Comparative Study of LST and SJF Scheduling Algorithm in Soft Real-Time System with its Implementation and Analysis

Abstract

The Least Slack Time First (LST) algorithm is a dynamic scheduling algorithm and also known as Least Laxity First. It decides the dynamic priority of the task based on slack time; The task having minimum slack time will be considered the highest priority. It is the most suitable algorithm for scheduling of tasks in soft Real-Time Operating System (RTOS). The Shortest Job First (SJF) algorithm is a static scheduling algorithm and decides the priority of the task based on execution time required for a given task. Task which has minimum execution time considered as the highest priority task in SJF. It is not directly used for scheduling Soft Real-Time system. In this paper, we have implemented the LST and SJF for the soft real-time operating system. These algorithms have been executed on periodic task set, and observations are gathered. We have observed Success Ratio & Effective CPU Utilization and compared both the algorithm in the same conditions. It is noted that the LST algorithm performs well in underload scenario but not well in an overload situation. SJF not able to schedule specific task even in underload situation but it comparatively performs well in an overload situation. Practical experiments have been conducted on a large dataset. Data Set contains 7500 task set, and each task set includes 1 to 9 processes. CPU load for each process set varies from 0.5 to 5. It has been tested on 500-time unit to validate the correctness of both algorithms.