Revisiting conservative time synchronization protocols in parallel and distributed simulation

摘要

Computer simulations have become an indispensable tool for the empirical study of large-scale systems.rnThe timely simulation of these systems, however, is not without its challenges. Simulators have to be ablernto harness the full computational power of modern multicore architectures through parallel execution andrnovercome the memory limitations of a single computer. In this paper, we evaluate the performance of arnparallel and distributed simulator using several conventional time synchronization protocols executed onrnmodern multicore hardware. In addition, we comprehensively analyze a hybrid approach, combining tworntraditional protocols, increasing robustness, and enabling improved performance in a wider range of simulationrnscenarios. Finally, an adaptive algorithm to automatically configure this hybrid protocol is introducedrnand evaluated, eliminating manual user intervention and further improving robustness with respect to varyingrnsimulation conditions.