As a researcher in the P2P area, I'm writing my first paper about characterizaton and analysis of Gnutella networks and users.
My group and I performed some experiments in order to prove the presence of free loaders and to point out some problematic features on the gnutella protocol.
This is the list of experiments made:
- Evaluate the amount of files avaiable for download;
- Evaluate the number of QUERY HITs sent by the peers and make a relarionship with the demeanor of the users;
- Evaluate the number of PONGs for each PING and test the overhead in the network;
- Compare the amount of PINGs and QUERYs received by a peer;
- Evaluate the number of connections kept by a peer;
- Evaluate how long the session of a peer is;
- Characterize the user throught the kind of files he shares;
- Measure the popularity of files in gNet;

Well, my English is terrible I know, but I believe you can understand this.
If anyone have any other ideas, sugestions, doubts ... let me know !!!

Thanks