
With the rapid development of Wireless Powered Communication Networks (WPCNs), the Hybrid Access Point (HAP), as an important auxiliary means of the network, plays a significant role in improving network performance through its dynamic deployment and task scheduling strategies. This paper aims to study the dynamic deployment and task scheduling of HAPs for WPCNs, in order to achieve the efficient use of network resources and optimization of service quality. Firstly, this paper analyzes the basic architecture of WPCNs and the role of HAPs within them, discussing the importance and challenges of HAP dynamic deployment. Secondly, a dynamic deployment strategy for HAPs based on network performance indicators is proposed, which considers factors such as network coverage, user distribution, energy consumption, and service quality, determining the optimal position and height of HAPs through optimization algorithms. Following that, the paper studies the task scheduling problem of HAPs, designing a priority-based scheduling algorithm that dynamically adjusts the task execution order based on the urgency, importance, and network status of tasks. In addition, a simulation model is constructed in this paper, and the effectiveness of the proposed strategies is verified through simulation experiments. The experimental results show that compared with existing methods, the HAP dynamic deployment strategy proposed in this paper can significantly improve network coverage and user satisfaction, while the task scheduling algorithm can reduce task execution delay and enhance the overall throughput of the network. Finally, the research findings are summarized, and future research directions are prospected. Future research can further explore strategies for multi-HAP collaborative work and how to combine artificial intelligence technology to achieve more intelligent HAP deployment and task scheduling.
In the era of the Internet of Everything, a vast number of Internet of Things (IoT) devices are deployed across various corners of the globe, generating enormous volumes and diverse types of data. When processed, this data can be transformed into significant data assets. However, disputes over property rights during the generation, processing, and utilization of IoT data can hinder efficient circulation and adversely impact data sharing and protection. This paper begins by examining the characteristics of IoT data and proposes a rational classification and tiering system based on the stages of data formation, thereby establishing a practical mechanism for rights confirmation. Furthermore, it presents a technological pathway for confirming IoT data rights using blockchain technology, enabling the confirmation of rights and possession changes throughout the entire lifecycle of IoT data. This approach significantly supports the assetization and transactional circulation of IoT data.
As a crucial component of multimedia content delivery systems, including those deployed in sensor network-based environments, the digital rights management (DRM) system ensures that multimedia content is securely used only by authorized users. The logical architecture of the DRM system is continually improving, introducing new business requirements for the system in aspects such as disaster recovery, elasticity, and business scalability. In these regards, traditional single-host deployment methods face disadvantages. With the emergence of containerization and cloud deployment trends, the deployment architecture based on container technology and orchestration provides a new solution for deploying the DRM system. This paper proposes a cloud deployment solution for the server-side of the DRM system based on container technology. Leveraging the powerful computational capabilities of cloud platforms and the management capabilities of container orchestration platforms for business systems, this solution enhances the system's abilities in rapid startup, disaster recovery migration, and automatic scaling. Test results indicate that the proposed solution exhibits better performance in terms of business startup speed and scalability. In terms of cluster scaling, this paper models the cluster scaling problem, leveraging GRU to model the temporal patterns of service access traffic pressure. Building upon temporal modelling, reinforcement learning algorithms are employed to learn optimal cluster scaling policies. This results in an intelligent cluster scaling decision mechanism based on GRU and Advantage Actor-Critic (A2C) algorithms. Simulation results demonstrate that the proposed mechanism can quickly control the cluster size in various testing scenarios, ensuring better service quality for the business system. The proposed server-side DRM cloud deployment scheme is applicable to multimedia content delivery scenarios involving sensor networks, where large-scale sensor-generated multimedia data require secure access control and elastic cloud-side processing.
In the realm of cloud computing (CC) services, task scheduling (TS) and resource allocation (RA) are key components that have a considerable impact on system performance, particularly with the emergence of nextgeneration networks like 6G. While current approaches to scheduling and resource management in cloud environments rely mainly on conventional optimization techniques, they frequently fail to meet the dynamic and complicated needs of 6G networks, such as ultra-low delay, high flexibility, and network slicing. The research presents a new dynamic scheduling technique for CC services that uses 6G network slicing in ad hoc networks, which are highly dynamic and decentralized environments. The goal is to improve TS and RA using Intelligent Coyote Optimised Deep Reinforcement Learning (ICO-DRL) techniques to address the constraints of existing systems. In particular, the research employs a hybrid approach for RA and TS in a 6G-enabled cloud environment and ad hoc network scenarios that combines deep reinforcement learning (DRL) with an Intelligent Coyote Optimisation (ICO) algorithm. The DRL component offers flexibility in response to changing network circumstances, while the ICO offers effective RA. The findings demonstrate that the suggested approach outperforms conventional techniques, resulting in reduced latency, optimal cloud resource utilization, and increased efficiency in TS and RA. Task scheduling efficiency is 90.6% and resource allocation efficiency is 92.46%. The proposed model offers real-time, adaptive decision-making, which greatly enhances the quality of service in cloud computing systems based on 6G. The ICO-DRL algorithm improves task scheduling and resource allocation by employing real-time feedback, rather than the "static" approaches taken by traditional optimization models. Therefore, ICO-DRL is more amenable to deployment in a cloud context and ad hoc network environments, where applications, processing, and scalability need to occur instantaneously. Finally, the findings showed that combining sophisticated optimization approaches with deep learning (DL) can help handle the issues of TS and RA in 6G-enabled CC services.
Abnormal network traffic events occur frequently, and the methods and means that lead to abnormal network traffic are constantly updated, making traditional digital identification and network attack protection more difficult. As network security faces numerous challenges, detecting abnormal traffic and protecting privacy are both difficult tasks, particularly in sensor networks. It is necessary to introduce new technologies, methods, and perspectives, such as machine learning and blockchain, to solve some problems. Therefore, this paper expands the research content of network abnormal traffic detection and privacy protection based on machine learning and blockchain technology, proposes a network abnormal traffic detection model and privacy protection algorithm for sensor networks, and illustrates the effectiveness and advantages of this method with experiments. The experimental results also demonstrate that this method has a significant impact on anomaly detection and privacy protection.