Continue to Site

# [moved] Coding in Polar codes

Status
Not open for further replies.

##### Newbie level 3
Hi everybody, my question concerns channel coding using polar codes:

after reading a number of articles and tutorials on this subject a problem persists in my understanding:

We have :
On the one hand: the polarization of the channel which separates the good channels from the bad ones thanks to the Bhattacharyya parameter and, select the indices of the frozen bits...

On the other hand: we have the coding part thanks to the kronecker matrices (factor graph) ... etc.

If U = [u1 ... uN] is the vector at the input of the encoder which contains the information bits and the frozen bits and
X = [x1 ... xN] the vector at the output of the encoder. We know that the xi (i:1:N) are a, using XOR and copies, a combinations of frozen bits and information bits.

My questions are:

1) After the encoder do have we a single channel or multiple channels (N)?
2) If there are N channels, how will the frozen bits go through the wrong channels while they have been mixed , at the encoding part, to the right bits (information bits)?

Hi everybody, my questions concerns channel encoding using polar codes:

We have :
On one hand: the polarization of the channel which creates N channels and separates the good channels from the bad ones thanks to the Bhattacharyya parameter, which leads to the selection of the frozen bits indices and information bits indices.

On an other hand: we have the coding part thanks to the kronecker matrices (factor graph) ... etc.

If U = [u1 ... uN] is the vector at the input of the encoder which contains the information bits and the frozen bits and if
X = [x1 ... xN] is the vector at the output of the encoder. We know that the xi (i:1:N) are , using XOR and copies, .a combinations of frozen bits and information bits.

My questions are:

1) After the encoding stage do we have a single channel or multiple channels (N channels)?
2) If there are N channels, how will the frozen bits go through the wrong channels and the information bits through the good ones knowing that both frozen bits and information bits have been mixed together at the encoding part ( using XOR and copies)?