An Empirical Study on Multicriteria Scheduling

Abstract

This paper presents an empirical study of non-preemptive Multicriteria-Based, called MCB for short, scheduling policy. MCB scheduling policy uses multiple criteria of each request: arrival time, deadline, and processing time, to balance the requirements on both client and server sites. Weighted aggregation method is applied in this study to conduct the different measurements to a single figure of merit. For the empirical study, an M/G/1 queuing simulation system is implemented with MATLAB to represent a general server's incoming request scheduling system. Comparative simulation results of MCB with best effort scheduling policy on an overload situation show that MCB is an optimal scheduling policy.