforcement learning (RL) algorithms.Note to Practitioners—Frequent attacks by malware on IoV are recognized as being challenging to prevent, with these attacks posing threats to data security and potentially resulting in traffic accidents and vehicle malfunctions. In response, a novel mathematical model has been introduced within this study to better predict the propagation trends of malware in IoV, effectively managing its spread within the vehicular network systems. While RL methods have been extensively utilized in the domain of control systems, it is noted that current RL methods depend on rich experience pools, rendering them inapplicable to more complex systems without adaptation. To address this, an effective and pragmatic RL algorithm has been devised in this study. This algorithm, devoid of the requirement for complex model establishment, is capable of intelligently learning and adjusting to the sophisticated environment of IoV, thereby effectively countering the propagation of malware. It should be highlighted that the RL method proposed herein is applicable to the majority of epidemic systems, enabling the achievement of stable control while substantially minimizing control expenditures. The integration of this method is anticipated to augment the security and robustness of IoV in the face of malware attacks.

Internet of Underwater Things (IoUT), widely utilized in data collection, ocean exploration and disaster prevention, are prone to malware attacks. To accurately describe the spatiotemporal dynamics of malware propagation in IoUT, a new Reaction-Diffusion Sleep Control Model (RDSCM) based on Partial Differential Equations (PDEs) is established in this paper. To target heavily infected regions and optimize control costs, a spatiotemporal hybrid optimal control strategy is implemented. First, based on the formulation of the spatiotemporal optimal control problem, the existence, uniqueness, and some estimates of the strong solution of the controlled system are obtained using the truncation method and semigroup theory. Then, the existence of the optimal pair is verified through the minimization sequence technique. Subsequently, the first-order necessary optimality condition for the optimal control problem of PDEs is derived by proving the differentiability of the control-to-state mapping. To validate the effectiveness of the proposed model and control methodology, three comparative experimental studies are conducted. Finally, some discussions are provided on the extension and application of the proposed method.

The Internet of Underwater Things (IoUT) relies on wireless communication devices that are arranged in an open underwater environment and can interact with other devices through acoustic communication technology. However, due to their limited resources and open underwater environment, IoUT has been suffering from a high risk of malware attacks. As the two main parts of IoUT, autonomous underwater vehicles (AUVs) and underwater wireless rechargeable sensor networks (UWRSNs) are more favored by attackers and can be directly attacked by malware, which can lead to the cross-propagation of malware when AUVs and UWRSNs exchange information. To mitigate that threat, there is an urgent need to study the propagation patterns of malware and control their spread. Therefore, we establish a mathematical model based on the fractional-order theoretical framework to investigate the malware propagation patterns in two coupled networks (UWRSNs and AUVs). Then, we combine immune, charging, and quarantine delays in control and derive the optimal control strategies based on optimal control theory. Moreover, to improve the generality of control, we propose a machine learning (ML) controller that combines ML [e.g., deep neural network (DNN) and random forest (RF)] with control theory. Ultimately, our simulation experiments show that the proposed optimal control strategy is more effective in inhibiting the spread of malware while obtaining the minimum control cost under different fractional-order scenes. At the same time, the ML-based control results are close to the optimal control.

Unmanned aerial vehicle (UAV) and wireless sensor network (WSN) system widely used in emergency sites monitoring, disaster relief and data collection, is vulnerable to malware attack. In this article, we establish a novel epidemic model to accurately describe the characteristics of malware cross-propagation and stochastic disturbances between UAV and WSN. To control their spread, the hybrid charging and patching strategies are optimized by the proposed stochastic hybrid optimal control (OC) method while minimizing control costs. An effective stacking ensemble learning control (SELC) approach is developed to obtain a sub-OC strategy. Extensive comparison and simulations are shown to validate the effectiveness of the proposed stochastic hybrid OC method and SELC method.

With the proliferation of mobile devices and various sensors (e.g., GPS, magnetometer, accelerometers, gyroscopes) equipped, richer services, e.g. location based services, are provided to users. A series of methods have been proposed to protect the users' privacy, especially the trajectory privacy. Hardware fingerprinting has been demonstrated to be a surprising and effective source for identifying/authenticating devices. In this work, we show that a few data samples collected from the motion sensors are enough to uniquely identify the source mobile device, i.e., the raw motion sensor data serves as a fingerprint of the mobile device. Specifically, we first analytically understand the fingerprinting capacity using features extracted from hardware data. To capture the essential device feature automatically, we design a multi-LSTM neural network to fingerprint mobile device sensor in real-life uses, instead of using handcrafted features by existing work. Using data collected over 6 months, for arbitrary user movements, our fingerprinting model achieves 93% F-score given one second data, while the state-of-the-art work achieves 79% F-score. Given ten seconds randomly sampled data, our model can achieve 98.8% accuracy. We also propose a novel generative model to modify the original sensor data and yield anonymized data with little fingerprint information while retain good data utility.