COMPARATIVE STUDY OF LOAD BALANCING ALGORITHMS IN CLOUD COMPUTING

Abstract

Cloud computing has emerged as a transformative paradigm, offering scalable and flexible resource provisioning across diverse applications. As demand for cloud services intensifies, efficient load balancing becomes essential to ensure optimal performance, resource utilization, and user satisfaction. This paper presents a comparative study of eight load balancing algorithms—Round Robin, Random, ZHCJ (Dynamic Weighted Algorithm), LIF (Least Initiated First), ZHJZ (Adaptive Feedback Algorithm), ONWID (Optimized Node Weight with Intelligent Distribution), OFWID (Optimized Feedback with Intelligent Distribution), and MBFD (Modified Best Fit Decreasing). These algorithms are evaluated based on key performance metrics including response time, throughput, fault tolerance, and scalability. The study categorizes the algorithms into static, dynamic, and hybrid models, analyzing their behavior under varying workloads and cloud service models (IaaS, PaaS, SaaS). Experimental simulations and theoretical assessments reveal that while static algorithms offer simplicity, dynamic and hybrid approaches demonstrate superior adaptability and efficiency. This comparative analysis aims to guide cloud architects and researchers in selecting appropriate load balancing strategies to enhance cloud infrastructure performance.