Scholarlink

<p>The rapid proliferation of Internet of Things (IoT) devices has intensified the demand for secure and efficient data storage solutions. While blockchain technology offers strong guarantees for data integrity and security in IoT environments, the efficient allocation of cloud resources, such as storage capacity, energy consumption, and operational cost, remains a complex optimization problem. This paper presents a new approach that combines the Non-Dominated Sorting Genetic Algorithm III (NSGA-III) with the Support Vector Machines algorithm (SVM) to increase the size of data blocks and improve the use of cloud resources in blockchain-based IoT systems. The main idea is using a machine learning model (SVM) within the NSGA-III process to enhance convergence speed and maintain solution diversity across generations. Through extensive testing and careful comparisons, the new NSGAIII-SVM method consistently outperforms other algorithms (VEGA, NSGA-II, and standalone NSGA-III), showing improvements of up to 35% in hypervolume and 30% in spacing. Additionally, it yields a 25% reduction in block transmission time, a 20% decrease in block composition time, and a 10% gain in energy efficiency. These results confirm the validity, scalability, and practical applicability of the proposed method, offering a robust solution for resource-efficient blockchain-based IoT deployments.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Optimizing block size and cloud storage in blockchain technology using an NSGA-III and SVM hybrid approach

  • 期刊论文

  • 发表日期：

    2025年11月12日

查看全文
  • Issam Zidi,
  • Radhia Zaghdoud,
  • Salim El Khediri

摘要

The rapid proliferation of Internet of Things (IoT) devices has intensified the demand for secure and efficient data storage solutions. While blockchain technology offers strong guarantees for data integrity and security in IoT environments, the efficient allocation of cloud resources, such as storage capacity, energy consumption, and operational cost, remains a complex optimization problem. This paper presents a new approach that combines the Non-Dominated Sorting Genetic Algorithm III (NSGA-III) with the Support Vector Machines algorithm (SVM) to increase the size of data blocks and improve the use of cloud resources in blockchain-based IoT systems. The main idea is using a machine learning model (SVM) within the NSGA-III process to enhance convergence speed and maintain solution diversity across generations. Through extensive testing and careful comparisons, the new NSGAIII-SVM method consistently outperforms other algorithms (VEGA, NSGA-II, and standalone NSGA-III), showing improvements of up to 35% in hypervolume and 30% in spacing. Additionally, it yields a 25% reduction in block transmission time, a 20% decrease in block composition time, and a 10% gain in energy efficiency. These results confirm the validity, scalability, and practical applicability of the proposed method, offering a robust solution for resource-efficient blockchain-based IoT deployments.