Aswad, S.A.Sonuc, E.2024-09-292024-09-292020978-172819090-7https://doi.org/10.1109/ISMSIT50672.2020.9254893https://hdl.handle.net/20.500.14619/93464th International Symposium on Multidisciplinary Studies and Innovative Technologies, ISMSIT 2020 -- 22 October 2020 through 24 October 2020 -- Istanbul -- 165025This paper classifies the VPN network traffic flow using the time related features on the Apache Spark and artificial neural networks. Today's, internet traffic is encrypted using protocols like VPN/Non-VPN. This situation prevents the classic deep packet inspection approaches by analyzing packet payloads. For the implementation of this research, MATLAB 2019b would be forwarded in use as increasing demand for VPN networks has actuated the evolutionary technology. The proposed method will prevent unnecessary processing as well as flooding found in standard VPN network traffic classification. As the proposed system is trained on 80 of the dataset while 20% is kept for the testing and validation with 10-cross fold validation as well as 50 epochs of training. To the best of our knowledge, this is the first study that introduces and utilizes artificial neural networks and apache spark engine to implement the classification of VPN network traffic flow. The accuracy of the VPN classification using ANN and Apache Spark Engine is 96.76%. The accuracy of the Non-VPN classification using the proposed method is 92.56%. This study has shown that an approach using the CIC-Darknet2020 for packet-level encrypted traffic classification cannot incorporate packet header information, as it allows to directly map a packet to a specific application with high accuracy. Considering only non-VPN traffic, 96.76% of all packets in the dataset can be associated with an application. The remaining packets can still be classified with high probability by predicting based on the applications that use this flow. © 2020 IEEE.eninfo:eu-repo/semantics/closedAccessANNapache sparkclassificationinternetmachine learningNetworktraffic-flowVPNConference Object10.1109/ISMSIT50672.2020.92548932-s2.0-85097665927N/A