Autonomic classification of IP traffic in an NFV-based network

Abstract

Network Function Virtualization (NFV) is an emerging solution that improves the flexibility, efficiency, and manageability of networks by leveraging virtualization and cloud computing technologies to run networked devices in software. The implemen-tation of NFV presents issues such as the introduction of new software components, bottleneck performance and monitoring of hidden traffic. A considerable amount of NFV traffic is invisible using traditional monitoring strategies because it does not hit a physical link. The implementation of autonomous management and supervised algorithms of Machine Learning (ML) become a key strategy to manage this hidden traffic.

In this undergraduate work, we focus on analyzing NFV traffic features in two test environments with different components and traffic generation. We perform a benchmarking of the performance of supervised ML algorithms concerning its efficiency; considering that the efficiency of the algorithms depends on the trade-off between the time-response and the precision achieved in the classification. The results show that the NaiveBayes and C4.5 algorithms reach values greater than 90.68 % in a response time range between 0.37 sec and 3 sec.