A minimization of a function with two weighted terms using the semidefinte programming and CVX package

The objective is a weighted sum of two norms. If the objective is almost zero in the solution then either the problem is underdetermined or weirdly scaled, I think.