Load Balancing Algorithm: A Markov Chain Model Using Queuing Theory Approach

This study addresses efficient task scheduling in data centers, crucial for optimal resource use and timely task completion. It explores whether a custom algorithm can enhance load balancing efficiency compared to traditional algorithms like Round Robin. The study uses Markov Chain Approach simulations to evaluate the proposed algorithm's performance. Results show that while Round Robin's execution time grows exponentially, the proposed algorithm exhibits a more controlled increase, mitigating rapid execution time growth. A strong negative correlation was found between service rate and queue length. By analyzing parameters such as server utilization, arrival rate, average service time, total server capacity, and queue length, the study identifies the best loadbalancing approach. The practical implication is that organizations can implement the proposed algorithm to boost their Data Center's operational efficiency. The study's originality lies in its unique evaluation and comparison approach, proving invaluable for professionals and researchers in data center management.