Can anyone help me to solve this problem and get the closed-form expression?
That is a non-convex problem. and therefore out of scope of this forum. I have no reason to believe it has a closed-form solution. Unless you know that it does, assuming it does is not a good assumption.
If you want to solve it numerically, to global optimality, you could try the BMIBNB or BARON global optimization solvers, under YALMIP. Whether the problem is actually solvable, to within a reasonable optimality gap, in an acceptable amount of time, depends on, among other things, the problem dimension.
1 Like
Thank you so much for your response.