Energy Optimal Partial Computation Offloading Framework for Mobile Devices in Multi-access Edge Computing

Abstract

Multi-access Edge Computing (MEC), also known as mobile edge computing, facilitates mobile devices (MDs) to offload their excess computation and data to MEC Server (MECS). The MECSs are located in close proximity to the MDs. This improves the energy efficiency of resource limited MD. At MD, the data offloading results in an additional energy cost of data transactions with the MECS. It is observed that for achieving energy efficient computation offloading, partial computation offloads can be more beneficial compared to the binary offloads, i.e., either full- or no-offload. This paper focuses on the challenging question: how much computation to offload? Further, the data to be offloaded can be compressed before transmission to save the transmission energy at an additional cost of compression-decompression energy. The overall energy consumption of an MD is a combined effect of energy overheads and energy savings. The energy overheads and energy savings depend on interdependent factors, e.g., amount of offloaded computation, MD-MECS distance, channel conditions, application type, and compression efficiency. In this paper, we propose a framework for determining energy optimal computation offloading configuration considering application- and system-specific parameters. Next, we investigate the viability of compression-decompression at the energy constrained MD, while offloading. Simulation results show that using compression saves a significant amount (28%) of energy compared to the offloading without compression. Further, using the energy optimal partial-offloading configuration obtained by the proposed framework saves up to 35% energy vis-a-vis binary data offloading.