Capacity dimensioning of VPN access links for elastic traffic

This paper discusses research into the capacity dimensioning of Virtual Private Network (VPN) access links for elastic traffic, such as the Web or ftp. Assuming that the coreVPN network is provisioned with a sufficiently large capacity, managing the capacity of the VPN access link comes to sharing the bandwidth for the elastic traffic of the two bottlenecks, the ingress and egress access links. In the case of a single bottleneck with a limited capacity for access links, the processorsharing model provides a simple formula for mean transfer time, but here, the value may be less than the actual transfer time because multiple flow may compete the bandwidth of both ingress and egress links. In contrast, max-min fair sharing provides an accurate sharing model which is similar to the TCP, but it is difficult to obtain a closed form for performance statistics. We propose a closed form approximation for a max-min fair sharing model, within a specific but realistic topology, through an investigation into the difference between the max-min and the processor sharing model. Using approximation, we calculate the capacity dimensioning of VPN access links. key words: VPN, hose model, max-min fair sharing