Cloud-Based Intelligent Load Balancing Algorithm for Efficient Resource Allocation in IaaS Cloud Computing
Ramisetti Pavan Kumar1, Dr. G. S. V. R. Abhishek2
1Department of Computer Science and Engineering, Bhimavaram Institute of Engineering and Technology, India
2 Department of Computer Science and Engineering, Bhimavaram Institute of Engineering and Technology, India
Abstract: Cloud Computing has become the leading approach for providing on-demand computing infrastructure over the Internet. Within this ecosystem, the Infrastructure as a Service (IaaS) model plays a central role by offering fundamental computing resources such as processing power, storage, and networking, all managed by Cloud Service Providers (CSPs). However, achieving efficient resource allocation and proper load balancing in IaaS environments remains a significant challenge. Uneven workload distribution can result in SLA violations, increased Makespan, reduced throughput, and an overall decline in user satisfaction.This paper introduces a cloud-based web application that implements an improved Load Balancing Algorithm (LBA) using Python 3.10, the Django framework, and a MySQL relational database. The proposed two-layer architecture integrates a Cloudlet Scheduler Time Shared mechanism for dynamic task allocation along with an SLA-aware Virtual Machine (VM) migration module that automatically adjusts MIPS allocations when deadline breaches are identified. Experimental comparisons with existing Dynamic LBA and Round Robin approaches show that the proposed solution achieves 78% resource utilization, lowers Makespan by 37%, and reduces SLA violations by 63%.The application features a comprehensive three-role management system—Task Scheduler, Cloud Administrator, and User—making it suitable for practical deployment in both academic and enterprise cloud environments.
Keywords: Cloud Computing, Cloudlet Scheduler, Django, IaaS, Load Balancing, Machine Learning, Makespan, MySQL, Python, Resource Allocation, SLA, Task Scheduling, Virtual Machine, Virtualization
