International Journal of Network Management最新文献

The rapid advancement of communication technologies, particularly in English language learning, is sharing education with the implementation of sixth-generation (6G) networks, offering immersive and interactive learning experiences. The purpose of the research is to establish an advanced method for sharing English education resources across multiple virtual networks enabled by 6G technology. Traditional resource-sharing systems lack the effectiveness and optimization requirement for large-scale instructional assignments, especially in virtual settings with various user demands. To address this, the study proposed a novel Dynamic Tunicate Swarm Refined Graph Neural Networks (DTS-RGNN) model to optimize resource allocation and improve the efficiency of resource sharing among educational tasks. The approach uses TSO for resource allocation scalable through 6G technology and GNN for task assignment according to the previous performances and interaction with the students to balance resource utilization. The experimental group performed writing (90%), sharing (91%), listening (85%), and reading (75%), finishing the task in 5.5 s at 1000 GB. Throughput increased by 5.0 GBps and resource utilization efficiency improved to (96%) and student outcomes showed high satisfaction (93%), retention (89%), and engagement (90%). The findings demonstrated the proposed method significantly improves the sharing of online English education resources, promoting more interactive and effective language learning experiences in virtual networks.

In this paper, a novel workflow scheduling framework is proposed using genetically modified golden jackal optimization (GM-GJO) with recurrent autoencoder. An integrated autoencoder and bidirectional gated recurrent unit (iAE-BiGRU) are used to forecast the number of virtual machines (VMs) needed to manage the system's present workload. The following step involves assigning the tasks of several workflows to cloud VMs through the use of the GM-GJO method for multiworkflow scheduling. GM-GJO provides optimal workflow scheduling by considering minimal maximizing utilization rate, minimizing makespan, and minimizing the number of deadline missed workflows. The proposed approach attempts to allocate the best possible set of resources for the workflows based on objectives such as deadline, cost, and quality of service (QoS). Extensive experiments were conducted with the CloudSIM tool, and the performance is evaluated in terms of scheduling length ratio, cost, QoS, etc. The execution time of 513.45 ms is achieved with a Sipht workflow of 30 tasks. When comparing the suggested strategy to the current methodologies, the suggested approach performs better.

Electronic archives are now widely used in many different industries and serve as the primary method of information management and storage because of the rapid growth of information technology and mobile networks. To enhance the security of electronic archives in mobile networks, the research utilizes the federated learning mechanism to design a federated learning model based on homomorphic aggregation cryptographic signature scheme combined with mobile network management. The use of homomorphic encryption technology in the signing process of electronic archives enables the aggregation of multiple electronic file signatures into a single signature without exposing the data of the electronic archives. This reduces the computational and storage requirements for signature verification. At the same time, a secure aggregation signature scheme is used to ensure the integrity and security of the data in the aggregation process. A novel approach is presented in this study, whereby trusted federated learning models are innovatively combined with homomorphic aggregate signature technology. This integration ensures data integrity through aggregate signature schemes. The results showed that, under mobile network management, the longest encryption time of the trusted federated learning model was 52 ms, and the longest decryption time was 44 ms. The accuracy of the optimized learning model reached 97.49%, and the loss value was significantly reduced to 0.09. To summarize, the electronic archive security protection method based on homomorphic aggregation signature scheme effectively improves the archive data protection efficiency and security.

Effective spectrum allocation in 5G and beyond intelligent ubiquitous networks is vital for predicting future frequency band needs and ensuring optimal network performance. As wireless communication evolves from 4G to 5G and beyond, it has brought about remarkable advancements in speed and connectivity. However, with the growing demand for higher data rates and increased network capacity, new challenges in managing and utilizing network frequencies have emerged. Accurately forecasting spectrum requirements is critical to addressing these challenges. This research explores how machine learning (ML) plays a pivotal role in optimizing network performance through intelligent decision-making, predictive analysis, and adaptive management of network resources. By leveraging ML algorithms, networks can autonomously self-optimize in real time, adjusting to changing conditions and improving performance in 5G and beyond. The effectiveness of our approach was demonstrated through an extensive case study, which showed that it not only meets spectrum requirements in various environments but also significantly reduces energy consumption by pinpointing the appropriate spectrum range for each location. These results underscore the approach's potential for enhancing spectrum management in future networks, offering a scalable and efficient solution to the challenges facing 5G and beyond.

The increasing scale and complexity of network infrastructure present a huge challenge for network operators and administrators in performing network configuration and management tasks. Intent-based networking has emerged as a solution to simplify the configuration and management of networks. However, one of the most difficult tasks of intent-based networking is correctly translating high-level natural language intents into low-level network configurations. In this paper, we propose a general and effective approach to perform the network intent translation task using large language models with fine-tuning, dynamic in-context learning, and continuous learning. Fine-tuning allows a pretrained large language model to perform better on a specific task. In-context learning enables large language models to learn from the examples provided along with the actual intent. Continuous learning allows the system to improve overtime with new user intents. To demonstrate the feasibility of our approach, we present and evaluate it with two use cases: network formal specification translation and network function virtualization configuration. Our evaluation shows that with the proposed approach, we can achieve high intent translation accuracy as well as fast processing times using small large language models that can run on a single consumer-grade GPU.

Wrong parking incidents pose a pervasive challenge in urban environments, disrupting the smooth flow of traffic, compromising safety and contributing to various logistical issues. Unauthorized parking occurs when vehicles are parked in locations not designated for such purposes, leading to a myriad of problems for both authorities and the general public. This research introduces a pioneering approach to confront the persistent challenge of unauthorized parking incidents in urban environments. The study focuses on harnessing the advanced capabilities of the FusionNet model to enhance the accuracy of license plate detection. This paper introduces the YOLO v8 Model, a deep learning architecture designed to enhance urban parking management by accurately detecting vehicles parked in unauthorized slots. The objective is to enhance parking management efficiency by accurately detecting vehicles and their occupancy status in designated parking areas. The methodology begins with data collection and preprocessing of images of parking spaces, followed by the training of YOLO v8 to identify vehicles and parking spaces in real time. Leveraging a diverse dataset encompassing various parking scenarios, including instances of unauthorized parking, the model achieves an accuracy of 98.50% in identifying vehicles outside designated areas. This model segments characters from detected license plates, enabling the accurate extraction of alphanumeric information associated with each vehicle. The integrated system provides timely identification of parking violations and facilitates effective enforcement actions through captured license plate data. Results demonstrate the model's effectiveness in real-world scenarios, showcasing its potential for improving urban safety and efficiency. The implementation of FusionNet in the Python programming language, the proposed solution aims to streamline parking management, improve compliance with parking regulations and enhance overall urban mobility., with robust precision 96.17%, specificity 97.42% and sensitivity 96.19%, surpassing other MobileNet, CNN, ANN, DNN and EfficientNet models.

The acceleration of the digitalization process in enterprise and university education management has generated a massive amount of electronic archive data. In order to improve the intelligence, storage quality, and efficiency of electronic records management and achieve efficient storage and fast retrieval of data storage models, this study proposes a massive data storage model based on HBase and its retrieval optimization scheme design. In addition, HDFS is introduced to construct a two-level storage structure and optimize values to improve the scalability and load balancing of HBase, and the retrieval efficiency of the HBase storage model is improved through SL-TCR and BF filters. The results indicated that HDFS could automatically recover data after node, network partition, and NameNode failures. The write time of HBase was 56 s, which was 132 and 246 s less than Cassandra and CockroachDB. The query latency was reduced by 23% and 32%, and the query time was reduced by 9988.51 ms, demonstrating high reliability and efficiency. The delay of BF-SL-TCL was 1379.28 s after 1000 searches, which was 224.78 and 212.74 s less than SL-TCL and Blockchain Retrieval Acceleration and reduced the delay under high search times. In summary, this storage model has obvious advantages in storing massive amounts of electronic archive data and has high security and retrieval efficiency, which provides important reference for the design of storage models for future electronic archive management. The storage model designed by the research institute has obvious advantages in storing massive electronic archive data, solving the problem of lack of scalability in electronic archive management when facing massive data, and has high security and retrieval efficiency. It has important reference for the design of storage models for future electronic archive management.

Centralization of Internet-based services in a few key players has been a topic of study in recent years. One of such services, the domain name system (DNS), is one of the pillars of the Internet, which allows users to access websites on the Internet through easy-to-remember domain names rather than complex numeric IP addresses. In this DNS context, the reliance on a small number of large DNS providers can lead to (a) risks of data breaches and disruption of service in the event of failures and (b) concerns about the digital sovereignty of countries regarding DNS hosting. As several essential services are provided through electronic government (E-Gov), it is highly important to be able to measure the digital sovereignty of a nation and the impacts that the lack of such feature can bring to its citizens. This work approaches the issue of DNS concentration on the Internet by presenting a solution to measure DNS hosting centralization and digital sovereignty in different countries, such as Brazil, India, China, Russia, and South Africa. With the data obtained through these measurements, relevant questions are answered, such as which are the top-10 DNS providers, if there is DNS centralization, and how dependent countries are on such providers to manage domains using their country code top-level domains (ccTLD). Future opportunities could investigate the impacts on sovereignty under the lens of other layers of the open systems interconnection (OSI) Network Sovereignty representation model presented in this work.