Streaming Solutions for Fine-Grained Network Traff
no vote
Online network trafficmeasurements and analysis is critical for detecting and preventing anyreal-time anomalies in the network. We propose, implement, and evaluate anonline, adaptive measurement platform, which utilizes real-time traffic analysisresults to refine subsequent traffic measurements. Central to our solution isthe concept of Multi-Resolution Tiling (MRT), a heuristic approach thatperforms sequential analysis of traffic data to zoom into traffic subregions ofinterest. However, MRT is sensitive to transient traffic spikes. In this paper,we propose three novel traffic streaming algorithms that overcome thelimitations of MRT and can cater to varying degrees of computational andstorage budgets, detection latency, and accuracy of query response. We evaluateour streaming algorithms on a highly parallel and programmable hardware as wellas a traditional software-based platform. The algorithms demonstratesi