An adaptive GA-based slice provisioning method for vertical industries in 5G and beyond networks

The fifth-generation of mobile communication networks are expected to support a large number of vertical industries requiring services with diverging requirements. In this vein, mobile networks have been witnessing a radical transformation to satisfy three key services types, ultra-reliable low-latency communication, enhanced mobile broadband, and massive machine-type communication, that will coexist on the same network infrastructure thanks to the network slicing paradigm. Moreover, a distributed user-plane deployment is introduced, allowing edge deployment of delay-sensitive applications thanks to the multi-access edge computing technology. However, designing resilient, elastic, end-to-end slice provisioning mechanisms is still an open issue. In this paper, we study the problem of end-to-end network slice provisioning for different vertical industries whose requested services are modeled as service function chains spanning from the radio access to the application and including user plane and control plane functionalities. Two slice provisioning strategies are considered by formulating a mixed integer linear programming problem and considering different objective functions allowing network operators to deploy the services while pursuing specific business logic. Further, an adaptive genetic algorithm based approach is proposed to solve the same problem in a short time scale by improving the search-space exploration while the population evolves. The simulation results demonstrate the effectiveness of the approach providing near-optimal solutions while drastically reducing the computational complexity.