Comments on: Viterbi decoder

By: Krishna Sankar

Krishna Sankar — Sun, 29 Jan 2012 00:54:29 +0000

@Anuj: Should be reasonably easy to figure out. Check out the size of ipHat_v to figure out why it is smaller than N

By: Anuj Kilje

Anuj Kilje — Sat, 28 Jan 2012 08:17:47 +0000

Try using this and edit in the original file:

% Transmitter
%ip = rand(1,N)>0.5; % generating 0,1 with equal probability

%Generate Random Data[i/p matrix..]
ip=zeros(1,64);
d=rand(1,64);
for i=1:64
if(d(i)>=0.5)
d(i)=+1;
else
d(i)=-1;
end
end
for i=1:64
ip(1,i)=d(i);
end

but the problem in the original code then pops up:

??? Index exceeds matrix dimensions.

Error in ==> Coded_nofinal_2 at 144
nErrViterbi(yy) = size(find([ip- ipHat_v(1:N)]),2)

Im using matlab and not octave.

By: Krishna Sankar

Krishna Sankar — Wed, 04 Jan 2012 00:43:30 +0000

@george: Where did you see the usage of conv2?

By: george

george — Sat, 17 Dec 2011 14:08:08 +0000

Hello,
I run this code in matlab . And it Doesn’t Work the message of matlab is :

??? Error using ==> conv2
First and second arguments must be single or double.

Error in ==> conv at 39
c = conv2(a(:),b(:),shape);

Error in ==> Untitled at 20
cip1 = mod(conv(ip,[1 1 1 ]),2);

PLEASE HELP. What is the problem with code??????

By: Krishna Sankar

Krishna Sankar — Thu, 26 May 2011 00:33:10 +0000

@Rodrigo: Are you sure? I re-ran the code. It’s working.

By: Rodrigo Mendoza

Rodrigo Mendoza — Wed, 25 May 2011 04:00:12 +0000

amm……..
I don`t know why but the code I download from this page doesnt work.
nErrViterbi(yy) = size(find([ip- ipHat_v(1:N)]),2);
the dimensions of ipHat_v are smaller than N
Can you tell me how can I resolve this problem I just cant

By: Krishna Sankar

Krishna Sankar — Fri, 19 Nov 2010 00:57:09 +0000

@guzeltayyar: I guess, one way is to give less weights to softbits affected by impulse noise. This can be done by scaling down those softbits. I have not tried any metric for non-gaussian noise.