Hi， Mark, why f1 can be approximated by its lower bound. As far as I know， a minimization optimal problem, we can use upper bound to approximate objective function, not lower bound. Paper title “Secrecy Rate Beamforming for Multi-cell Networks with Information and Energy Harvesting”
This is not a CVX question. I suggest you post at or.stackexhange.com to get help. Then come back to this forum if you have a CVX-specific issue.
ok，thank you very much.