Delay-Outage Analysis of OFDMA-Based Task Offloading in Edge Computing Networks.

In this paper, we evaluate the performance of a multiple access edge computing system for task offloading over Rayleigh and Nakagami-m wireless fading channels. In our model, the user device simultaneously offloads a complex task to multiple edge nodes for processing by means of orthogonal frequency-division multiple access (OFDMA) scheme. The proposed scheme is suitable for millimeter Wave (mmWave) bands considered in 5G and beyond mobile networks. To assess the system efficiency, the delay-reliability metric is introduced in the context of the delay-outage probability. We investigate the delay-outage probability expression analytically for the proposed scheme and determine the optimized task size to minimize the offloading delay. Dependency of the system performance on various parameters is also analyzed in this work. Simulation results are provided to numerically evaluate the system reliability. It is found that the selection based scheme in which a single edge node is selected is recommended for OFDMA-based task offloading.