This discrete event simulator models queueing operations with various configurations. Follow these steps to begin:
The simulator is divided into three main panels:
Models random events with a constant average rate. Commonly used for arrival processes where events occur independently.
Events occur with equal probability within a specified range.
Models events that cluster around a mean value with symmetric variation.
Fixed, predictable intervals between events.
All customers join one queue, and go to the next available server. This typically minimizes average waiting time.
Customers are assigned to specific servers (often round-robin) and must wait in that queue even if other servers are free.
Customers join the queue with the fewest waiting customers. This attempts to balance load across servers dynamically.
| Metric | Interpretation | Ideal Range |
|---|---|---|
| Avg Waiting Time (Wq) | Time customers spend waiting in queue | As low as possible |
| Avg System Time (W) | Total time from arrival to departure | Close to service time |
| Throughput | Customers served per second | Higher is better |
| System Utilization | Percentage of server capacity being used | 70-90% (balance efficiency vs. wait times) |
Use the "Light load" preset to see a system with ample capacity. Notice how customers are served immediately with minimal waiting.
The "Heavy load" preset demonstrates a system near capacity. Observe how queues build up and waiting times increase significantly.
This scenario uses a normal distribution with high variability. Watch how the system handles sudden rushes of customers.
Try the same parameters with different routing policies. Notice how single queue typically outperforms separate queues.
This fundamental principle states: L = λW, where:
You can verify this relationship in the simulator by comparing the metrics.
This measures how busy the system is: ρ = λ/(cμ), where:
When ρ approaches 1, the system becomes unstable with growing queues.
Adjust the "Time scale" parameter. Lower values slow down the simulation for better observation.
Check that your arrival rate isn't significantly higher than your service capacity. The system needs λ < cμ to remain stable.
Verify that arrival parameters are reasonable (not extremely large values). Try resetting with a preset.
Toggle the event log to see a detailed timeline of customer arrivals, service starts, and departures.
Use the "Step" button to advance the simulation one event at a time for detailed analysis.
Monitor the queue length and average waiting time charts to identify trends and patterns over time.