We imagine that customers arrive at the times of a

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

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: re are n people in the system when 138 CHAPTER 4. CONTINUOUS TIME MARKOV CHAINS the customer arrives, then the amount of time he needs to enter service has a gamma(n, µ) density, so using (2.12) in Chapter 2 ◆ ✓ ◆n 1 X✓ µn xn 1 f (x) = 1 e µx µ µ (n 1)! n=1 Changing variables m = n 1 and rearranging, the above becomes ✓ ◆ 1 X m xm =1 e µx = (µ )e (µ )x µ m! µ m=0 Recalling that P (TQ > 0) = /µ, we can see that the last result says that the conditional distribution of TQ given that TQ > 0 is exponential with rate µ . From this we see that WQ = ETQ = 1 · µµ To compare with the Pollaczek-Khintchine formula, (3.7), we note that the service times si have Es2 /2 = 1/µ2 to conclude: i WQ = E (s2 /2) /µ2 1 i = =· 1 E si 1 /µ µµ With the waiting time in the queue calculated, we can see that the average waiting time in the system is W = WQ + Esi = 1 · µµ + 1µ · µµ = 1 µ To get this result using Little’s formula L = W we note that the queue length...
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