Scholarlink

<p>The rise of medical IoT devices has transformed healthcare through real-time monitoring and data-driven decision-making. However, traditional security methods often fall short in terms of energy efficiency, adaptability, and resilience against cyber threats. This work introduces BBMDH-OTBAAT—a secure framework that integrates Blockchain-based Bipartite Modified Diffie–Hellman (CBGT-mDH) encryption with an optimized Two-Branch Attention Adversarial Transformer (TBAADFPT) for anomaly detection in IoT networks. The system begins with data preprocessing, which involves noise removal, standardization, and anonymization. Secure encoding is achieved using a UFD-based bipartite graph encryption, with optimal key selection via the Black-Winged Kite Algorithm (BKA). A Proof of Authority (PoA) blockchain ensures decentralized, immutable storage. The TBAADFPT model extracts spatial–temporal patterns for robust intrusion detection, fine-tuned by BKA for enhanced accuracy. The proposed framework achieves 99.2% detection accuracy, 98.9% precision, 0.4% false positive rate, and operates with low encryption/decryption energy costs (8.2&#xa0;mJ / 7.5&#xa0;mJ). It maintains a modest communication overhead (2.7%) and storage requirement (12.8&#xa0;MB), making it suitable for secure, real-time healthcare IoT applications.</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

Secure IoT Medical Data Storage and Intrusion Detection Using Blockchain Encryption and Optimized Two-Branch Attention Adversarial Transformer

  • 期刊论文

  • 发表日期：

    2025年12月18日

查看全文
  • B. H. Baba Fakruddin Ali,
  • M. A. Gunavathie,
  • P. Karthigaikumar,
  • T. Giri babu

摘要

The rise of medical IoT devices has transformed healthcare through real-time monitoring and data-driven decision-making. However, traditional security methods often fall short in terms of energy efficiency, adaptability, and resilience against cyber threats. This work introduces BBMDH-OTBAAT—a secure framework that integrates Blockchain-based Bipartite Modified Diffie–Hellman (CBGT-mDH) encryption with an optimized Two-Branch Attention Adversarial Transformer (TBAADFPT) for anomaly detection in IoT networks. The system begins with data preprocessing, which involves noise removal, standardization, and anonymization. Secure encoding is achieved using a UFD-based bipartite graph encryption, with optimal key selection via the Black-Winged Kite Algorithm (BKA). A Proof of Authority (PoA) blockchain ensures decentralized, immutable storage. The TBAADFPT model extracts spatial–temporal patterns for robust intrusion detection, fine-tuned by BKA for enhanced accuracy. The proposed framework achieves 99.2% detection accuracy, 98.9% precision, 0.4% false positive rate, and operates with low encryption/decryption energy costs (8.2 mJ / 7.5 mJ). It maintains a modest communication overhead (2.7%) and storage requirement (12.8 MB), making it suitable for secure, real-time healthcare IoT applications.