摘要： 根据网络上的流统计呈现很强的重尾分布的特性，该文提出了使用周期抽样和counting bloom filter(CBF)技术相结合的方法，即Sample-CBF方法来实现长流识别，并根据抽样策略的不同，将其具体化为两种方法：PSample-CBF方法和FSample-CBF方法。理论分析和仿真结果表明，在存在可容忍流长度测量误差的条件下，两种方法都可以准确识别长流，有效地减少存储空间和提高处理速度。

关键词: 周期抽样, 长流, 哈希

Abstract: According to the characteristics that the network flow statistics show a strong heavy-tail distribution, this paper puts forward a method using Sample-CBF to realize elephant flows identification, which combines period sampling and counting bloom filter(CBF) technique. According to the different sampling strategies, the method is translated into two specific methods: PSample-CBF method and FSample-CBF method. Theoretical analysis and simulation result indicate under the condition of existing some tolerable measurement error about the length of flows, the two methods both can identify elephant flows accurately, reduce the storage space and improve the processing speed efficiently.

Key words: Period sampling, elephant flow, Hash

中图分类号: 

• TP393.02