Scholarlink

<p>The Internet of Things (IoT) refers to the connection of everyday things to the Internet, enabling them to collect, transmit, and share data. This interconnected network of countless devices transforms existing frameworks, enabling them to interact intelligently with each other and their environment. Today, decentralized systems of smart devices are emerging as a new paradigm in the IoT, due to reasons such as interoperability and limitations in massive data processing within centralized systems. In these distributed systems, IoT components provide data resources for applications, and also offer distributed computing services and communication storage. This enables the creation of a distributed data stream. Numerous distributed computing technologies, along with hardware virtualization, service-oriented architecture, and automated computing, have fueled the development of cloud computing. Researchers have developed and analyzed new algorithms and techniques in the field of distributed computing on the IoT network. However, these often show disadvantages in guaranteeing Quality of Service (QoS) parameters. To address these drawbacks, we propose a novel solution using a combination of an evolutionary algorithm based on the collective defense of elephants and a gray model, named ECDOAGM. This method provides the possibility of distributed processing in the IoT architecture while optimally distributing network load. Evaluation results from the IoT DDoS Honeypot Dataset (IoT-DH) demonstrate that the proposed method is more efficient than other approaches. The proposed approach outperforms the methods of Yosuf et al., Fawzy et al., and Wang et al. by 1.18%, 1.86%, and 0.86% in accuracy for managing distributed data streams, respectively. It also demonstrates faster processing times by 10.12&#xa0;s, 1.3&#xa0;s, and 0.9&#xa0;s compared to the same methods, respectively. Consequently, the proposed approach offers superior QoS.</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

ECDOAGM: an elephant collective defense optimization algorithm with a gray model for managing distributed massive data streams in IoT architectures

  • 期刊论文

  • 发表日期：

    2025年6月15日

查看全文
  • Mohammad Farrokhzadeh Ajirlou,
  • Abbas Jalilvand,
  • Madjid Khalilian

摘要

The Internet of Things (IoT) refers to the connection of everyday things to the Internet, enabling them to collect, transmit, and share data. This interconnected network of countless devices transforms existing frameworks, enabling them to interact intelligently with each other and their environment. Today, decentralized systems of smart devices are emerging as a new paradigm in the IoT, due to reasons such as interoperability and limitations in massive data processing within centralized systems. In these distributed systems, IoT components provide data resources for applications, and also offer distributed computing services and communication storage. This enables the creation of a distributed data stream. Numerous distributed computing technologies, along with hardware virtualization, service-oriented architecture, and automated computing, have fueled the development of cloud computing. Researchers have developed and analyzed new algorithms and techniques in the field of distributed computing on the IoT network. However, these often show disadvantages in guaranteeing Quality of Service (QoS) parameters. To address these drawbacks, we propose a novel solution using a combination of an evolutionary algorithm based on the collective defense of elephants and a gray model, named ECDOAGM. This method provides the possibility of distributed processing in the IoT architecture while optimally distributing network load. Evaluation results from the IoT DDoS Honeypot Dataset (IoT-DH) demonstrate that the proposed method is more efficient than other approaches. The proposed approach outperforms the methods of Yosuf et al., Fawzy et al., and Wang et al. by 1.18%, 1.86%, and 0.86% in accuracy for managing distributed data streams, respectively. It also demonstrates faster processing times by 10.12 s, 1.3 s, and 0.9 s compared to the same methods, respectively. Consequently, the proposed approach offers superior QoS.