vn and hence the set of arrival times t1 t2

This preview shows page 1. Sign up to view the full content.

This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: ng (A.22) we see that in the long run the number of calls in the system will be Poisson with mean Z1 (1 G(r)) dr = µ r =0 That is, the mean number in the system is the rate at which calls enter times their average duration. In the argument above we supposed that the system starts empty. Since the number of initial calls still in the system at time t decreases to 0 as t ! 1, the limiting result is true for any initial number of calls X0 . 2.4.2 Superposition Taking one Poisson process and splitting it into two or more by using an i.i.d. sequence Yi is called thinning. Going in the other direction and adding up a lot of independent processes is called superposition. Since a Poisson process can be split into independent Poisson processes, it should not be too surprising that when the independent Poisson processes are put together, the sum is Poisson with a rate equal to the sum of the rates. Theorem 2.13. Suppose N1 (t), . . . Nk (t) are independent Poisson processes with rates 1 , . . . , k , then N1 (t) + · · · + Nk (t) is a Poisson process with rate 1 + · · · + k....
View Full Document

This document was uploaded on 03/06/2014 for the course MATH 4740 at Cornell University (Engineering School).

Ask a homework question - tutors are online