VARIANCE REDUCTION FOR SENSITIVITY ESTIMATES OBTAINED FROM REGENERATIVE-SIMULATION

We apply the method of concomitant control variables to reduce the var iance of derivative estimates obtained from regenerative simulations u sing the likelihood ratio method. We restrict our attention to systems that contain a Poisson process. The control variables that we use ari se naturally from likelihood ratios. After deriving an expression for the optimal coefficients of the control variables, we numerically inve stigate their performance. Using six controls with an M/G/1 queue we a chieve a reduction in the standard deviation of the derivative estimat e of up to 60%.