/lattice-attack

Lattice ECDSA attack

Primary LanguagePythonGNU General Public License v3.0GPL-3.0

Lattice ECDSA Attack

Recover an ECDSA private key from partial "k" nonces data.

The partial "k" information can be recovered from side channels : timing gives its size (leading 0), modular operations can give parity, ...

This software requires at least 4 known bits per nonce : LSB (last bits known) or MSB (first bits known). It computes what is the minimum amount of signatures data required to recover the key, before performing the recovery.

It uses linear matrices and lattice basis reduction to solve a Shortest Vector Problem from a Hidden Number Problem.

Requirements

  • Python3
  • cryptography library
  • fpylll library

cryptography : pip3 install cryptography or apt install python3-cryptography

fpylll : doesn't work in Windows, only in Linux.
-> apt install python3-fpylll

We recommend that you install fpylll with the distribution package manager. Else this requires lots of compilation tools and Python low level libraries. The package managers, such as apt or dnf, are providing all these, with pre-compiled binaries.

Use

Read a JSON data file with the signatures, with the following keys :

  • The name of the curve, see CURVES_ORDER in ecdsa_lib.py
  • Target public key is given as list of the coordinates : [x,y]
  • Message given as bytes integers list
  • The type of bits known : "LSB" or "MSB"
  • The number of known bits per secret k
  • The signatures as a list of dictionaries, with integers values : {"r": intR, "s": intS, "kp": knownNoncePart }
{
    "curve": curveString,
    "public_key": [pubx, puby],
    "message": [a,b,c,...], // In case same message for all signatures
    "known_type": "LSB"/"MSB",
    "known_bits": 6
    "signatures": [ {"r": intR, "s": intS, "kp": leakednoncepart }, {...}, ... ]
}

2^known_bits is a upper bound for kp.
Example if the LSB known for "k" are 0b000101 for a given signature
-> { "hash": xyz, "r": xxx, "s": xxx, "kp": 5 }

"hash" needs to be provided as integer in the signatures data when there's no "message" key. That means each signature has its own hash.

MSB shall be provided reduced like LSB, means only the known bits :
0b000101... -> "kp": 5
If the known bits are all 0 : "kp": 0

Run :

python3 lattice_attack.py [-f data.json]

There's a demo mode, provided by gen_data.py. Before performing an attack, it generates a random EC key pair (private/public), signs many messages with this key, and builds the data file accordingly ("data.json" by default).

Call lattice_attack and gen_data with the "-h" argument to see more about the options.

Real data

A real data set is provided in the Data/ECDummyRPA directory.

Bibliography

These related publications helped to develop this software. You can also find out deeper insights about the underlying mathematics used.

Hardness of Computing the Most Significant Bits of Secret Keys in Diffie-Hellman and Related Schemes
by Dan Boneh and Ramarathnam Venkatesan
CRYPTO '96: Proceedings of the 16th Annual International Cryptology Conference on Advances in Cryptology, August 1996
PDF Link

Return of the Hidden Number Problem
A Widespread and Novel Key Extraction Attack on ECDSA and DSA
by Kegan Ryan
IACR Transactions on Cryptographic Hardware and Embedded Systems (TCHES), December 2018
PDF Link

Minerva : The curse of ECDSA nonces
Systematic analysis of lattice attacks on noisy leakage of bit-length of ECDSA nonces
by Ján Jančár, Vladimír Sedláček1, Petr Švenda and Marek Sýs
Cryptographic Hardware and Embedded Security (CHES 2020), September 2020
PDF Link

Biased Nonce Sense : Lattice Attacks Against Weak ECDSA Signatures in Cryptocurrencies
by Joachim Breitner and Nadia Heninger
Financial Cryptography and Data Security 23rd International Conference, January 2019
PDF Link

A Side Journey to Titan
Side-Channel Attack on the Google Titan Security Key
by Victor Lomne and Thomas Roche
January 2021
PDF Link

Return of ECC dummy point additions
Simple Power Analysis on efficient P-256 implementation
by Andy Russon
SSTIC 2021, June 2021
PDF Link

License

Copyright (C) 2021 Antoine FERRON - BitLogiK

This program is free software: you can redistribute it and/or modify it under the terms of the GNU General Public License as published by the Free Software Foundation, version 3 of the License.

This program is distributed in the hope that it will be useful, but WITHOUT ANY WARRANTY; without even the implied warranty of MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU General Public License for more details.

You should have received a copy of the GNU General Public License along with this program. If not, see http://www.gnu.org/licenses/.