Prev: Direct form II transposed filter implementation
Next: Difference between pi/4 QPSK and pi/4 DQPSK
From: sairamesh on 26 May 2010 21:59
Hi,
I am trying to design a receiver structure for system with a
channel code (Convolutional code/block code) followed by a repetition
code at the transmitter side. I understand MAP/ML decoder for the
combined code is optimal in terms reducing the Bit error rate of the
system.
I would like to use two optimal decoders for the each individual
decoders. Do you think just adding up the repeated symbols(received)
is optimal as an independent decoder? Even then combination of each
optimal decoder is optimal/close to optimal for the whole system ?
Any suggestions are welcome.

Thanks for your help.

Best Regards,
-SaiRamesh..


From: Steve Pope on 27 May 2010 00:51
sairamesh(a)gmail.com <nammisairamesh(a)gmail.com> wrote:

> I am trying to design a receiver structure for system with a
>channel code (Convolutional code/block code) followed by a repetition
>code at the transmitter side. I understand MAP/ML decoder for the
>combined code is optimal in terms reducing the Bit error rate of the
>system.
>I would like to use two optimal decoders for the each individual
>decoders. Do you think just adding up the repeated symbols(received)
>is optimal as an independent decoder?

If these signals are all the same nominal magnitude, then yes add
them up. If not, use MRC (Maximum Ratio Combining). If it
is binary signalling, MRC reduces to summing the squares of the
signals given the usual assumptions.

>Even then combination of each
>optimal decoder is optimal/close to optimal for the whole system ?

Not necessarily. You have to be careful how you pass information
from the inner decoder to the outer decoder. And in some
cases, you may have to iteratively decode the two decoders.
It depends on details.

Good luck.

Steve
 | 
Pages: 1
Prev: Direct form II transposed filter implementation
Next: Difference between pi/4 QPSK and pi/4 DQPSK