Bottleneck Detection via Aggregate Rate Analysis : A Real Case in a 3G Network

We address the problem of inferring the presence of a bottleneck from passive measurement in a 3G core network. The work is based on two weeks of packet traces collected in the core network of Mobilkom Austria AG&CoKG, the leading mobile telecommunications provider in Austria, EU. During the measurement period a bottleneck link in the 3G core network was revealed and removed, therefore the traces enable the accurate analysis and comparison of the traffic behavior in the two network conditions. The proposed diagnosis approach is based on the marginal rate distribution of the aggregate traffic. Since TCP is the dominant component in the overall traffic mix, the presence of a bottleneck strains the aggregate rate distribution, compressing it against the bottleneck capacity limit during the peak hour. The analysis of the trajectories of the first central moments (variance, skewness) can reveal such strain without any a priori knowledge of the provisioned capacity.