Q-SHIELD: a forward-secure hybrid ML-KEM/AEAD framework with session-aware key management for post-quantum network security
The emergence of large-scale quantum computers will render current cryptosystems like RSA and ECC useless; hence, communication protocols will have to move to quantum-resistant techniques. The paper introduces Q-SHIELD, a UI-driven hybrid protocol framework facilitating the transition from existing protocols to quantum resistance. Q-SHIELD integrates an ephemeral ML-KEM-768 key encapsulation (FIPS 203) authenticated by an ML-DSA-65 signature (FIPS 204), an HKDF-SHA-384 transcript-based key derivation function, and AES-256-GCM record encryption using deterministic nonces. Per-session KEM key pairs, erased after being decrypted, provide forward secrecy, while game-based security proof sketches its security as a function of its underlying components. Q-SHIELD is implemented in Python using Open Quantum Safe and PyCryptodome libraries and evaluated using replay UNSW-NB15, CIC-IDS2017 and MAWI datasets in terms of setup latency, throughput, communication overhead, memory, energy consumption and concurrency. In 50 trials, the KEM path latency was measured as 4.18 ms, 15.2% faster than Kyber-768 standalone (4.93 ms) and analysis at the operation level showed that the performance gains were due to the reduction in framework overhead (86.7%) rather than speed in lattice arithmetic. A power model based on execution times estimated a 23.9% reduction in key establishment energy consumption (20.4 vs 26.8 mJ). Scalability was tested with up to 64 concurrent sessions; larger deployments are analytical projections only.
Authors
- Satish Raosaheb Jondhale (ORCID: https://orcid.org/0000-0003-2908-5610)
- R Kanthavel
- Amruta Jondhale
- R Dhaya
- Manoj Wakchaure
Institutions
- Papua New Guinea University of Technology (PG)
Publication Details
- Journal
- Journal of Cyber Security Technology
- Published
- 2026-11-18
- DOI
- https://doi.org/10.1080/23742917.2026.2759014
- Primary Topic
- Cryptography and Data Security
- Type
- article
- Field-Weighted Citation Impact
- 0.00