A single server queue with random arrivals and balking: Confidence interval estimation (Q756337)

A single server works a shift of length L with customers arriving according to a Poisson process rate \(\lambda\) ; arriving customers finding the server busy leave without service and the server works overtime, if necessary, at the end of the shift to complete service of a customer being served at that time. For a single shift, distributional properties of the number of services (X), the overtime worked (Y), and the total server idle time (T) are derived both unconditionally and conditional on the number of arrivals (n), when the arrival times are not recorded. These are used to provide estimates and confidence intervals for parameters from a single realization of the queue over a shift. When the service times are fixed (w), estimates of w are given when (n,X) or (n,T) are observed, and it is noted that confidence intervals based on T are shorter than those based on X for all confidence coefficients.