Agile Design and Dimensioning of MEC-NFV Infrastructure to Support Heterogeneous Service Chains

Mobile edge computing (MEC) has recently become a key technology that appeared in response to the increasing focus on the emergence of new heterogeneous service chain requests, the resource-constrained mobile devices, and the long delay provided by the conventional Cloud Data Centers. Although many researchers have investigated how the heterogeneous service function chain (SFC) requests can interact with the MEC system, very few have tackled how to deploy a flexible MEC infrastructure at the mobile operator for the expected mobile traffic. This paper addresses the design and dimensioning of an agile MEC infrastructure problem with aid of Network Function Virtualization (NFV) technology. To improve physical resource utilization, we consider virtual instances resources overheads and virtual instances shareability in the deployment process. A mixed-Integer Linear Program (MILP) is used to construct a physical MEC system design by mapping the received SFCs' virtual networks to the dimensioned MEC nodes while minimizing the deployment cost. Numerical results highlight the value gained in reducing the deployment cost while guaranteeing the delay requirements for different SFCs with high acceptance rates.