PariRandom : Entropy distribution in a p2p network

Current software entropy generators bring entropy to a level which is not always suitable for its final use. There is a particular trade-off between the quality of entropy and the speed needed to generate it. We propose a novel method to improve the quality of entropy generators that leverages the ever growing phenomenon of Peer-to-peer networks. PariRandom is a pseudo-random number generation system that may be used to extend all other existing Pseudo Random Number Generation (PRNG) algorithms, ensuring they have an equal or greater level of entropy. An important aspect of our system is that it does not noticeably increase the underlying traffic, "piggybacking" instead on packets that would be transmitted anyway. The theoretical and experimental results demonstrate an increase in performance, which on a wide-scale, comes close to the performance achieved by hardware entropy generators. Moreover, they guarantee resistance to every kind of attack by malicious nodes in the network