Page 386 - A First Course In Stochastic Models
P. 386

MULTI-SERVER QUEUES WITH POISSON INPUT            381

                This gives for P (z) the following explicit expression:

                                   c(1 − ρ)(1 − z)  c−1    z − z k  !

                             P (z) =    c λD(1−z)            ,  |z| ≤ 1.     (9.6.8)
                                    1 − z e           1 − z k
                                                 k=1
                This expression for P (z) allows for a direct application of the FFT method. It
                is important to have an accuracy check on the calculated complex roots z k and
                the subsequent calculations by the discrete FFT method. Such an accuracy check
                is provided by Little’s relation (9.6.5). Another accuracy check is obtained by
                calculating the average queue size L q both from formula (9.6.7) and from the
                                     ∞

                direct expression L q =  (j − c)p j .
                                     j=c
                Waiting-time probabilities
                In the paper of Crommelin (1932) an explicit expression has been derived for W q (x)
                in terms of an infinite alternating series. However, this explicit expression turns out
                to be of little computational use and is therefore not further discussed. It is possible
                to deduce a recursion scheme for W q (x) from Crommelin’s original derivation, but
                this recursion scheme is also hampered by numerical difficulties. It took more than
                sixty years before a satisfying solution was found for the computation of W q (x).
                An elegant and numerically stable algorithm was found by Franx (2001) using an
                ingenious argument. The following expression holds for W q (x):

                         kc−1                          j
                                     −λ(kD−x)  [λ(kD − x)]
                 W q (x) =   Q kc−1−j e                 ,  (k − 1)D ≤ x < kD (9.6.9)
                                                 j!
                          j=0
                for k = 1, 2, . . . , where

                                          c+j

                                     Q j =   p i ,  j = 0, 1, . . . .
                                          i=0
                The first step in the proof is to assume that the arriving customers are assigned
                in cyclic order to the servers: the customers with labels i, i + c, i + 2c, . . . are
                assigned to server i for i = 1, . . . , c (the nth arriving customer gets label n). This
                service discipline is not violating the assumption of service in order of arrival since
                the service times are deterministic. Denote by W j the waiting time in queue of the
                customer with label j. It is assumed that there is a single queue in front of all c
                servers. Fix now x > 0. Also fix the positive integer k by (k − 1)D ≤ x < kD.
                Next choose any integer n such that n > kc. Consider now the customers with the
                labels n − kc and n. Both customers are served by the same server. This server is
                called the marked server. To derive P (W n ≤ x), we condition upon the number
                of waiting customers in the queue just after the epoch at which the customer with
                label n − kc enters service with the marked server. Distinguish between two cases:
   381   382   383   384   385   386   387   388   389   390   391