FSRmSTS—An Optimize Task Scheduling with a Hybrid Approach: Integrating FCFS, SJF, and RR with Median Standard Time Slice

Cloud Computing enables multiple users to access a network of computing resources. To satisfy the expectations of cloud providers and consumers, it has become more focused on delivering Quality of Service (QoS). QoS is geared towards reducing task completion time, also referred to as makespan, and response time, while enhancing the efficiency of resource utilization. To achieve QoS, novel task scheduling strategies are employed, as traditional schedulers have often failed to meet the required standards. Several schedulers have prioritized reducing waiting or response times, without considering the impact on specific processes. To meet QoS requirements, new task scheduling techniques are frequently utilized. This article proposes the FSRmSTS task scheduling algorithm, which combines First Come First Serve (FCFS), Shortest-Job-First (SJF), and Round Robin (RR) with Median Standard Time Slice schedulers. In order to balance waiting times for both short and long tasks, it uses a dynamic task quantum. It also splits the ready queue into two sub-queues according to task duration. Tasks are assigned to resources from each sub-queue in a mutually exclusive manner. The performance of the proposed algorithm was evaluated using the CloudSim environment toolkit 3.0.3 and compared against six other scheduling algorithms, namely FCFS, SJF, RR, Improved RR, HFSR (Hybrid First Come First Serve, Shortest job first, Round Robin), and HFSR with Median. The results revealed that FSRmSTS outperformed the other algorithms by reducing waiting and turnaround time, and addressing the issue of long task starvation.