A Transport Theoretic Approach for Computational Task Migration in Multi-Access Edge Computing

Sarder Fakhrul Abedin, Md. Shirajum Munir, SeokWon Kang, Choong Seon Hong

http://doi.org/10.5626/JOK.2019.46.10.1061

In the present work, the problem of computational task migration in the Multi-Access Edge Computing (MEC) Network has been addressed and the goal is to minimize the computational cost including the task migration cost of the MEC network. Apparently, at first, we have formulated a Hitchcock-Koopmans transportation problem, which corresponds to the task migration from the over-utilized MEC servers to the under-utilized MEC server. Second, we have solved the transportation problem using the Vogel’s Approximation Algorithm (VAM), where the optimal task migration was achieved. Finally, in the simulation, we have demonstrated that the proposed approach significantly outperforms the baseline approach in terms of the task migration cost, average response time, and average queuing delay in the MEC network.