> Matthieu Latapy and Frédéric Aidouni
In most P2P file exchange systems, the file sizes are visible without downloading the files. Here, we recorded many (22 765 093) such file sizes and plotted, for each file size (in megabytes, horizontal axis), the number of files with this size (i.e. the file size distribution).
One observes that there are many small files (probably music files), and that there are clear peaks at 700 MB (typical size of a CD-ROM), at portions (1/2, 1/3, 1/4) of this value, or multiples of it (1.4 GB = 2 x 700 MB). The peak at 1 GB is still a mystery to me, even though I guess that it may indicate that users split very large files (DVD images for instance) into 1 GB pieces…
This plot reveals the fact that, even though in principle files exchanged in P2P systems may have any size, their actual sizes are strongly related to the space capacity of classical exchange and storage supports.
For more information, see our project on Measurement and Analysis of P2P Activity Against Paedophile Content