Discovery-based P2P data distribution

Using the peersim engine to simulate distribution of data with no infrastructure. The node connections are set up by discovery.

How to run it

Requirements

• java 1.8 or superior
• python3 with pandas, textable and matplotlib installed
• Coffee and patience.

Let's dive into it !

Run an experiment

Clone the repo on your computer (if you plan to contribute, a fork to your personnal account will make things easier) then compile the code. A makefile is provided to ease things.

make all will compile the code.

To run a simulation you'll need to provide a configuration to the simulation engine. This is done through a text file. make run CONFIG=example.txt will feed the engine with example.txt.

Now, try with make run CONFIG=myfile.txt. I see numbers ! Yes, these are the simulation time, it gives you an idea of how fast the simulation goes.

But what happened ?

A lot of data has been created in a folder called "raw_dat". These are text file with simultaion data. You could try to graph them ! make graph will create a folder called figs in which figures are created. Have a look ! You'll see how the nodes are mooving and their noighbors relationships. Try to fiddle a bit with the config file and experiment !