I wish I should have documented this code better. But alas it has been 4 years since I last worked on this. Probably a task for ChatGPT and alike someday. Till then archiving this.
I can help you if you have questions on a specific piece of code. Feel free to reach out on [email protected]
The main idea behind the code is to predict network polling intervals by looking at curvature instead of rate (double derivative instead of single derivative). Other details can be found in the paper An Efficient Method to Collect Statistics in SDN Using Curvature-Based Sampling
The code of this repository is not available under any open source license. Feel free to reach out in case you want to use some part of the code.