Comparative Performance and Computational Complexity Analysis of Hybrid WGO–DRL and Heuristic, Metaheuristic and Bio-Inspired Scheduling Algorithms in Cloud Computing

Abstract

Cloud computing environments are becoming increasingly dynamic and heterogeneous, making efficient task scheduling a critical challenge for cloud service providers. Traditional heuristic scheduling methods such as Round Robin (RR), Min–Min, Max–Min, and First Come First Serve are commonly used because of their simplicity and low computational overhead. However, these approaches often struggle to adapt to fluctuating workloads and heterogeneous resource conditions, which may result in inefficient resource utilization, increased energy consumption, and longer execution times. To address these limitations, this study investigates a hybrid scheduling framework that integrates Hybrid Wild Goose Optimization (HWGO) with Deep Reinforcement Learning (DRL). In the proposed framework, the DRL component observes the current system state and generates adaptive scheduling decisions, while the HWGO algorithm refines these decisions through multi-objective optimization. The hybrid model aims to enhance scheduling efficiency by reducing makespan, minimizing energy consumption, balancing CPU load across virtual machines, and decreasing migration overhead. The performance of the HWGO–DRL framework is evaluated through a comparative analysis with conventional heuristic scheduling algorithms including(RR, FCFS, Min-Min, Max-Min), metaheuristic algorithms (GA, PSO, ACO), and bio-inspired algorithms (CSO, DFA, FDA, LOA) under identical cloud simulation conditions. Experimental results demonstrate that the hybrid approach achieves improved scheduling performance across multiple evaluation metrics. Notable improvements are observed in execution efficiency, workload distribution, and energy utilization. The proposed model achieves up to 30–40% reduction in makespan and 20–30% improvement in energy efficiency compared to baseline methods.. These findings indicate that intelligent hybrid optimization techniques can provide adaptive and efficient task scheduling solutions for modern cloud computing environments.