O
ouadid abdelkarim
Hi every body,
I want to have any suggestion from somebody who designed a kind of soft-decision decoding algorithm.
I have a little problem with some notions. I'm trying to design a new codec with soft-decision,i'll use this example:
011 is the most significant 0 in my design, and 010 is less significant and so on till 000 who is the worst 0. Same thing for 1, 111 is the most confident one, and then whe have 110,101,100.
The problem is that i'm using an addition to have a weighed evaluation of ny symbol. I have to add 4 symbols that are over 3 bits so to avoid the overflow my output is over 5 bits. In some case i can have a zero at the end of my adder and i have to feed it back to my input so i have to resize it over 3 bits (kind of saturation will work) but how can i deal with the zero that i don't use in my symbols and that i can't use ( both 100 and 000 are in decimal -1 and +1 so ...
_ )
Any suggestion or question?
thanks for any help
I want to have any suggestion from somebody who designed a kind of soft-decision decoding algorithm.
I have a little problem with some notions. I'm trying to design a new codec with soft-decision,i'll use this example:
011 is the most significant 0 in my design, and 010 is less significant and so on till 000 who is the worst 0. Same thing for 1, 111 is the most confident one, and then whe have 110,101,100.
The problem is that i'm using an addition to have a weighed evaluation of ny symbol. I have to add 4 symbols that are over 3 bits so to avoid the overflow my output is over 5 bits. In some case i can have a zero at the end of my adder and i have to feed it back to my input so i have to resize it over 3 bits (kind of saturation will work) but how can i deal with the zero that i don't use in my symbols and that i can't use ( both 100 and 000 are in decimal -1 and +1 so ...
_ )Any suggestion or question?
thanks for any help