Please show us your proof that this is a convex optimization problem.

Use `log_sum_exp()`

. Check

Thank you. But do you know how to deal with V_n, Because V_n is also included by log

Define a new variable A, then add a constraint `A==log(V_n)`

, so V_n=exp(log(V_n))=exp(A).

To clarify, `A=log(V_n)`

is a MATLAB assignment, and would not be a CVX equality constraint `A==log(V_n)`

, because `V_n`

is a constant.

V_n is just a constant, have no relationship with n, I wrote has some problems.