Analyzing the Sample Path: Analyzing the Sample Path

Analyzing the Sample Path

The QSIM application provides some built-in analysis capabilities. You can obtain sample means, variance estimates, and other simple statistics on several measures of performance, including time spent in system components and resource utilization. You can also display simple line plots, histograms, and box plots to dynamically observe system behavior.

An additional feature is the ability to accumulate the sample paths in SAS data sets. See Chapter 8, "Saving and Restoring," for more information. With these data in data sets, you can write your own programs to do more sophisticated analysis.

When calculating statistics, QSIM uses the following measures of system performance:

number in a queue
time in a queue
number of multiple-servers busy
server utilization
time in service
time in the system

In addition, statistics can be calculated on samples of random variables, which are user-defined attributes.

To illustrate some of the analysis features, this chapter uses the multiple-server, single-input queueing model shown in Figure 9.1.

Figure 9.1: A Single Queue 3-Server Model

The inter-arrival time distribution is ${\cal e}(.33)$ and the service time distribution for each server is ${\cal e}(1)$ . Transactions queue in a FIFO discipline and go to the first available server. This corresponds to the type of queueing that you encounter at check-in at many airports.

Analyzing the Sample Path

Running a Simulation

Statistics on Component State

Statistics on Transaction State