Scholarlink

<p>Cloud computing is pervasive in storing and processing confidential information, yet traditional authentication mechanisms such as passwords and tokens remain vulnerable to attacks, and these vulnerabilities compromise data confidentiality, integrity, and user trust, Therefore, a critical need is to develop robust cloud security models that combine biometric-based authentication with adaptive crypto approaches along with advanced attack detection features for effectively guarding confidential information. This paper introduces a new Decision Q-learning Camellia Authentication Framework (DQCAF) cloud data security model that incorporates dynamic Q-learning-based key generation, Camellia encryption, and biometric authentication to ensure strong protection of sensitive cloud-stored information. In the designed system, plaintext data is encrypted with dynamically generated keys, and the attempted unauthorized access is tracked by a Decision Tree, which identifies the anomalous activity, such as infrequent login attempts, request volume, and IP trends. An additional biometric authentication verifies the user. Finally, the authenticated user decrypts the data. Moreover, security validation is performed for establishing the correctness of biometric-based authentication. The performance of the model is validated with few standard metrics and obtained better performance such as 95% throughput, 0.67 ms latency, 99.24% Authentication accuracy, 1.66% False acceptance rate (FAR) and 1.29% False rejection rate (FRR), which shows improvement of 15% in throughput, 2% improvement in accuracy and 2–3% improvement in FAR and FRR. The developed framework demonstrates its ability to provide scalable security in the cloud compared to existing cloud security methods.</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

An adaptive cloud security approach for improving privacy with authentication

  • 期刊论文

  • 发表日期：

    2026年2月2日

查看全文
  • Kavuri K. S. V. A. Satheesh,
  • T. Krishna Sree,
  • S. Mary Evanchalin

摘要

Cloud computing is pervasive in storing and processing confidential information, yet traditional authentication mechanisms such as passwords and tokens remain vulnerable to attacks, and these vulnerabilities compromise data confidentiality, integrity, and user trust, Therefore, a critical need is to develop robust cloud security models that combine biometric-based authentication with adaptive crypto approaches along with advanced attack detection features for effectively guarding confidential information. This paper introduces a new Decision Q-learning Camellia Authentication Framework (DQCAF) cloud data security model that incorporates dynamic Q-learning-based key generation, Camellia encryption, and biometric authentication to ensure strong protection of sensitive cloud-stored information. In the designed system, plaintext data is encrypted with dynamically generated keys, and the attempted unauthorized access is tracked by a Decision Tree, which identifies the anomalous activity, such as infrequent login attempts, request volume, and IP trends. An additional biometric authentication verifies the user. Finally, the authenticated user decrypts the data. Moreover, security validation is performed for establishing the correctness of biometric-based authentication. The performance of the model is validated with few standard metrics and obtained better performance such as 95% throughput, 0.67 ms latency, 99.24% Authentication accuracy, 1.66% False acceptance rate (FAR) and 1.29% False rejection rate (FRR), which shows improvement of 15% in throughput, 2% improvement in accuracy and 2–3% improvement in FAR and FRR. The developed framework demonstrates its ability to provide scalable security in the cloud compared to existing cloud security methods.