Have you been able to formulate a convex or quasi-convex optimization problem? I don’t see what such a formulation would be.
I will assume this problem is non-convex, unless you show otherwise. Perhaps YALMIP could be used with a local or global optimization solver.