Enhanced Weighted Round Robin: A New Paradigm in Cloud Load Balancing

Abstract

Objectives: The aim of this study is to suggest and recommend a load balancing algorithm based on enhanced weighted round robin (EWRR) technique to distribute incoming tasks or requests among various servers or resources in a cloud computing architecture. Methods: The study is accomplished by designing the enhanced weighted round robin (EWRR) load balancing and implemented in Cloud infrastructure by varying the number of jobs and virtual machines and evaluating the efficiency of the model. The experiments are simulated using CloudSim an open source tool and the effectiveness with regard to response time, idle time, task migration and delayed tasks are recorded. Findings: The obtained results are compared with round robin (RR) and weighted round robin (WRR) techniques and with the other state of art methods in the literature. The proposed EWRR performs better than all other approaches, such as Pragmatic Load Balancing (PLB) (28 ms) and Reinforcement Learning and Proficient Hybrid Lyrebird Falcon Optimization (RL-HFLO) (30 ms), with the shortest reaction time of 20 ms. Further, in comparison to greater values like 150 units in Round Robin and 120 units in K-means with RR, it reduces cumulative idle time to 75 units. Among all the methods examined, EWRR further reduces the number of delayed tasks to a mere 5. Even though EWRR involves 3 task migrations, which is slightly more than others, it results in better overall efficiency and utilization of resources. Novelty: The suggested EWRR model can be utilized as load balancing in cloud infrastructure in the varied cloud environments in a predictive way and enable the cloud service providers to make informed decisions. Dynamic parameters used by EWRR include each virtual machine's processing power, the quantity of incoming jobs, the duration of each job, and the VM's current burden (in real time). All of these EWRR settings produce intelligent scheduling that takes workload into account by preventing the blind assignment of jobs in RR and LC. Keywords: Load Balancing, Weighted Round Robin, Enhanced Weighted Round Robin, Cloud Computing, Performance Evaluation