A discrete-event queuing simulator built as a domain module for an existing Policy Evaluation Framework.
The goal is to model real-world queueing situations (e.g., single-server, multi-server, priorities, routing) and evaluate different policies/strategies using the framework’s experiment + evaluation tooling.
- Implement queueing environments/simulators (state, transitions, events)
- Define queueing decision points (e.g., which job to serve next, routing, admission control)
- Produce metrics that the framework can evaluate (wait time, latency, throughput, utilization, fairness)
- A new policy evaluation framework
- A UI app or production scheduling system
- Queueing domain types (jobs, servers, queues, events)
- Deterministic baseline environments (easy to test)
- Stochastic workloads (configurable RNG seeds)
- Policy hooks (FCFS, SJF, Priority, Round Robin, etc.) implemented through the framework’s policy interface
- Experiment configs + reproducible runs