A metaheuristic approach for minimizing service creation time in slice-enabled networks

A fundamental technology towards the next generation of mobile networks is the Network Slicing, i.e., the capability to “slice” network resources and functions and to offer isolated end-to-end network services over shared physical infrastructures. From the performance perspective, one of the main challenges in network slicing is the minimisation of the slice set up time; a factor that directly effects a key performance indicator for the 5G networks, namely, the service creation time. Treating a network slice as a set of Virtual Network Functions (VNFs) that are installed in different hosts (central/edge clouds) during the slice creation procedure, we examine a nature-inspired metaheuristic approach for scheduling the VNFs installation process. Simulation results prove that the proposed metaheuristic method can offer high quality solutions for VNF scheduling problems and optimize service creation time for a set of requested network slices, by producing a close-to-optimal scheduling plan, for a large set of VNFs, in reasonable execution time.