"Hybrid Post-Quantum TLS 1.3 Network Flow Dataset for Anomaly Detection with Cryptographic Feature Augmentation"

"Dataset DescriptionThis dataset is a derived benchmark for anomaly detection in hybrid TLS 1.3 networks, created by integrating and augmenting two existing datasets: NF-CSE-CIC-IDS2018-v3 and CIC-PQC_OAV v1. It has been systematically enhanced with post-quantum cryptographic features to simulate realistic network conditions for the gradual adoption of post-quantum cryptography (PQC).The dataset is designed to support research on hybrid anomaly detection systems, in which classical TLS traffic is mixed with post-quantum traffic, and subtle anomalies must be distinguished from protocol-induced variations.Key Features and EnhancementsPost-Quantum Feature AugmentationNetwork flows are selectively enhanced with cryptographic features derived from CRYSTALS-Kyber operations.This allows models to capture both computational and communication characteristics of hybrid TLS handshakes.Probabilistic Post-Quantum InjectionEach flow is assigned a post-quantum indicator using a Bernoulli process, enabling mixed classical and PQC traffic to reflect realistic adoption scenarios.Extended Feature SetThe extended feature set includes is_pq, a binary indicator for post-quantum or classical flow; kyber_keygen_time, kyber_encap_time, and kyber_decap_time, which capture key generation, encapsulation, and decapsulation delays, respectively; kyber_pubkey_size and kyber_ciphertext_size, which denote the public key and ciphertext sizes in bytes; and hybrid_handshake_time and handshake_overhead_ratio, representing the total handshake duration (TLS + PQC) and relative latency overhead.\"Note: The research paper describing the creation and methodology of this dataset is currently under peer review. Users are encouraged to cite this dataset as hosted on IEEE DataPort and to acknowledge that it is derived from the original source datasets.For access to the dataset, don't hesitate to get in touch with [e-mail] or [e-mail].   "