Scholarlink

<p>The IoT or Internet of Things explosive expansion has raised the demand for dependable and safe mechanisms help protect sensitive data and connected devices from changing cyberthreats. This study proposes an integrated IoT security framework that combines optimized key generation, secure authentication, blockchain-enabled data storage, intelligent Deep learning-based intrusion detection and feature selection within a unified architecture. The framework employs Puma Optimizer (PO)-based key generation Hybrid and Adaptive Cryptographic (HAC) authentication, blockchain-based secure storage, Fuzzy Min–Max Neural Network (FMMNN) preprocessing, Hiking Optimization Algorithm (HOA) feature selection, Dual-Level Contextual Attention Scalable Quantum Non-local Network (DLCA-SQN) intrusion detection, and Draco Lizard Optimizer (DLO)-based loss optimization. The experimental approach proves demonstrating the recommended method outperforms the alternative methods regarding intrusion detection capabilities, security, and computation overhead reduction. Moreover, the results obtained using cross-validation prove that the suggested framework is robust and generalizable. The incorporation of features like authentication, encryption, security optimization, and intrusion detection into one framework makes it a reliable solution addressing security concerns in the age of the Internet of Things.</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

Dual-Level Contextual Attention Scalable Quantum Non-local Network–Driven Cryptographic Authentication for Secure IoT Data Protection and Intrusion Detection

  • 期刊论文

  • 发表日期：

    2026年7月8日

查看全文
  • P. R. Jebisha,
  • M. R. Geetha,
  • T. Rajesh,
  • M. R. Kavitha

摘要

The IoT or Internet of Things explosive expansion has raised the demand for dependable and safe mechanisms help protect sensitive data and connected devices from changing cyberthreats. This study proposes an integrated IoT security framework that combines optimized key generation, secure authentication, blockchain-enabled data storage, intelligent Deep learning-based intrusion detection and feature selection within a unified architecture. The framework employs Puma Optimizer (PO)-based key generation Hybrid and Adaptive Cryptographic (HAC) authentication, blockchain-based secure storage, Fuzzy Min–Max Neural Network (FMMNN) preprocessing, Hiking Optimization Algorithm (HOA) feature selection, Dual-Level Contextual Attention Scalable Quantum Non-local Network (DLCA-SQN) intrusion detection, and Draco Lizard Optimizer (DLO)-based loss optimization. The experimental approach proves demonstrating the recommended method outperforms the alternative methods regarding intrusion detection capabilities, security, and computation overhead reduction. Moreover, the results obtained using cross-validation prove that the suggested framework is robust and generalizable. The incorporation of features like authentication, encryption, security optimization, and intrusion detection into one framework makes it a reliable solution addressing security concerns in the age of the Internet of Things.