Mean Value Analysis for Blocking Queueing Networks

Mean value analysis is an exact solution technique for infinite capacity queueing networks and enjoyed widespread popularity during recent years. It considers the behavior of the system by stepwise increasing the number of jobs in the entire network, thus it is well-suited for the analysis of queueing networks with blocking. In this work, an approximation is introduced for the mean value analysis of queueing networks with transfer blocking. The blocking occurs when a job after completing service at a station wants to join a station which is full. The job resides in the server of the source station until a place becomes available in the destination station. The approximation is based on the modification of mean residence times due to the blocking events that occur in the network. Several examples are executed in order to validate the approximate results.