CLOUDSHIELD: AN AI-DRIVEN INTELLIGENT FRAMEWORK FOR DETECTION OF MULTI-CLOUD BASED CYBER (MCBC) ATTACKS IN MULTI CLOUD ENVIRONMENTS
Keywords:
Multi-cloud security, MCBC attacks, Intrusion detection system, Machine learning, Deep learning, Anomaly detection, Cybersecurity framework, Cloud computing, AI based security, CloudShield,,Abstract
Rapid adoption of multi-cloud computing architectures has been a major factor in increasing scalability, resilience and service availability for modern enterprises. However, the distributed and heterogeneous nature of multi-cloud environments leads to complex cybersecurity challenges, in particular, Multi-Cloud Based Cyber (MCBC) attacks that exploit inter-cloud dependencies, inconsistent security policies, and cross-platform vulnerabilities. Traditional intrusion detection systems (IDS) have limited ability to detect such coordinated and evolving threats because they rely on static rule sets and visibility into a single environment.
In this paper, we propose CloudShield, an intelligent AI-driven framework for real-time detection and mitigation of MCBC attacks in multi-cloud infrastructures. CloudShield is built on a five-layer architecture which consists of data acquisition, pre-processing and
normalization, feature engineering, intelligent detection, and automated response orchestration. The detection engine uses supervised learning models (Random Forest, XGBoost), unsupervised anomaly detection methods (Isolation Forest, One-Class SVM) and sequential deep learning models (LSTM networks) to detect known and unknown attack patterns. The framework also has an adaptive response module that leverages risk-based prioritization and reinforcement-inspired decision logic for automating mitigation actions such as access revocation, workload isolation, and traffic throttling. Specifically, the System is tested on the
standard datasets of cybersecurity (i.e. CICIDS2017, UNSW-NB15) for attack simulation in multi-clouds. Thus, compare the results of the analysis to determine that the CloudShield architecture presents an improved way for detection of attacks with better precision with reduced false positives and more rapid response times. All of this works along the two dimensions with intelligent cyber defense systems for multi-clouds and adaptive cyber defense systems models for future.
