A PARAMETERIZABLE METHODOLOGY FOR INTERNET TRAFFIC FLOW PROFILING

We present a parameterizable methodology for profiling Internet traffi c flows at a variety of granularities, Our methodology differs from ma ny previous studies that have concentrated on end-point definitions of hows in terms of state derived from observing the explicit opening an d closing of TCP connections. Instead, our model defines hows based on traffic satisfying various temporal acid spatial locality conditions, as observed at internal points of the network, This approach to how c haracterization helps address some central problems in networking base d on the Internet model, Among them are route caching, resource reserv ation at multiple service levels, usage based accounting, and the inte gration of IP traffic over an ATM fabric, We first define the paramete r space and then concentrate on metrics characterizing both individual flows as well as the aggregate flow profile, We consider various gran ularities of the definition of a flow, such as by destination network, host-pair, or host and port quadruple, We include some measurements b ased on case studies we undertook, which yield significant insights in to some aspects of Internet traffic, including demonstrating i) the br evity of a significant fraction of IP flows at a variety of traffic ag gregation granularities, ii) that the number of host-pair IP hows is n ot significantly larger than the number of destination network flows, and iii) that schemes for caching traffic information could significan tly benefit from using application information.