Ecosyste.ms: Awesome
An open API service indexing awesome lists of open source software.
https://github.com/jesseemond/random-prediction
Prediction of the outputs of poor random generators.
https://github.com/jesseemond/random-prediction
Last synced: about 1 month ago
JSON representation
Prediction of the outputs of poor random generators.
- Host: GitHub
- URL: https://github.com/jesseemond/random-prediction
- Owner: JesseEmond
- Created: 2015-06-04T05:01:12.000Z (over 9 years ago)
- Default Branch: master
- Last Pushed: 2017-05-12T14:30:48.000Z (over 7 years ago)
- Last Synced: 2024-03-18T05:56:04.144Z (8 months ago)
- Language: C++
- Homepage:
- Size: 17.6 KB
- Stars: 8
- Watchers: 3
- Forks: 1
- Open Issues: 0
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# random-prediction
Prediction of the outputs of poor random generators.
In the hopes that this might save someone's time while learning on the subject.
I tried to comment the parts that were not completely explicit for me from the links.
I did not bother to write classes as this was just a quick POC.
### Sample output
```
Predicting output for glibc's TYPE_1 random (default):
Success rate: 7493 / 10000 (0.7493)
Predicting output for glibc's TYPE_1 random (default) & 0x0F:
Success rate: 7504 / 10000 (0.7504)
Predicting output for linear congruential generator:
Success rate: 10000 / 10000 (1)
```
### How to compile
```
g++ predict.cpp --std=c++11 -o predict
```
## What this predicts
- glibc's random() TYPE_1 (additive feedback generator)
- [info about the maths behind](http://www.mathstat.dal.ca/~selinger/random/)
- [info about the probability bias](http://stackoverflow.com/a/14679656/395386)
- [info about glibc's random TYPEs](http://stackoverflow.com/a/25819262/395386)
- glibc's random() TYPE_1 (additive feedback generator) last 4 least significant bits
- higher bits do not influence the future lower bits, so nothing new here
- linear congruential generator
- [info about the maths behind](http://www.pcg-random.org/predictability.html)
## Notes
See [Challenge 23](https://github.com/JesseEmond/matasano-cryptopals/blob/master/src/23.py) of [my solutions](https://github.com/JesseEmond/matasano-cryptopals) to the [Matasano Cryptopals](http://cryptopals.com/) challenges to clone a Mersenne Twister generator from 624 numbers of output.
## Ideas
- predict the output of a LCG that only outputs high bits
- see if glibc's TYPE_1 random can be predicted with 100% accuracy (finding the seed/the states exact values?)