The key is to know that if and are i.i.d. random variables with density function f(y) then the joint density function of and is

for

and zero elsewhere. I can prove this result if you want, but you should have a copy of the proof somewhere.

For your problem for and .

Then .