Mobile edge offloading using markov decision processes

Abstract

Considering where to process data and perform computation is becoming a more difficult problem as Mobile Edge Computing (MEC) and Mobile Cloud Computing (MCC) continue to evolve. In order to balance constraints and objectives regarding items like computation time and energy consumption, computation and data should be automatically shifted between mobile devices, the edge, and the cloud. To address this issue, this study proposes a Markov Decision Process (MDP) based methodology to intelligently make such choices while optimizing multiple objectives. Results demonstrate an 17.47% or greater increase in performance.