Energy Efficient Multi-Tenant Resource Slicing in Virtualized Multi-Access Edge Computing

With the help of multi-access edge computing (MEC) system, traditional mobile network operators (MNOs) can provide various services to their mobile users with the minimum delay by installing micro-datacenters at the base stations (BSs)(i.e., at the edge of the radio access network). However, the capital and operational expenditures become significant challenge for the MNOs. Fortunately, multiple MNOs can coexist on the same infrastructure and share the network resources with the help of upcoming technologies such as network virtualization (i.e., network slicing) and software defined networking (SDN). In this work, we introduce a virtualized MEC system in which an infrastructure provider (InP) deploys a BS integrated with a micro-datacenter and owns the wireless network resource i.e., bandwidth. Then, InP creates the virtual network by slicing its network resources including bandwidth and the computation resource of the MEC server, and shares these resource slices to multiple virtual network operators (MVNOs) where MVNOs provide specific services to their mobile users. To solve our proposed problem, we first decompose the original problem into two subproblems. Then, we apply the Karush-Kuhn-Tucker (KKT) conditions to solve each subproblem. Moreover, simulation results prove that our proposed algorithm for joint communication and computation resources sharing outperforms the existing schemes.