Multicast routing and virtual network function placement in NFV-SDN networks
Germán García, Diego Pedro Pinto-Roa · 2022
Communication trends demand services that require flexible multicast transmission with requirements for data processing functions at the network nodes. This challenge is achieved with software-defined networking (SDN) and network functions virtualization (NFV) technologies. In the multicast routing problem in NFV-SDN networks, the goal is to compute a multicast tree and the location of virtual network functions (VNFs) at nodes, satisfying the traffic demand and processing functions with the least possible resource. The traffic flow must pass through a VNF node before reaching a destination node. Since the problem has high computational complexity and current proposals consider only one type of virtual function in all demands, the development of scalable solutions for multiple sessions with different virtual functions is necessary. This paper addresses the multicast session routing and VNF placement (MSVNFP) problem in NFV-SDN networks as a joint optimization problem. In this context, we propose an approach based on Genetic Algorithms (GA) called MSVNFP-GA. Given a set of multicast demands, the proposed algorithm calculates for each demand a tree and the location of VNFs seeking to minimize the total cost of links and node activation. Numerical simulations on different network topologies and traffic loads show that MSVNFP-GA is promising compared to the state-of-the-art competitive algorithms.