The Art of Error Correcting Coding
BCH encoder and decoder written using modern C++17 features, multithreading, Object Oriented Programming and many good practices.
This program simulates a bch encoded transmission through a noisy channel. It divides the given image into message polynomials which are then encodedand the errors are simulated with a given probability. They are afterwards decoded.
After the whole process there is a table with the informations of the process andif you wish you can take a look at three pictures: the original image, image with errorsand the image with errors fixed. Currently there are 4 BCH codes supported.
How to build: mkdir build cd build cmake .. make
Usage:
./bch_simulator [-h] -i image -p err_prob -c code_type [-v]
Options:
-i image Choose one image from images folder, example: images/image2.bmp.
(image has to be in bmp format without compression)
-p err_prob Give probability between 10 and 10000000. A 1 in err_prob error will occur in
the codeword during the simulated transmission through a noisy medium, example: 1000.
-c code_type Choose bch code type:
0 - BCH(63, 51, 5)
1 - BCH(63, 45, 7)
2 - BCH(48, 36, 5)
3 - BCH(48, 30, 7)
Optional arguments:
-h Show this help message.
-v Enable bch_logger encoding and decoding logs which will print out the whole process to the
terminal, is disabled by default. WARNING! This option causes the threads to run sequentially instead
of in parallel which combined with printing operations to console causes a severe performance degradation.
Example:
./bch_simulator -i images/image2.bmp -p 100 -c 3
Example program output:
./bch_simulator -i images/lenna.bmp -p 100 -c 3
number of detected threads: 16
Parsing image file...
Parsing image file done and it took: 0.000 seconds
Primitive polynomial:
p(x) = x^6 + x^4 + x^3 + x + 1
This is a (48,30,7) binary bch code
g(x) is 1001010101000110011
Coding and decoding process done and it took: 0.033 seconds
Converting modified and recovered data from bitsets to bytes...
Converting modified and recovered data from bitsets to bytes done and it took: 0.000 seconds
+-----------------------------------------+-------------------+ | Code used | (48,30,7) | +-----------------------------------------+-------------------+ | Original image used | images/lenna.bmp | +-----------------------------------------+-------------------+ | Given error probability | 0.010000 | +-----------------------------------------+-------------------+ | Real error probability | 0.010000 | +-----------------------------------------+-------------------+ | Number of all data bits | 2001000 | +-----------------------------------------+-------------------+ | Number of all data + redundant bits | 3201600 | +-----------------------------------------+-------------------+ | Number of all introduced errors | 31829 | +-----------------------------------------+-------------------+ | Number of all message polynomials | 66700 | +-----------------------------------------+-------------------+ | Number of successful decodings | 66612 | +-----------------------------------------+-------------------+ | Number of decoding errors | 88 | +-----------------------------------------+-------------------+ | Number of uncaught decoding errors | 18 | +-----------------------------------------+-------------------+ | Number of over t errors in codewords | 83 | +-----------------------------------------+-------------------+ | Final data bits difference | 293 | +-----------------------------------------+-------------------+ | Encoding and decoding time | 0.033 seconds | +-----------------------------------------+-------------------+
Results have been written to BCH_logs.txt file
To view made images on linux use "feh -F -Z --force-aliasing -d images/lenna.bmp images/lenna_with_errors_BCH4830.bmp images/lenna_fixed_BCH4830.bmp"
Inspiration: http://the-art-of-ecc.com