Optimized Self-Guided Quantum Generative Adversarial Network Based Scheduling Framework for Efficient Resource Utilization in Cloud Computing to Enhance Performance and Reliability

Abstract

Cloud computing enables dynamic resource access, but efficient resource allocation remains challenging due to interference and performance limitations in virtual machine (VM) management. Efficient resource allocation in cloud computing is crucial for minimizing interference and optimizing virtual machine (VM) performance. This study proposes a Self-Guided Quantum Generative Adversarial Network with Prairie Dog Optimization Algorithm (SGQGAN-PDOA) to reallocate tasks across VMs dynamically. The framework integrates Inception Transformer (IT) for feature extraction and Spatial Distribution–Principal Component Analysis (SD-PCA) for feature reduction, enhancing processing efficiency. Implemented in Java with CloudSim, the proposed model improves resource utilization, achieving 80% reliability for 150 VMs with a 200 ms processing time. Experimental results demonstrate significant reductions in waiting time, response time, and load imbalance, outperforming existing methods. By leveraging quantum generative modeling and optimization, this approach enhances scalability, energy efficiency, and system responsiveness in dynamic cloud environments. The findings suggest that quantum-inspired scheduling frameworks offer a promising solution for adaptive and high-performance resource management in cloud computing.