Theoretical and experimental analysis of MSES and Kshemkalyani-Singhal algorithms under distributed simulation environment

Abstract

In conservative distributed simulation systems, data synchronization can be achieved by using time stamps or causality tracking. Message ordering techniques based on time-stamps usually cause serious performance problems by impeding possible concurrency. On the other hand, causality based methods introduces minimal delay on message processing but requires large control information to be appended to messages. Both transmission and handling of control information are drawbacks of such methods. In this study, two causality based message ordering algorithms MSES (Modified Schipper-Egglie-Sandoz) and KS (Kshemkalyani-Singhal) are studied. The algorithms are implemented and tested on a common simulation environment. Both theoretical and experimental analysis and comparison of the algorithms are presented. © 2009 Elsevier B.V., All rights reserved.