IP traffic classification in NFV: A benchmarking of supervised Machine Learning algorithms

The Network Functions Virtualization (NFV) is an emerging solution that improves the flexibility, efficiency, and manageability of networks by leveraging virtualization and cloud computing technologies to run network appliances in software. The implementation of NFV presents issues such as the introduction of new software components, bottleneck performance, and monitoring hidden traffic. A considerable amount of traffic in NFV is invisible using traditional monitoring strategies because it does not hit a physical link. The implementation of autonomic management and supervised algorithms from Machine Learning (ML) become a key strategy to manage this hidden traffic. In this work, we focus on analyzing the traffic features of an NFV-based network while performing a benchmarking of the behavior of supervised ML algorithms in the IP traffic classification regarding their efficiency; considering that the efficiency of an algorithm depends on the trade-off between the response time and the precision. Our results demonstrate the NaiveBayes algorithm as the best traffic classifier. NaiveBayes reaches values of 99.9% with precision in 1.1sec.