Scholarlink

<p>The growth of the Information Technology (IT) sector comprises various risks and attacks on user data. The prevailing works did not focus on both attack detection and security of legitimate data on Virtual Machine (VM). Hence, in this work, the insider attack is detected and the non-attacked data is encrypted to improve the data security. Initially, the VM is registered on the cloud server and the generated QR code is updated on IPFS. Then, the QR code is hashed in IPFS using Gausigmoidalm-SWIFFT (G-SWIFFT). Meanwhile, the user registers and logs in to the system, and selects the required file for storing it in the cloud. Further, the signature is created using a Digital Signature Algorithm (DSA) for the hashed QR code by the user and sent to the cloud. The VM of the cloud verifies the signature using DSA. Then, the insider attack is detected in case the signature gets verified by the cloud. For attack detection, the dataset is preprocessed and then the word embedding is performed using the BERT algorithm. Further, the features are extracted, and the optimal features are selected using the Camkowsky-based Rat Swarm Optimization Algorithm (C-RSOA). Then, the attacked data is identified using the proposed Rootsig Ridsso-based Gated Recurrent Unit (R<sup>2</sup>GRU) model. Finally, the non-attacked data is encrypted using Leech Lattice–Based Secure Encryption (LL-SE) to ensure security and is stored in the cloud. The experimental outcome exhibits that the proposed methodology detects attacks with enhanced accuracy of 98.45% and precision of 98.74% than the existing techniques.</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

VM-Based Enhanced Data Security and Insider Attack Detection Using R2GRU and LL-SE

  • 期刊论文

  • 发表日期：

    2026年6月15日

查看全文
  • Surendar Rama Sitaraman,
  • Poovendran Alagarsundaram,
  • Kalyan Gattupalli,
  • Venkata Surya Bhavana Harish Gollavilli,
  • Harikumar Nagarajan,
  • Haris M. Khalid

摘要

The growth of the Information Technology (IT) sector comprises various risks and attacks on user data. The prevailing works did not focus on both attack detection and security of legitimate data on Virtual Machine (VM). Hence, in this work, the insider attack is detected and the non-attacked data is encrypted to improve the data security. Initially, the VM is registered on the cloud server and the generated QR code is updated on IPFS. Then, the QR code is hashed in IPFS using Gausigmoidalm-SWIFFT (G-SWIFFT). Meanwhile, the user registers and logs in to the system, and selects the required file for storing it in the cloud. Further, the signature is created using a Digital Signature Algorithm (DSA) for the hashed QR code by the user and sent to the cloud. The VM of the cloud verifies the signature using DSA. Then, the insider attack is detected in case the signature gets verified by the cloud. For attack detection, the dataset is preprocessed and then the word embedding is performed using the BERT algorithm. Further, the features are extracted, and the optimal features are selected using the Camkowsky-based Rat Swarm Optimization Algorithm (C-RSOA). Then, the attacked data is identified using the proposed Rootsig Ridsso-based Gated Recurrent Unit (R2GRU) model. Finally, the non-attacked data is encrypted using Leech Lattice–Based Secure Encryption (LL-SE) to ensure security and is stored in the cloud. The experimental outcome exhibits that the proposed methodology detects attacks with enhanced accuracy of 98.45% and precision of 98.74% than the existing techniques.