GoPro: a Low Complexity Task Allocation Algorithm for a Mobile Edge Computing System

In an Internet of Things (IoT) based network, tasks arriving at individual nodes can be processed in-device or at a local Mobile Edge Computing (MEC) server. In this paper, we focus on the optimal resource allocation problem for tasks arriving in an MEC based IoT network. To address the inherent trade-off between the computation time and the power consumption, we aim to minimize the average power consumption subject to a constraint on the deadline violation probability. The problem is formulated as a Constrained Markov Decision Process (CMDP) problem. To address the high complexities of achieving optimality, we propose a low-complexity heuristic task scheduling scheme. Efficacy of our approach is demonstrated using simulations.