FINGERTRAP: A Self-Defending Cryptographic Protocol for Network Communications

Abstract

Fingertrap is a network encryption and authentication protocol that extends the X3DH and Double Ratchet frameworks with three novel mechanisms inspired by the Chinese finger trap (zhĭ wăng): a friction ratchet that exponentially increases computational cost for each failed authentication attempt; a recursive annihilation protocol that irreversibly destroys all cryptographic state after a configurable failure threshold; and a commit-then-challenge handshake that requires a counterintuitive “inward” action for legitimate authentication. A bidirectional weave hash extends the Double Ratchet’s transcript binding to cover every message in both directions. Together, these mechanisms provide per-message forward secrecy, post-compromise security (self-healing), clock-free operation, and a self-destruct capability. The individual ingredients-client puzzles, key erasure, and ratcheting-each build on established lines of work; their combination into a single stateful protocol, in which failed authentication attempts cryptographically tighten the session state and ultimately destroy it, is not to our knowledge offered by deployed transport protocols such as TLS 1.3, Signal, or WireGuard. The design targets deployments in which interception or capture of a device implies endpoint compromise, such as Unmanned Aerial Vehicle (UAV) telemetry links and body-worn sensors, where denial of exploitation requires guaranteed loss of past and future session material. We describe the full protocol, provide game-based security arguments under an explicit adversarial model, give analytic cost estimates for the friction mechanism, analyse the denial-of-service surface and a two-layer mitigation strategy, and specify a post-quantum extension using hybrid X25519/ML-KEM-768 ratcheting.