This is a nonconvex MINLP problem. If I implement it in GAMS using BARON, the optimal solution I got is a global solution? As BARON is a global solver.
In the general case, indepently of the software used, you must to organize an special search process for falling in a solution "close to the global optimal one".
As posed, the answer is: I do not know. You really offer almost no detail about your problem. But in general the answer is: there is no guarantee, that an algorithm finds the global optimum, when the constraint set, apart from integrality constraints, is non convex . However, there is a bunch of literature, about objective functions, that can be written as a linear function divided by a linear function.
Just consult an expert in nonlinear programming, which I am not any more.
By the way, your problem is two-dimensional, so you could try to make a picture of the constraint set. That also gives you a better idea about the nature of your problem.