Τίτλος:
A genetic algorithm approach for service function chain placement in 5G and beyond, virtualized edge networks
Γλώσσες Τεκμηρίου:
Αγγλικά
Περίληψη:
Network Function Virtualization (NFV) is already considered as a structural enabler of today's networking technology and particularly the 5th Generation of Broadband and Cellular Networks (5G). NFV provides the means to flexibly and dynamically manage and allocate resources, without being restricted to the hardware limitations of the network/cloud infrastructure. Resource orchestration for specific 5G vertical industries and use case families, such as Industry 4.0 and Industrial Internet of Things (IIoT), often introduce very strict requirements in terms of network performance. In such a dynamic environment, the challenge is to efficiently place directed graphs of Virtual Network Functions (VNFs), named as SFCs (Service Function Chains), to the underlying network topology and to dynamically allocate the required resources. To this end, this work presents a novel framework, which makes use of a delay and location aware Genetic Algorithm (GA)-based approach, in order to perform optimized sequential SFC placement. Evaluation results clearly demonstrate the effectiveness of the proposed framework in terms of producing solutions that approximate well the global optimal, as well as achieving low execution time due to the employed GA-based approach and the incorporation of an early stopping criterion. The performance benefits of the proposed framework are evaluated in the context of an extensive set of simulation-based scenarios, under diverse network configurations and scales. © 2021 Elsevier B.V.
Συγγραφείς:
Magoula, L.
Barmpounakis, S.
Stavrakakis, I.
Alonistioti, N.
Περιοδικό:
Computer Networks
Λέξεις-κλειδιά:
5G mobile communication systems; Directed graphs; Genetic algorithms; Petroleum reservoir evaluation; Transfer functions; Virtual reality, Cellular network; Cloud infrastructures; EDGE Networks; Genetic algorithm approach; Network functions; Networking technology; Performance; Resource orchestration; Service functions; Virtualizations, Network function virtualization
DOI:
10.1016/j.comnet.2021.108157
