This framework simulates a modified version of the DiSC tool developed by Rasmus P. This is modified to support better network simulations, and the possibility for the controller/asset to run slower than the sampling time.
To install the framework, clone or download the git repository to a PC with MATLAB installed. The framework requires some consumption data which can be found as a zip-file via google drive:
https://drive.google.com/open?id=0B1d_5ZAxrCWacnIxc0ppU1hXWlk
In the zip-file the data is located in folders corresponding to where they must be placed in the repository. Further more the folders required are already included in the repository, and just needs to be populated. Therefore, the files located in the data folder, must be put in the data folder in the root of this repository. Similarly, the files located in DiSC/matlab/data folder must be placed in the DiSC/matlab/data folder of this repository.
The file RunAll.m runs through 3 nested for loops. The first one determines the access scheme. 1 means the reactive scheme, 2 means the proactive scheme. The second loop determines the seed for the simulation. The final loop determines the mean network delay as a multiple of the sampling time (Ts, default 5s), this delay will be applied in both upstream and downstream. Based on the parameters of these 3 loops, the script determines a file name in which to store data generated by the simulation. If a file by this name exists in the results folder, it will skip the simulation, and continue to the next parameter value in the loops.
The RunAll.m script is set to only create data using the reactive scheme, which is why the script RunAll0.m is the exact same, except that this runs the proactive scheme. The results folder is currently populated with empty data files as place holders for all the simulations that have already been done. The actual data should not be uploaded to github, but handled via some other service.
The makeFiles.m script can be used to generate empty data files, if new placeholders are needed.
The ShowResults.m script can be used to handle the data, and generate plots for visualisation.