Abstract: Simulating large scale Peer-to-Peer（P2P） networks efficiently is still challenging.In this paper，FALPS，a discrete event simulator for large scale network simulation，is designed.BackboneNet model is used in the simulator.Based on this model，the authors present three strategies to make FALPS efficient：event combination，piece discarding and event queue size control.Results from performance experiments show that except for its extra height speed and accuracy，FALPS reduces the memory consumption significantly.FALPS can be used to simulate a P2P system of 100 000 nodes in 1 664 s，only consuming about 60 MB memory.

Key words: network simulation, packet level, flow level, Peer-to-Peer（P2P）

摘要： 随着P2P技术的发展，传统的网络仿真平台已经不能满足研究需要。如何设计支持大规模P2P内容分发仿真的平台是亟待解决的问题。通过建立BackboneNet模型，并采取了“事件”合并、非尾片段丢弃、“事件队列”大小控制三个关键算法设计了一种用于大规模P2P内容分发系统的包级离散事件驱动网络仿真平台FALPS。该平台具有内存消耗低、速度快、精确度高的特点，可用于仿真具有105数量级节点规模的P2P系统。

关键词: 网络仿真, 包级, 流级, Peer-to-Peer（P2P）