Guler, EvrimKarakus, MuratAyaz, Furkan2024-10-042024-10-0420231084-80451095-8592https://doi.org/10.1016/j.jnca.2022.103538http://hdl.handle.net/20.500.12403/3349The recent network virtualization technology enables the multi-tenancy, where various virtual network requests can share the same physical network by decoupling network services from the underlying hardware architecture. The process of virtual node and link mapping onto a shared Substrate Network (SN) by satisfying the requested network resources (i.e., bandwidth, computing capacity, etc.) is referred to as Virtual Network Embedding (VNE), which is known as an NP-Hard problem. The problem of VNE aims to exhibit one-to-one (unicast) communication. However, the motivation of this research is to explore how to efficiently map virtual networks with one-to-many (multicast) communications, which are in the form of Virtual Multicast Trees (VMTs), onto an SN. This problem differs from the traditional VNE problem and has not been well-studied by the research community. To this end, we propose a novel algorithm, Modified Genetic Algorithm for Virtual Multicast Tree Embedding (MGA-VMTE), to embed VMTs onto a shared SN in this research. The proposed MGA-VMTE algorithm focuses on minimizing the network resource consumption (i.e., bandwidth) under end-to-end delay constraint in the SN while satisfying the computing request of virtual nodes. Our extensive simulations demonstrate that the MGA-VMTE algorithm outperforms the dynamic impact factor and traditional greedy-based virtual multicast tree embedding approaches regarding bandwidth consumption, acceptance ratio, and resource depletion ratio metrics on NSFNET, USNET, Random-60 node, and Random-120 node network topologies.eninfo:eu-repo/semantics/closedAccessEmbeddingVirtualizationMulticastGeneticSDNArticle20910.1016/j.jnca.2022.1035382-s2.0-85141892595Q1WOS:000912938400001Q1