Scholarlink

<p>Cloud infrastructure development has created new security problems which demand real-time detection systems to fight complex cyber threats. The given manuscript introduces an Adaptive Cloud Security Risk Assessment Framework (ACSRF) which uses Pattern-Aware Countermeasure Optimization through a four-layered system namely (i) <i>Attack Pattern Identification Layer</i> that uses deep learning and graph-based anomaly detection to predict advanced attack methods, (ii) <i>Countermeasure Generation Layer</i> that generates real-time defense strategies through reinforcement learning (RL) with adversarial simulation (iii) <i>Ontology Design Layer</i> that selects optimal countermeasures by merging trust-weighted policy evaluation with multi-criteria decision-making (MCDCM) and (iv) <i>System Optimization and Validation Layer</i> that implements federated learning and Bayesian optimization (BO) for real-time policy optimization based on performance requirements. The proposed framework demonstrates superior performance than traditional rule-based and static defense systems through its ability to detect attacks with 3.7 times better accuracy and respond 2.8 times faster while delivering enhanced adaptive countermeasure effectiveness without affecting cloud system performance.</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 novel and optimized framework to assess the impact of intrusions in autonomous cloud computing environment

  • 期刊论文

  • 发表日期：

    2026年1月14日

查看全文
  • Banpreet Kaur,
  • Sunil Gupta

摘要

Cloud infrastructure development has created new security problems which demand real-time detection systems to fight complex cyber threats. The given manuscript introduces an Adaptive Cloud Security Risk Assessment Framework (ACSRF) which uses Pattern-Aware Countermeasure Optimization through a four-layered system namely (i) Attack Pattern Identification Layer that uses deep learning and graph-based anomaly detection to predict advanced attack methods, (ii) Countermeasure Generation Layer that generates real-time defense strategies through reinforcement learning (RL) with adversarial simulation (iii) Ontology Design Layer that selects optimal countermeasures by merging trust-weighted policy evaluation with multi-criteria decision-making (MCDCM) and (iv) System Optimization and Validation Layer that implements federated learning and Bayesian optimization (BO) for real-time policy optimization based on performance requirements. The proposed framework demonstrates superior performance than traditional rule-based and static defense systems through its ability to detect attacks with 3.7 times better accuracy and respond 2.8 times faster while delivering enhanced adaptive countermeasure effectiveness without affecting cloud system performance.