Analysis of DTLS Implementations ... State Fuzzing
summary
Extends TLS-Attacker with DTLS support and uses protocol state fuzzing to learn Mealy-machine models for 13 DTLS implementations, with models up to ~66 states and learning runs taking hours. Produces concrete workflow traces and finds 4 serious vulnerabilities, including a full client authentication bypass in JSSE, plus epoch/retransmission tricks that let unauthenticated or plaintext application data be accepted. Framework generates reproducible exploits against UDP DTLS servers and was responsibly disclosed.
AI-written summary · about
chapters (10)
- 0:00Introduction
- 0:51What is fuzzing
- 3:34What is TLS
- 11:03Models
- 22:13DTLS State
- 25:03Results
- 26:36Expert Knowledge
- 27:20Model of Pie
- 31:20Model of Jesse
- 38:58Demo of Jesse
related talks
01h 05mTheory and Security
45mAutomatic Discovery of Evasion Vulnerabilities Using Targeted Protocol Fuzzing
36mWhy TLS is better without STARTTLS ...
53mThe BEAST Wins Again: Why TLS Keeps Failing to Protect HTTP
20mFUZZVPN: Finding Vulnerabilities in OpenVPN
29mFuzzing and Breaking Security Functions of SIMATIC PLCs
16mLost in Translation: Enabling Confused Deputy Attacks on EDA Software...
13mX.509DoS: Exploiting and Detecting Denial-of-Service Vulnerabilities in
