A dynamic-server queuing simulation

We proposea new dynamic-server queuing model that increases system efficiency and customer satisfaction compared with current practice. It fits multi-server systems sharing a common queue, typical in banks and airlines check-in among others. This is a proposal of a computer program. The program sends a signal periodically to add (or dismiss) a server when the ratio of customers waiting to the number of servers reaches (or falls below) a pre-specified threshold policy ratio. If this ratio is set equal to one for example, then each waiting customer is practically at most next in line to be served. Currently, management adds or dismisses a server subjectively, which is inefficient and lacks theoretical foundation. This work invites the development of a similar algorithm for multi-server, multi-queue systems. We develop a queuing simulation algorithm that systematically adjusts the number of servers in a system. It applies to systems with a single waiting line but multiple servers. We impose a few realistic assumptions then use the algorithm to simulate random Poisson arrivals and exponential service. The results show that the proposed algorithm increases the systems efficiency and customer satisfaction relative to existing models.