Deep Neural Classification of Darknet Traffic

Darknet is an encrypted portion of the internet for users who intend to hide their identity. Darknet’s anonymous nature makes it effective tool illegal online activities such as drug trafficking, terrorist activities, and dark marketplaces. traffic recognition essential in monitoring detection malicious activities. However, due anonymizing strategies used darknet conceal users’ identity, practically challenging. The state-of-the-art systems are empowered by artificial intelligence techniques segregate data. Since they rely on processed features balancing techniques, these suffer from low performance, inability discover hidden relations data, high computational complexity. In this paper, we propose a novel decision support system named Tor-VPN detector classify raw into four classes Tor, non-Tor, VPN, non-VPN. discovers complex non-linear our deep neural network architecture with 79 input neurons 6 layers. To evaluate performance proposed method, analyses conducted benchmark dataset DIDarknet. Our model outperforms classification accuracy 96%. These results demonstrate power handling without using any preprocessing like feature extraction or techniques.