Stochastic differential equation-based general P2P file distribution model
基于随机微分方程的通用P2P文件分发模型*

This paper proposed a stochastic differential equation-based model to describe the dynamics of peer-to-peer(P2P)file distribution systems.The nodes of the system were divided into two categories according to the implementation of this kind of system. The dynamics of these two kinds of nodes was analyzed quantitatively. The proposed model employed an abstract neighbor selection policy to support different neighbor selection algorithms. The arrival rate of new nodes was a function of time, which could describe invariable and variable arrival rates. Compared with existing models and trace data, the proposed model is simpler, more general, and practical than other models. The new model is very useful for evaluation existing paradigms and designing new ones.