Scholarlink

<p>Mobile ad hoc networks (MANETs) incorporate attack detection as a mechanism to guarantee security by detecting and countering dangerous activities or unauthorized access. The existing techniques are currently facing severe issues regarding latency, energy consumption, complexity, and trust models. Taking into account all the issues, a secure routing and data transfer mechanism in the blockchain-based MANET is proposed. This solution integrates the blockchain protocol with cluster formation by using the Improvement Frog Snake prey-predator Relationship Optimization Algorithm (IFSROA) and also includes attack detection by using the Bi-directional Cascade Residual Convolutional Neural Network (BCRCNN). The selfish node trust-aware elk herd multi-objective routing methodology ensures safe data transfer while evaluating its performance based on different parameters in the NS3 simulator. This comprehensive technique represents a significant step forward in the domain of attack detection methodologies. The proposed network shows excellent performance with an accuracy of 99.56%, a precision of 99.35%, an F1-score value reaching 99.7%, and a packet delivery ratio equal to 99.9%. Integration of blockchain in the proposed network enhances network security by providing advanced learning techniques that guarantee high detection rates for several attack types, routing, and data transmission.</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 routing and attack detection in mobile ad hoc networks via bi-directional cascade residual convolutional neural networks

  • 期刊论文

  • 发表日期：

    2026年4月3日

查看全文
  • B. Uma,
  • S. Sumathi

摘要

Mobile ad hoc networks (MANETs) incorporate attack detection as a mechanism to guarantee security by detecting and countering dangerous activities or unauthorized access. The existing techniques are currently facing severe issues regarding latency, energy consumption, complexity, and trust models. Taking into account all the issues, a secure routing and data transfer mechanism in the blockchain-based MANET is proposed. This solution integrates the blockchain protocol with cluster formation by using the Improvement Frog Snake prey-predator Relationship Optimization Algorithm (IFSROA) and also includes attack detection by using the Bi-directional Cascade Residual Convolutional Neural Network (BCRCNN). The selfish node trust-aware elk herd multi-objective routing methodology ensures safe data transfer while evaluating its performance based on different parameters in the NS3 simulator. This comprehensive technique represents a significant step forward in the domain of attack detection methodologies. The proposed network shows excellent performance with an accuracy of 99.56%, a precision of 99.35%, an F1-score value reaching 99.7%, and a packet delivery ratio equal to 99.9%. Integration of blockchain in the proposed network enhances network security by providing advanced learning techniques that guarantee high detection rates for several attack types, routing, and data transmission.