Welcome to the single queue simulator #1
The simulator enables analysis of single queue systems with single inputs
and single service without impatience and with state independent service time. To start simulation you have to specify the system
and to set simulation control unless you want to keep the default values.
If something is not clear go first back to the Simulator's page and read
Introduction to Queuing Systems (click the picture in the heading or the link).
Note also that Help buttons explain the meaning of all parameters and results.
If all this does not help, contact the author.
- Enter the distribution of intervals between arrivals. Either select a theoretical
distribution and enter its parameters or enter a user defined distribution in table form.
For details read the help by pressing the Help button.
- Enter the distribution of service duration in similar way, all channels are
- For models with limited population check the option and enter the total
number of customers. Then decide whether the intervals between arrivals should
be recalculated with respect to the number of customers in the system. Normally
intervals should be modified.
- For models with limited capacity check the option and enter the maximum queue length.
Note that it can be zero when no waiting is allowed.
- Select the queue organization, either FIFO or LIFO.
- Enter the number of channels. You can model ample service (self-service) by
entering a big number of channels. Then check whether the maximum queue length
is zero. If not, increase the number of channels. Maximum and average system
size is then the maximum and the average number of working channels.
There are two ways how to finish an experiment. If both are active, the experiment
is terminated by what comes first.
- Experiment duration is the maximum experiment length. This is the typical way
how to end experiments.
- Number of served customers can also be used as the termination condition. It may
happen that the experiment is terminated earlier. In this case increase the experiment
duration. The default value 0 means that the number of served customers is not tested.
- Enter the number of repetitions. Note that the seeds entered in the two
distribution tables are used only if there is a single repetition. Otherwise
the seeds are randomized before each experiment.
- Select the place for results.
- For long experiments uncheck the system status. If unchecked, only the run
number is reported.
- Start with shorter experiment duration to check the capabilities of your computer
and your browser. The experiment duration is related to the length of arrival
intervals and the service duration.
If these times are in units or tens, the default experiment length 1000 might be a reasonable value
to start with.
- The simulator is based on abstract time units whose interpretation depends on particular
simulated systems. All time values must of course be expressed in the same time units.
- If a user defined distribution is used, the table has to be confirmed that also checks the
table for consistency and offers some adjustments.
To start with a simple user defined distribution, try a uniform one.
For example the cumulative distribution defined by two points (0,0)
and (5,1) is the uniform distribution on the interval (0,5) - all values from this
interval are equally likely. Simulator is able to adjust distribution tables, but to allow
editing, the tables may be temporarily inconsistent.
- The simulator was tested with various browsers. To get a neat
screen the resolution should be at least 800x600. All comments,
especially from other than PC platforms will be very appreciated.