HQ-CABC-ALS: A Hybrid QoS-Aware Chaotic Artificial Bee Colony With Adaptive Local Search For Efficient Cloud Task Scheduling

Task scheduling in cloud computing remains a challenging optimization problem due to the dynamic and heterogeneous nature of cloud environments, where efficient resource allocation is essential for maintaining Quality of Service (QoS). Although Artificial Bee Colony (ABC)-based scheduling approaches have shown promising performance, they often suffer from premature convergence, limited population diversity, and inadequate consideration of multiple QoS objectives. To address these limitations, this study proposes a Hybrid QoS-aware Chaotic ABC with Adaptive Local Search (HQ-CABC-ALS) framework. The proposed method introduces (1) chaotic population initialization to enhance solution diversity, (2) an adaptive local search mechanism integrated into the scout phase to improve exploitation, and (3) a QoS-aware multi-objective fitness function incorporating makespan, energy consumption, deadline violation, and task priority. The algorithm is evaluated using a Python-based simulation environment with varying workloads of 20 − 100 tasks and 6 − 10 virtual machines to model heterogeneous cloud scenarios. Performance is compared against benchmark algorithms, including ABC, Genetic Algorithm (GA), Particle Swarm Optimization (PSO), First-Come-First-Serve (FCFS), and Min-Min scheduling. Experimental results demonstrate that HQ-CABC-ALS achieves up to 42% reduction in makespan, 23% lower energy consumption, and 17% improvement in load balancing efficiency compared with baseline methods. These findings confirm the effectiveness of integrating chaotic search, QoS-aware optimization, and adaptive local search for robust and efficient cloud task scheduling.