math - What is the rationale for the encoding procedure described in the DVB-S2 standard for LDPC codes?

Question

I have not found an explanation on why accumulating the given parity check addresses and then xoring the accumulated values gives the codeword of the LDPC code. The close it gets the Wikipedia explanation of "Example Encoder" but it happens to have no references. Does it has to do with its Tanner graph representation? Maybe related to Richardson and Urbanke? On every book I've visited the procedure is the usual matrix multiplication and the 2 step Richardson-Urbanke procedure.

It the standard, to be clear, an array of addresses is given with locations of where to accumulate the incoming bit in the parity check bits (Similar to wikipedia's repeat-accumulate). After the last bit is exhausted, the whole parity bit vector is xored with itself (again, see wikipedia's picture).

My question is why? Where does it come from?

question from:https://stackoverflow.com/questions/65602947/what-is-the-rationale-for-the-encoding-procedure-described-in-the-dvb-s2-standar

Answer 1

I think i've found an answer, irregular repeat accumulate codes link1 or link2 .