Open Access

ARTICLE

Enhancing Cloud Performance Using File Format Classifications

Muhammad Junaid^1,*, Adnan Sohail¹, Monagi H. Alkinani², Adeel Ahmed³, Mehmood Ahmed³, Faisal Rehman⁴

1 Department of Computing, Iqra University, Islamabad, Pakistan
2 Department of Computer Science and Artificial Intelligence, College of Computer Science and Engineering, University of Jeddah, Jeddah, Saudi Arabia
3 Department of Information Technology, The University of Haripur
4 Department of Computing, COMSATS University, Abbottabad Campus, Pakistan

* Corresponding Author: Muhammad Junaid. Email:

Computers, Materials & Continua 2022, 70(2), 3985-4007. https://doi.org/10.32604/cmc.2022.019962

Received 03 May 2021; Accepted 15 June 2021; Issue published 27 September 2021

Abstract

Metaheuristic approaches in cloud computing have shown significant results due to their multi-objective advantages. These approaches are now considering hybrid metaheuristics combining the relative optimized benefits of two or more algorithms resulting in the least tradeoffs among several factors. The critical factors such as execution time, throughput time, response time, energy consumption, SLA violations, communication overhead, makespan, and migration time need careful attention while designing such dynamic algorithms. To improve such factors, an optimized multi-objective hybrid algorithm is being proposed that combines the relative advantages of Cat Swarm Optimization (CSO) with machine learning classifiers such as Support Vector Machine (SVM). The adopted approach is based on SVM one to many classification models of machine learning that performs the classifications of various data format types in the cloud with best accuracy. In CSO, grouping phase is used to divide the data files as audio, video, image, and text which is further extended by polynomial Kernel function based on various input features and used for optimized load balancing. Overall, proposed approach works well and achieved performance efficiency in evaluated QoS metrics such as average energy consumption by 12%, migration time by 9%, and optimization time by 10%, in the presence of competitor baselines.

---

Keywords

---