A Task set based Adaptive Round Robin (TARR) scheduling algorithm for improving performance

CPU scheduling is one of the key concepts of any operating system for optimum use of resources. Round Robin scheduling is one of the efficient techniques but has a limitation that time quantum is static and must be taken fairly. Several works have been proposed to make it dynamic one for increasing the efficiency by decreasing the average waiting time, average turnaround time and context switching. The advantage of Round Robin and Dynamic Round Robin is that both are starvation free. The proposed Task set based Adaptive Round Robin (TARR) improves the performance of system in terms of average waiting time, average turnaround time and context switching. This method does not focus on entire ready queue but only a particular task set that is applicable for processing at an instance. Based on the available processes which are less than or equal to the arrival time at a particular instance, this method selects the time quantum from the task set and executes only that task set. The result obtained by the experimental setup clearly indicated that TARR gives improvement in the system performance.