Round Robin CPU Scheduling with Dynamic Quantum using Vague Sets

Abstract

Size of quantum is a key factor in round-robin scheduling. When the chosen quantum is very large, the algorithm degenerates to FCFS, while too short quantum leads to increased overheads because of more numbers of context switching. This work intends to select a suitable quantum dynamically in every round. A dynamic weighting policy that assigns weight in proportion to the remaining CPU burst timewould keep on changing the weights of the process in every round.The fuzzy technique for assigning weights has been used by previous authors. Membership values in fuzzy sets are based only on the consideration of the favorable evidences;for this reason,it may prove to be inadequate to cover up all the uncertainties.Vague sets are characterized by two membership values: truth membership and false membership value. These are therefore better capable of obtaining the combined effect of favorable and unfavorable evidences together.The present paper deals with vague sets to assign the weights to the processes and computes the quantum dynamically in every round. Numerical computations have been carried out for different sets of processes and the results have also being compared with fuzzy-based round-robin and conventional round-robin techniques.

Keywords: Vague Set, Truth Membership Function, False Membership Function, Vague Inference Engine, Context Switching.