Recursive computation of steady state probabilities in M/M/1/B queues

Several aspects related to the performance of wireless networks have been traditionally modeled with queueing theory. This paper presents a recursive procedure to find the steady state probabilities in an M/M/1/B queue from the steady state probabilities of an M/M/1/(B-1) queue. The procedure provides an alternative approach for the computation of state probabilities and provides better insight on how those probabilities change when the buffer size increases. A formula is developed for computing the relative decrease in the state probabilities when the buffer size is increased by one.