Joint power and coverage control of massive UAVs in post-disaster emergency networks: An aggregative game-theoretic learning approach