Parallel discrete event simulation with predictors

Due to the character of the original source materials and the nature of batch digitization, quality control issues may be present in this document. Please report any quality issues you encounter to digital@library.tamu.edu, referencing the URI of the item.Includes bibliographical references.Issued also on microfiche from Lange Micrographics.The motivation for this research has been its applicability in sequence checking in a spacecraft's control commands. Spacecrafts are controlled by sequences of time-tagged control commands which are essentially onboard computer programs. 'The sequence of control commands needs to be verified for correct execution before the commands are executed on the spacecraft to avoid the catastrophic effects of incorrect execution. The sequence checking employed (presently) is inherently sequential. The main objective of this work is to replace this existing sequential algorithm with a parallel implementation to gain performance. Parallel (distributed) discrete event simulation (PDES) refers to the execution of a single discrete event simulation program on a parallel computer. PDES mechanisms broadly fall into two categories : conservative mechanisms, like the Chandy-Misra approach and optimistic mechanisms, like the Time Warp mechanism. Conservative approaches strictly avoid the possibility of a causality error occurring whereas optimistic approaches use a detection and recovery approach wherein causality errors are detected and a rollback mechanism is invoked to recover. We present a technique for PDES that is based upon prediction techniques which performs very well for deterministic systems