Review on Network Traffic Measurements and Analysis

摘要

The network system usually falls into the complexity of preventing the real time anomalies in the created architecture, some streaming analysis and measurements are created for preventing the anomalies but it is critical for detecting and preventing those problems. The traditional mechanism of slow open-loop measurement scheme is not feasible for the high speed and complex networks. A new approach called Closed-loop measurement is introduced and demonstrated the practical realization and functionalities of an efficient network mechanism. Three streaming algorithms are proposed to provide a tight integration between the measurement platform and the measurements. The algorithms provide to varying degrees of computational budgets, detection latency, and accuracy. We empirically evaluate our streaming solutions on a highly parallel and programmable measurement platform. The algorithms demonstrate a marked 100% accuracy increase from a recently proposed MRT algorithm in detecting Denial of Service attacks made up of synthetic hard-to-track elephant flows. The proposed algorithms maintain the worst case complexities of the MRT, while empirically demonstrating a moderate increase in average resource utilization.