关键词: 流量测量, 分层抽样, 累积业务流数量, 资源限制, NetFlow

Abstract: The technique of traffic sampling measurement is an important scalable solution in high-speed network.NetFlow is one of the applications which is widely deployed for traffic measurement.However，the sampling method of NetFlow has shortcomings.In order to overcome those deficiencies，this paper proposes a novel sketch called resource constraints adaptive packet sampling based on flow counting.Based on the idea of stratified sampling，the proposed sketch introduces a parameter called cumulative number of flows in order to control the memory resource.The easily-implemented packet sampling method presented can not only automatically adapt the sampling rate，but also give the right tradeoff between resource consumption and accuracy for all traffic mixes.Experiments are also conducted based on real network traces.Results demonstrate that the proposed method can achieve simplicity，adaptability and controllability of resource consumption without sacrificing accuracy compared with other sampling methods.

Key words: traffic measurement, stratified packet sampling, cumulative number of traffic flows, resource constraints, NetFlow