On stationary queue length distributions for G/M/s/r queues (Q1120916)

Consider a G/M/s/r queue, where the sequence \(\{A_ n\}^{\infty}_{n=-\infty}\) of nonnegative interarrival times is stationary and ergodic, and the service times \(S_ n\) are i.i.d. exponentially distributed. (Since \(A_ n=0\) is possible for some n, batch arrivals are included.) In case \(r<\infty\), a uniquely determined stationary process of the number of customers in the system is constructed. Furthermore, we give a proof of the relation \(\min (i,s)\bar p(i)=\rho p(i-1),\quad 1\leq i\leq r+s,\) between the time- and arrival- stationary probabilities \(\bar p(i)\) and p(i), respectively.