A Novel Server Selection Strategy for Multi-access Edge Computing
Gayan Dilanka, Lakshan Viranga, Rajitha Pamudith, Tharindu D. Gamage, Pasika Sashmal Ranaweera, Indika A. M. Balapuwaduge, Madhusanka Liyanage · 2021
To realize the 5G mobile technology evolution, Multi-access Edge Computing is quite essential. Bringing cloud services towards the edge of the mobile network produces ultra-low latency and higher bandwidths to the end-users in the IoT domain. As a result, IoT devices can achieve the dynamic requirements of the running services. Virtualization enhances edge computing by reducing the number of physical resources required. With this knowledge, automated infrastructure of 5G deployments becomes more cost-effective. Virtualization technologies should be used in the IoT space to fulfill the edge computing paradigms' goal of creating computing infrastructure at the network's edge. With virtualization, 5G server platforms will be fast and affordably accessible for gNodeBs. Virtualized edge computing server platforms are managed and controlled via orchestrators. Even though cloud-based orchestration platforms exist, the optimum MEC server selection is not reached with maximum infrastructure utilization and autonomous infrastructure scaling. This paper proposes a mobile edge MEC server selection strategy, dependent on the MEC server's resources and wireless link parameters. The performance of the proposed strategy was simulated and analyzed in terms of blocking probability and algorithm complexity via simulations.