Hybrid Fuzzy Metaheuristic Technique for Efficient VM Selection and Migration in Cloud Data Centers

The rapid expansion of cloud computing has made maintaining Quality of Service (QoS) across dynamic workloads essential. Virtual machine (VM) migration is crucial for optimizing resource management; however, traditional migration techniques, which rely on static parameters, often lead to inefficiencies, such as increased energy consumption, higher migration costs, and suboptimal resource utilization. To address these challenges, a novel fuzzy-based hybrid optimization technique, FCSFFC, is proposed, integrating fuzzy logic with advanced optimization methods, Cuckoo Search and Firefly Colony Optimization. This technique introduces a dynamic threshold-based load prediction mechanism that adapts to real-time conditions, ensuring efficient VM placement and migration. The performance of this algorithm was rigorously evaluated using real workload data in a CloudSim simulation environment. Compared to state-of-the-art algorithms, the proposed approach demonstrated a 31.7% improvement in migration cost, achieving the lowest migration cost. Additionally, the proposed approach achieved the lowest energy consumption, using 2% less energy than other methods. In terms of load management and resource availability, the algorithm showed a significant reduction in the load parameter and the highest resource availability, minimizing unnecessary migrations. It also achieved the shortest computation time, completing tasks in 4.003 seconds compared to up to 9.2 seconds for traditional techniques. These results underscore the effectiveness of the proposed method in enhancing cloud service efficiency by optimizing energy consumption, reducing migration costs, and improving overall system performance. © 2024 Slovene Society Informatika. All rights reserved.