opp-repl-parameter-optimization

What problem does it solve?

This skill enables automated tuning of simulation parameters to achieve targeted outcomes using a derivative-free optimization approach (SciPy Nelder-Mead) within opp_repl-managed OMNeT++ simulations. It iterates over a single SimulationTask, adjusting parameter values until the expected result vectors are met, accommodating stochastic variability and non-differentiable models.

Core Features & Use Cases

• Automated parameter search: jointly tunes multiple parameters to meet specified scalars such as throughput, error rate, or utilization.
• Flexible task definition: works with get_simulation_task(...) and supports fixed_parameter_names/values, parameter_names/assignments, initial_values, min_values, max_values, and units to preserve semantics.
• Stochastic-aware optimization: designed for noisy simulations; guidance includes longer runs or repeated evaluations to stabilize results.
• Use Case: tune an OMNeT++ configuration until the observed metrics align with target values within an acceptable tolerance.

Quick Start

Configure a SimulationTask with your targets and call optimize_simulation_parameters to start tuning.