The robustness and integrity of IP networks require efficient tools fortraffic monitoring and analysis, which scale well with traffic volume andnetwork size. We address the problem of optimal large-scale flow monitoring ofcomputer networks under resource constraints. We propose a stochasticoptimization framework where traffic measurements are done by exploiting thespatial (across network links) and temporal relationship of traffic flows.Specifically, given the network topology, the state-space characterization ofnetwork flows and sampling constraints at each monitoring station, we seek anoptimal packet sampling strategy that yields the best traffic volume estimationfor all flows of the network. The optimal sampling design is the result of aconcave minimization problem; then, Kalman filtering is employed to yield asequence of traffic estimates for each network flow. We evaluate our algorithmusing real-world Internet2 data.
展开▼
