Variance reduction in the simulation of a stationary M/G/1 queuing system using refined descriptive sampling

This article analyzes and simulates the stationary FIFO M/G/1 queuing system with different service time distributions using Simple Random Sampling (SRS) and Refined Descriptive Sampling (RDS). Consequently, we design and realize a software under Linux using C language which establishes the performance measures of the M/G/1 queuing system, computes their variances and the variance reduction in order to compare both sampling methods. The simulation results demonstrate that RDS produces more accurate and efficient point estimates of the true parameters and can significantly improve performance sometimes by a huge variance reduction factor in the M/G/1 queue compared to SRS.