Scholarlink

<p>A Long Short-Term Memory (LSTM) auto-encoder is used for anomaly detection in this study’s hybrid computation-and-security approach, while a Random Forest algorithm is used to assume multi-class attack classification. However, Apache Kafka is used for real-time traffic analysis, and because the alerting system is built using smart contracts on the Blockchain, the alert storage is auditable and immutable. Synthetic Minority Over-sampling Technique (SMOTE) was used to balance the Adaptive Blockchain-based Trustworthy Access Policy (AB-TRAP) dataset, and Principal Component Analysis (PCA) was used to achieve dimensional reduction in order to assess the suggested system. Despite preventing log tampering and maintaining a false alarm rate below 2%, the proposed system demonstrated an extremely high detection accuracy of 96.58%. This design is therefore a cutting-edge, scalable, and secure solution for attack detection and log auditing.</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 hybrid LSTM-auto encoder and random forest-based cybersecurity framework with blockchain and Kafka integration

  • 期刊论文

  • 发表日期：

    2025年11月21日

查看全文
  • D. S. Sahana,
  • Sneha Sureddy

摘要

A Long Short-Term Memory (LSTM) auto-encoder is used for anomaly detection in this study’s hybrid computation-and-security approach, while a Random Forest algorithm is used to assume multi-class attack classification. However, Apache Kafka is used for real-time traffic analysis, and because the alerting system is built using smart contracts on the Blockchain, the alert storage is auditable and immutable. Synthetic Minority Over-sampling Technique (SMOTE) was used to balance the Adaptive Blockchain-based Trustworthy Access Policy (AB-TRAP) dataset, and Principal Component Analysis (PCA) was used to achieve dimensional reduction in order to assess the suggested system. Despite preventing log tampering and maintaining a false alarm rate below 2%, the proposed system demonstrated an extremely high detection accuracy of 96.58%. This design is therefore a cutting-edge, scalable, and secure solution for attack detection and log auditing.