https://github.com/heat1q/libldpc
Fast C++17 Simulation Tool for LDPC Codes with Multithreading Support
https://github.com/heat1q/libldpc
decoding error-correcting-codes forward-error-correction ldpc parallel simulation simulator sum-product-algorithm
Last synced: 5 months ago
JSON representation
Fast C++17 Simulation Tool for LDPC Codes with Multithreading Support
- Host: GitHub
- URL: https://github.com/heat1q/libldpc
- Owner: heat1q
- License: mit
- Created: 2019-05-08T18:36:49.000Z (almost 7 years ago)
- Default Branch: master
- Last Pushed: 2021-08-19T22:15:32.000Z (over 4 years ago)
- Last Synced: 2025-04-03T16:11:15.456Z (10 months ago)
- Topics: decoding, error-correcting-codes, forward-error-correction, ldpc, parallel, simulation, simulator, sum-product-algorithm
- Language: C++
- Homepage:
- Size: 42.1 MB
- Stars: 5
- Watchers: 1
- Forks: 2
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
- License: LICENSE
Awesome Lists containing this project
README
# LDPC Code Simulation
[](https://opensource.org/licenses/MIT)
### Requirements
* CMake Version 3.10
* C++17
### Getting Started
Building the binary:
```
$ cmake .
$ cmake --build . --target TARGET
```
where the target can be the following:
* `--target ldpcsim` produces an executeable containing the simulator, which is used by e.g. the commandline. See **Running the Simulator**.
* `--target ldpc` produces a shared library containing the simulator for external usage. See **Python Wrapper**.
### Running the Simulator
After successful build the simulator can be executed. Note the usage:
```
$ ./ldpcsim --help
Usage: ldpc [options] codefile output-file snr-range
Positional arguments:
codefile LDPC parity-check matrix file containing all non-zero entries.
output-file Results output file.
snr-range {MIN} {MAX} {STEP}
Optional arguments:
-h --help shows help message and exits
-v --version prints version information and exits
-G --gen-matrix Generator matrix file.
-i --num-iterations Number of iterations for decoding. (Default: 50)
-s --seed RNG seed. (Default: 0)
-t --num-threads Number of frames to be decoded in parallel. (Default: 1)
--channel Specifies channel: "AWGN", "BSC", "BEC" (Default: AWGN)
--decoding Specifies decoding algorithm: "BP", "BP_MS" (Default: BP)
--max-frames Limit number of decoded frames.
--frame-error-count Maximum frame errors for given simulation point.
--no-early-term Disable early termination for decoding.
```
### Code File Format
The code file lists the the non-zero entries of the matrix line-by-line,
where the left number refers to the row index and the right to the column index. For example, the matrix
```
0 0 1 0
0 1 0 0
0 0 0 1
1 1 1 0
```
is represented as
```
0 2
1 1
2 3
3 0
3 1
3 2
```
Optionally, a puncturing (or shortening) patter may be specified a the top of the file:
*Puncture 2 bits associated with column index 1 and 3*
```
puncture [2]: 1 3
```
A sample `k=128`, `n=1024` LDPC code with parity-check matrix `h.txt` and generator matrix `g.txt` can be found in `tests/code/`.
### Python Wrapper
The simulator may be used as Python Module in a threaded application.
```
from pyLDPC import ldpc
# initialize with codefile & simfile
code = ldpc.LDPC("doc/exmp_code.txt")
# simulate AWGN with 100 iters and MinSum decoding
code.simulate(snr=[0, 6, 0.2], iterations=100, decoding="BP_MS")
# stop the simulation
code.stop_simulation()
# retrieve the results
print(code.get_results())
```
### Contributing
If you found a bug or have any suggestions or improvements, feel free to start a discussion or submit a PR!