The framework’s engine is based on an Event-Driven Architecture. The framework allows to build and configure simple models of real systems, conduct experiments, collect and analyze statistical data. Also, this framework contains a module allowing to build models of systems and networks from the Queueing Theory.
The framework contains the following modules
Implementing a simple timer through an event (HandledEvent) and its handler (TimeoutHandler).
The timer is activated at intervals that are distributed exponentially with some rate.
public static void simpleQueueingSystems() {
Environment env = new EnvironmentImpl();
Clock clock = new ClockImpl();
EventProvider eventProvider = new EventProviderImpl(Collections.emptyList());
SimulationContext context = new SimpleSimulationContext(env, clock, eventProvider);
SimulationModelImpl model = new SimulationModelImpl(context);
Queue queue = new QueueFIFO();
Event event = createDemandEvent(2, queue, context);
eventProvider.add(event, context.getCurrentTime());
model.setStopCondition(new TimeStopCondition(100000));
model.run();
}This is an example of a Queuing System type MM1.
Arrivals occur at rate λ according to a Poisson process. Service times are exponentially distributed with rate parameter μ so that 1/μ is the mean service time. A single server serves customers one at a time from the front of the queue, according to a first-come, first-served discipline. When the service is complete the customer leaves the queue and the number of customers in the system reduces by one.
Implementation through agent-based modeling.
public class AgentSimulationMM1 {
public static void main(String[] args) {
double simulationDuration = 10_000_000;
SimulationContext context = SimulationContextImpl.getContext();
double lambda = 1;
SourceAgent source = new SourceAgent(context,
new ExponentialRandomVariable(lambda),
"Source");
double mu = 2;
Queue queue = new QueueFIFO();
QueueingSystemAgent queueingSystem = new QueueingSystemAgent(context, queue,
new ExponentialRandomVariable(mu),
"Server"
);
source.setReceiver(queueingSystem);
AgentBasedSimulationModel agentSimulationMM1 = new AgentBasedSimulationModel(simulationDuration,
context);
agentSimulationMM1.addAgents(Arrays.asList(source, queueingSystem));
source.sendDemand();
long startTime = System.nanoTime();
agentSimulationMM1.run();
long endTime = System.nanoTime();
long timeElapsed = endTime - startTime;
System.out.println("Elapsed time = " + (double) (timeElapsed) / 1_000_000_000);
// correct answer = 1/(mu - lambda) = 1
System.out.println("Average service time is "
+ queueingSystem.getStatisticForVariable("SOJOURN_TIME").mean());
}
}