To　realize　a　seamlessly　connected　living　in　smart　city　is　an　attractive　topic　nowadays.　With　the　help　of　massive　machine　type　communication　(mMTC)　technology　and　drone-based　networking　strategy,　we　can　create　ubiquitous　computing　framework　which　is　both　agile　and　green.　In　this　paper,　we　propose　a　cooperative　data　collecting　mechanism　in　unmanned　aerial　vehicle　(UAV)　covered　internet　of　things　(IoT)　community,　aiming　at　delay-sensitive　context　information　update　and　system　energy　efficiency　(EE)　optimization.　To　better　collect　small　and　fragmentized　IoT　data　scattering　around,　we　first　propose　a　self-organizing　algorithm　to　compress　them　into　a　hotspot　area　and　derive　the　theoretical　collecting　efficiency　(CE)　therein.　Then,　a　system　EE　maximizing　power　control　algorithm　is　proposed,　exploiting　cooperative　communication　theory.　When　dealing　with　this　highly　non-convex　programming　problem,　we　look　carefully　into　its　Hessian　matrix　and　prove　that　under　two　physically　meaningful　conditions,　the　modified　problem　can　be　turned　into　a　strictly　concave　one　and　tractably　solved.　Finally,　a　social　augmented　reality　(AR)　use　case　is　proposed　and　tested　with　our　mechanism.　Simulation　results　validate　our　analysis　and　demonstrate　nearly　20%　improvement　in　EE　performance　comparing　with　benchmark　solutions.　©　2019　IEEE.