Evaluation of a Local Adaptive Protocol for Distributed Discrete Event Simulation

Abstract

The performance of a local adaptive protocol for distributed discrete event simulation is evaluated. This protocol allows each process to adapt at runtime to its environment on a per channel basis with the possible range of behavior varying from conservative to optimistic. This evaluation includes a comparison of the performance of the algorithm with that of both a conservative and an optimistic protocol, as well as, an examination of a number of unique characteristics of the algorithm. Results from simulations using a network of workstations on a closed queueing system are presented. The results demonstrate the importance of various protocol features and show that it performs better than either the conservative or optimistic protocols in certain simulations.