Title :
Eliminating event cancellation in discrete event simulation
Author :
Savage, Eric L. ; Schruben, Lee W.
Author_Institution :
Sch. of Oper. Res. & Ind. Eng., Cornell Univ., Ithaca, NY, USA
Abstract :
The cancellation of previously scheduled events not only results in a model running less efficiently, it precludes the application of some analysis techniques such as infinitesimal perturbation analysis. While same simulation languages (SIMSCRIPT, SIGMA) include an explicit facility for event cancellation, others do not (SLAM, GPSS, SIMAN). From computation theory, it is known that event cancellation is never necessary; but it is sometimes a convenient modeling technique. Unfortunately, there has been no general methodology developed for eliminating event cancellation from a simulation model. We present a simple general approach. Applications to two classical models where event cancellation is typically used serve as illustrations of the method
Keywords :
discrete event simulation; modelling; queueing theory; scheduling; simulation languages; computation theory; discrete event simulation; event cancellation elimination; infinitesimal perturbation analysis; modeling technique; previously scheduled event cancellation; simulation languages; Computation theory; Computational modeling; Customer service; Delay effects; Discrete event simulation; Logic; Mathematical model; Performance analysis; Scheduling; State-space methods;
Conference_Titel :
Simulation Conference Proceedings, 1995. Winter
Conference_Location :
Arlington, VA
Print_ISBN :
0-78033018-8
DOI :
10.1109/WSC.1995.478852