In tactical scenarios, there is limited knowledge available about an adversary’s encrypted traffic. To improve traffic classification performance in these scenarios, a new modified naïve Bayes kernel classifier (MNBK) is proposed based on optimal weight-based kernel bandwidth selection. By generating several traffic types expected in modern tactical edge networks, we demonstrate that the proposed MNBK classifier not only improves classification performance on the existing classes, but also detects unknown traffic with very high accuracy, precision, and recall compared with the traditional classifiers. In addition, a real time learning model is proposed based on MNBK and applied to real time traffic classification.
Machine Learning-Based Traffic Classification of Wireless Traffic
01.05.2019
695135 byte
Aufsatz (Konferenz)
Elektronische Ressource
Englisch
GPS-Based Traffic Conditions Classification Using Machine Learning Approaches
Transportation Research Record | 2022
|Federal learning traffic prediction method based on traffic mode classification
Europäisches Patentamt | 2025
|Classification of Traffic Accident Severity Using Machine Learning Models
Springer Verlag | 2025
|WIRELESS ROADSIDE MACHINE, TRAFFIC COMMUNICATION SYSTEM, AND TRAFFIC COMMUNICATION METHOD
Europäisches Patentamt | 2022
|