-
Sequence features dataset
The dataset consists of sequences of packet sizes, payload sizes, and inter-arrival times generated from the TLS handshake. -
Real User Dataset
The real-user dataset is used to evaluate the model on a traffic flow collected by a real user. It contains three sub-datasets: uni-label, binary-label, and multi-label. -
Supplementary Test Dataset
The supplementary dataset is used to test the model's performance on triples not seen during the training process. It contains 843 triple labels. -
Main Dataset
The dataset is used for multi-session website fingerprinting over TLS handshake. It contains 190 labels, each consisting of three websites from the twenty selected websites.