Scholarlink

<p>The rapid expansion of the Internet of Things (IoT) ecosystem has increased vulnerability to distributed attacks and data manipulation threats, necessitating robust and efficient attack detection mechanisms. Therefore, to address these challenges, this paper proposes a blockchain (BC)-enabled smart contract framework for IoT threat detection that combines an Enhanced Residual Gated Recurrent Unit (E-RGRU) with a cryptographic one-way compression function. The E-RGRU model incorporates residual connections to improve temporal feature learning and mitigate gradient degradation. The Adaptive Coati Optimization Algorithm (ACOA) is used to&#xa0;optimize&#xa0;the hyperparameters of RGRU to enhance detection performance. To ensure data integrity and confidentiality, the Miyaguchi–Preneel cryptographic algorithm is used to generate tamper-resistant hash for normal input data. The suggested technique is implemented in Python. The proposed framework was evaluated on the N-BaIoT and DDoS benchmark datasets. Experimental results demonstrate superior performance, achieving high accuracy of 98.59 and 97.74%, confidentiality rates of 98.5 and 98%, and data integrity rates of 97.8 and 97.5% on the N-BaIoT and DDoS datasets, respectively. The proposed model significantly outperforms the state-of-the-art models, confirming its effectiveness in enhancing IoT security.</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

A smart contract based on enhanced residual gated recurrent unit with the one-way compression function for attack detection in IoT

  • 期刊论文

  • 出版模式：

    开放获取

  • 发表日期：

    2026年3月29日

查看全文
  • T. Nishitha,
  • Akhil Khare,
  • V. Sumalatha,
  • Swathi Sambangi

摘要

The rapid expansion of the Internet of Things (IoT) ecosystem has increased vulnerability to distributed attacks and data manipulation threats, necessitating robust and efficient attack detection mechanisms. Therefore, to address these challenges, this paper proposes a blockchain (BC)-enabled smart contract framework for IoT threat detection that combines an Enhanced Residual Gated Recurrent Unit (E-RGRU) with a cryptographic one-way compression function. The E-RGRU model incorporates residual connections to improve temporal feature learning and mitigate gradient degradation. The Adaptive Coati Optimization Algorithm (ACOA) is used to optimize the hyperparameters of RGRU to enhance detection performance. To ensure data integrity and confidentiality, the Miyaguchi–Preneel cryptographic algorithm is used to generate tamper-resistant hash for normal input data. The suggested technique is implemented in Python. The proposed framework was evaluated on the N-BaIoT and DDoS benchmark datasets. Experimental results demonstrate superior performance, achieving high accuracy of 98.59 and 97.74%, confidentiality rates of 98.5 and 98%, and data integrity rates of 97.8 and 97.5% on the N-BaIoT and DDoS datasets, respectively. The proposed model significantly outperforms the state-of-the-art models, confirming its effectiveness in enhancing IoT security.