Cross-Layer AI-Driven Resource Optimization for Secure and Resilient Distributed Computing Systems

Abstract

Distributed computing systems are increasingly deployed to support large-scale applications requiring high reliability, security, and efficient resource utilization. However, dynamic workloads, cyber threats, and heterogeneous computing environments pose significant challenges to conventional resource allocation mechanisms. This paper proposes a Cross-Layer AI-Driven Resource Optimization (CLAIRO) framework for secure and resilient distributed computing systems. The proposed architecture integrates machine learning-based workload prediction, reinforcement learning-enabled scheduling, adaptive security management, and cross-layer communication optimization to improve overall system performance. The framework enables coordinated optimization across the application, network, and infrastructure layers while ensuring resilience against failures and malicious attacks. Experimental analysis demonstrates that the proposed approach achieves 27.8% reduction in response time, 24.5% lower energy consumption, 31.6% improvement in throughput, and 96.8% security detection accuracy, outperforming existing resource management approaches. The results indicate that AI-assisted cross-layer optimization provides an effective solution for next-generation distributed computing environments.