Abstract: Efficient cloud resource management is vital for optimizing system performance and ensuring balanced workloads across servers. Effective load balancing not only improves resource utilization but also enhances throughput and reduces response times, which are critical for achieving high availability and fault tolerance in cloud environments. Traditional job scheduling strategies often struggle to prioritize tasks with the same priority and to allocate jobs to virtual machines optimally, leading to performance inefficiencies. Despite extensive research, many existing scheduling algorithms fail to provide optimal solutions consistently.

This study proposes a Bio-Inspired Cat Swarm Optimization (CSO) approach integrated with a modified K-means clustering technique to address the shortcomings of current load balancing methods. The CSO method mimics the natural behavior of cats to search for optimal solutions, while the modified K-means clustering ensures efficient grouping and prioritization of tasks. The new priority-based scheduling algorithm introduced in this research aims to eliminate the drawbacks of existing systems, thereby enhancing the overall performance and efficiency of cloud computing. This approach not only improves resource allocation but also ensures a more balanced and resilient cloud infrastructure, capable of meeting the increasing demands of users and applications.

Keywords: Cloud computing, load balancing, resource management, Cat Swarm Optimization, K-means clustering, job scheduling, high availability, performance optimization.

| DOI: 10.17148/IARJSET.2024.11769