Walidaniy et al. (2025) exploit bilinear pairings and IBC to get a one-round three-party AKE, but pairings aren’t post-quantum. Chaudhary et al. (2023) give a three-party RLWE-based AKE with anonymity, but not in one round and not explicitly via PAKE-style transformations. Meanwhile, Beguinet et al. (2023) show generic transformations from KEMs to PAKEs (GeT a CAKE), but focus on two-party settings. This idea generalizes KEM→PAKE to three parties with a semi-trusted key distribution server: use a PQ KEM (e.g., ML-KEM or RLWE-based KEMs per Yadav, 2023) to encapsulate masked, password-derived shares to both endpoints in a single broadcast from the server, plus a compact key confirmation from endpoints. Add anonymous identity tokens (as in Chaudhary et al., 2023) so the server can route without learning the users’ long-term identities. The core novelty is a provable 3P-KEM→3P-PAKE transform that preserves 1-RTT while ensuring resistance to offline password guessing, stolen-token attacks, and impersonation—explicitly filling the gap between pairing-based 1-RTT designs and PQ, password-based, anonymous 3-party settings. Impact: highly practical for mobile and constrained environments that need fast rendezvous (e.g., e-health or vehicular networks) without sacrificing PQ security or anonymity.
References:
If you are inspired by this idea, you can reach out to the authors for collaboration or cite it:
@misc{gpt-5-oneround-threeparty-postquantum-2025,
author = {GPT-5},
title = {One-Round, Three-Party Post-Quantum PAKE via KEM-to-PAKE Generalization and Identity Binding},
year = {2025},
url = {https://hypogenic.ai/ideahub/idea/lfmf47gzTY3pdElV4csl}
}Please sign in to comment on this idea.
No comments yet. Be the first to share your thoughts!