Mobile Edge Computing for Task Offloading in Small-Cell Networks via Belief Propagation

A large number of computation-hungry mobile applications have led to an ever-increasing computation demands. Mobile edge computing (MEC) has been considered as an emerging paradigm to alleviate the demand effectively by offloading the computationally intensive tasks from mobile devices (MD) to the adjacent MEC servers. It is expected that the quality of computation experience, e.g., the computing energy and the execution latency, can be greatly improved by the MEC. In this paper, we will investigate the computing task offloading problem via the MEC in the context of small-cell base-station (SBS) networks, where each SBS is equipped with an MEC server. Specifically, we first formulate the optimization problem to minimize the objective, i.e., the weighted sum of energy consumption and execution duration. The parameters to be optimized are the allocations of the MD's tasks to be offloaded to the MEC servers. Then we propose a novel belief propagation (BP) algorithm to optimize the task allocation in a distributed manner. Next, we develop the factor graph according to the network topology and decompose the object function into multiple local utilities to fit the factor graph. Finally, we transform local utilities into the estimations of marginal distributions and propose a distributed BP algorithm to solve the estimations. Simulations demonstrate that our BP algorithm can effectively approach the optimal solutions via exhaustive search.