With the popularity of wireless networks, as well as the continuousdevelopment of the mobile terminal, real-time fine-grained traffic monitoringfor mobile user is more and more important in wireless networks. However,the high granularity sampling of wireless user’s network traffic will degradesystem performance, while producing the transmission and storage of datawill increase system overhead.In this paper, based on the analysis of large-scale campus wireless LANuser traffic data, it is observed that wireless network traffic is high time andspace correlated by using information entropy to measure a large-scale realdata in campus wireless LAN. This paper presents a new traffic samplingscheme based on wireless user’s traffic spatial-temporal correlation to reducenetwork sampling frequencies of wireless user’s network traffic and ensurehigh accuracy. The experiment result proved that this scheme can reach morethan80%accuracy of sampling wireless user’s network traffic by using onlyless than20%of the sampling rate. Finally, we use an open source deeppacket inspection technology OpenDPI to analyze campus wireless networkdata, to verify the realiability algorithm based on compressed sensing indifferent application protocol. |