Re: Channel Coding
take a (n,k) binary code. Of the 2^n possible combinations, only 2^k are valid codewords. So bigger the n, the further spread out your codewords can be, in the set of all 2^n possibilities. And the more dissimilar your codewords are, the better your error correction performance will be. So increasing n increases the coding gain (of course, code has to be good). Does this explain?
Downside of course is that your rate goes down very badly. You end up sending more redundnat info that real data.
- b