Scholarlink

<p>The recent growth in cloud computing has impacted industries betterment, but has also created complicated cybersecurity problems. In protecting the cloud infrastructure, IDS (intrusion detection systems) play a crucial role. The use of Advanced Persistent Threats (APTs) has rendered traditional methods and models ineffective along with the WID-ALL principle. To deal with this problem, deep learning techniques have been adopted by researchers, especially with the use of Convolutional Neural Networks (CNNs) as well as Recurrent Neural Networks, performing exceptionally well in traffic anomaly detection and classification. This article focuses on the performance hybrid frameworks of CNN and RNN improves in temporal and spatial feature learning related to detection in cloud environments. The paper focuses on the discussion of the architectural components of CNNs and RNNs, describing their individual strengths and the benefits gained through hybridization. Moreover, it explains the problems with implementing such systems in actual cloud infrastructure, particularly about scalability, probative value, interpretability of results, and availability of datasets. This review provides an analysis of the cloud that could help in devising intelligent, adaptive, and scalable IDS systems by studying emerging trends and possible research areas. Such findings enhance understanding of hyper deep learning architectures as next generation security system foundations.</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

Intrusion Detection in Cloud Environment: A Survey of Hybrid CNN-RNN Framework

  • 期刊论文

  • 发表日期：

    2026年3月30日

查看全文
  • Parminder Singh,
  • Sudhanshu Prakash Tiwari,
  • Kanika Sharma,
  • Samriti .,
  • Ashu Bhardwaj

摘要

The recent growth in cloud computing has impacted industries betterment, but has also created complicated cybersecurity problems. In protecting the cloud infrastructure, IDS (intrusion detection systems) play a crucial role. The use of Advanced Persistent Threats (APTs) has rendered traditional methods and models ineffective along with the WID-ALL principle. To deal with this problem, deep learning techniques have been adopted by researchers, especially with the use of Convolutional Neural Networks (CNNs) as well as Recurrent Neural Networks, performing exceptionally well in traffic anomaly detection and classification. This article focuses on the performance hybrid frameworks of CNN and RNN improves in temporal and spatial feature learning related to detection in cloud environments. The paper focuses on the discussion of the architectural components of CNNs and RNNs, describing their individual strengths and the benefits gained through hybridization. Moreover, it explains the problems with implementing such systems in actual cloud infrastructure, particularly about scalability, probative value, interpretability of results, and availability of datasets. This review provides an analysis of the cloud that could help in devising intelligent, adaptive, and scalable IDS systems by studying emerging trends and possible research areas. Such findings enhance understanding of hyper deep learning architectures as next generation security system foundations.