MATLAB: How to reduce the simulation time of the model in SimRF 4.0 (R2013a)

RF Blockset

I have a very large and complex SimRF model. When I simulate my model, it takes a very long time to run. I would like to know how I can speed up the execution time of my model.

Best Answer

How do I reduce the simulation time of my model in SimRF 4.0 (R2013a)?
I have a very large and complex SimRF model. When I simulate my model, it takes a very long time to run. I would like to know how I can speed up the execution time of my model.
Depending on the specifics of your model, there are several ways that it may be possible to reduce the simulation time:
1) If possible, use Equivalent Baseband (EBB) instead of Circuit Envelope (CE). EBB is faster than CE, although the tradeoff is less accuracy and modeling flexibility.
2) If you are using the Normal Simulink simulation mode, please consider switching to Accelerator or Rapid Accelerator mode. Accelerator mode is available for both EBB and CE models, while Rapid Accelerator mode is only available for EBB. For more information on the Accelerator/Rapid Accelerator modes, please see the following link:
3) Try changing the solver settings to see if that reduces the simulation time. In general, it is recommended to use the stiff solvers for physical modeling. If your model is CE, try changing the local solver setting. For more information on solvers in general and the local solvers in particular, please see the following links:
4) Try to simplify the model itself:
- Combine series and parallel components into equivalent single components.
- Derive the equivalent S-parameter representation for complex subsystems and replace the subsystem with an S-parameter block with these values.
- If the model is CE, try to reduce the number of simulated frequency harmonics/carriers in the SimRF Configuration block (trading some accuracy for higher speed).
5) If the model is CE, please consider upgrading to MATLAB release R2013a or later. The simulation speed of CE in R2013a was significantly improved relative to R2012b and earlier releases.