This preview shows pages 1–2. Sign up to view the full content.
This preview has intentionally blurred sections. Sign up to view the full version.
View Full Document
Unformatted text preview: Urban OR Fall 2006 Problem Set 4 (Due: Wednesday, November 8, 2006) Problem 1 [This is a review problem on M/G/1 queues. We have already done the first four parts in class, as an example.] Consider a singleserver queueing system with infinite queue capacity. Arrivals of customers to this system occur in a Poisson manner at the rate of 36 per hour. The service times, S , of customers are mutually independent and their duration is uniformly distributed between 1 and 2 minutes, in other words, f S ( t ) = U [1, 2] in units of minutes. (1) Compute the expected waiting time in queue, W q , and the expected number of customers in the queue, L q , when this queueing system is in steady state. [Numerical answers are expected.] (2) Assume now that service times have a negative exponential probability density 2 function with f S ( t ) = 2 e 3 t for t 0 , in units of minutes. Repeat part (1), i.e., 3 compute numerically the values of W q and of L q for this case. (3) Assume now that service times are constant and all have duration equal to 1.5 minutes. Repeat part (1), i.e., compute numerically the values of W q and of L q for this case. (4) Using the G/G/1 bounds discussed in class, compute numerically upper and lower bounds for W q and for L q for the case of Part (1). (5) Return to the original case (uniform service times) from here until the end of the problem. Suppose that immediately after the completion of a service, exactly 2 customers are left in the system. Carefully write an expression for the probability that after the end of the next service, exactly 4 customers will be left in the system. Do NOT evaluate numerically this expression. (6) Repeat part (4) under the supposition that after the completion of a service, exactly 0 customers are left in the system. Carefully write an expression for the probability that after the end of the next service, exactly 4 customers will be left in the system. Do NOT evaluate numerically this expression. (7) Consider again part (5) where after the completion of a service, exactly 2 customers are left in the system. Carefully write an expression for the probability that TWO service completions later, exactly 4 customers will be left in the system. Do NOT evaluate numerically this expression. (8) Suppose that George begins observing the queueing system at a random time...
View
Full
Document
 Fall '06
 hansman

Click to edit the document details