You haven’t shown us the solver and CVX output. Did the solver numerically fail, or did it report the problem as infeasible?
Do not use cvx_quiet(true); until you have the program working satisfactorily.
If you have Mosek available, use it as solver (wirh CVX 2.2). Otherwise follow the directions at CVXQUAD: How to use CVXQUAD's Pade Approximant instead of CVX's unreliable Successive Approximation for GP mode, log, exp, entr, rel_entr, kl_div, log_det, det_rootn, exponential cone. CVXQUAD's Quantum (Matrix) Entropy & Matrix Log related functions .
If after doing so the problem is reported as infeasible, follow the advice, except for section 1, at https://yalmip.github.io/debugginginfeasible ,