definitely. that's the only reason I know about them. usually you have to do a little more, because the errors are very bursty, so you have to use very wide windows or interleave in order to spread the errors out so that the redundancy can cover them.

if you're saying 'thats a link layer problem' then I agree, but it would be better to change your link level encoding strategy than to just start sending multiple copies at the transport layer.