Solving an MILP (MIP) problem within MIP problem

Dear all ,

I have following doubt:

My problem is roughly of following nature:

sets 
	it "iter" /it1*it10/
;
variables x(it) ,y(it)
loop
(it,
MIP_problem(it)
)

For each iteration above, I solve MIP problem to get x(it) and y(it)

Now, I have to solve another MIP problem, which should be as follows:

variables xsum , ysum

equations
y_eqn;
y_eqn..
ysum = sum(it,y(it))
;

model eqn/y_eqn/

solve eqn minimizing ysum using MIP

Now since ysum depends on y(it), it should also start solving the loop again such that the ysum will be minimum. Is it even doable? Can I invoke an MILP problem from another ?

Thanks

Are you referring to bilevel optimisation?