Xavi Artigas - 2004-09-25

Hello,

I am relatively new to it++ and to Error Correcting Coding, so you'll have to excuse me if my questions are out of place.

First off, congratulations. I think this project is great, and has helped me a lot with my PhD thesis.

I am working with the Turbo Codes class, but I need High-Rate constituent encoders. The current implementation of the Recursive Systematic Convolutional Encoder adds as many parity bits as you want to each input data bit, resulting in code rates ranging from 1/2 downwards.
I need higher code rates, that is, 4/5, 8/9, etc.
i.e. I want that for every 4 input bits, only one parity bit is generated (0.2 parity bits per input bit)
To do so, I would need to change the encoding function, but I am not quite sure about how to do that. The ouput of this function won't be a parity matrix as it is now, but a parity vector 4 times smaller than the input vector. Moreover, as of now, I don't know what changes would the decoder need.

Any help will be greatly appreciated.
Thanks a lot, and keep up the good work!

Xavi