Creating a "wait point" or a roll-back mechanism

[Javi]

Hello, 

I would like to ask the community if there is a way to "go back in time" in simulations. I will elaborate the idea: 

I have built a new user interface for OMNeT++ in order to control the simulation from an external program (basically what is described in the manual in the "Embedding the kernel" section). Now, I would like to: once the simulation has reached time t1, change some parameters (some link's properties for instance), run the simulation until t2 ( with t2 > t1, obviously), check some other parameters and, depending on the results, continue the simulation from t2 or "re-run" the part from t1 to t2 with a different set of parameters. 

Since I am controlling the simulation externally, I have a cSimulation variable that basically "is" the simulation (if I understood correctly). So, I thought about the possibility to clon this variable (still dont know how) in order to keep a "freeze" version of the simulation in t1 and continue the simulation with the cloned version. 

obviously, an easy way of doing this is re-run the whole simulation from the beginning, but, given the application in which I want to use this, re-starting from the beginning would not be a solution

Does anybody see a clear way of doing this?

Any input would be greatly appreciated!

Best, 

Javier.