Author: Yasir Zaki, Carmelita Görg
Short Description:
In simulations it is often a mistake to use correlated random number generators. Normally more than one stochastic distribution are needed to be used in a simulation and if the base random number generators for these distributions are not chosen in a correct way that doesn’t have correlation in between the simulation results would be completely wrong. In the attached presentation there is one example that shows the impact of using correlated and uncorrelated RNGs on the results.
Keywords:
Correlated random number generators
M/M/1 queuing system
Download as PDF