Algoritmi raslojavanja mreže 5G temeljeni na strojnom učenju

master's thesis

• Title (english): 5G network slicing algorithms based on machine learning
• Author: Hrvoje Grabovac
• Mentor: Željko Ilić (mentor)
• Committee member: Željko Ilić (predsjednik povjerenstva)
• Committee member: Gordan Šišul (član povjerenstva)
• Committee member: Adrian Satja Kurdija (član povjerenstva)
• Granter: University of Zagreb; Faculty of Electrical Engineering and Computing; Zagreb
• Defense date and country: 2022-06-28, Croatia
• Scientific / art field, discipline and subdiscipline: TECHNICAL SCIENCES; Computing

Abstract

Tradicionalne mreže sada dosežu svoje granice. Povećanje propusnosti i broja korisnika ne može se zadovoljiti jednostavnom evolucijom hardvera. Način upravljanja mrežom također treba evoluirati. Uvođenje SDN-a i NFV-a započelo je ovu evoluciju. Ove dvije paradigme omogućuju programibilnu mrežu i apstrakciju resursa, čime se dopušta automatsko upravljanje i puno veća kontrola prometa i resursa. 5G dodatno povećava potrebu za evolucijom i inovacijama u upravljanju mrežom. Uvodi snažna ograničenja za kašnjenje, mobilnost, propusnost, energetsku učinkovitost, sigurnost i prije svega QoS. Infrastruktura mobilnih mreža također se razvija, ali to mora biti povezano s promjenom pristupa upravljanju mrežom. Velika heterogenost slučajeva korištenja i njihovih zahtjeva dovela je do uvođenja mrežnog raslojavanja. Ova nova paradigma uključuje podjelu fizičke mreže na više neovisnih virtualnih mreža, od kojih svaka pruža specifičnu uslugu kako bi zadovoljila preciznim zahtjevima. Naravno, sa novom paradigmom dolaze i novi algoritmi koji zahtijevaju daljnje razvijanje.

Abstract (english)

Traditional networks are now reaching their limits. The increase in throughput and in the number of users cannot be satisfied with a simple hardware evolution. The management of networks also needs to evolve. The introduction of SDN and NFV started this evolution. These two paradigms enable a programmable network and an abstraction of resources, thus allowing an automatic management and a much greater control of traffic and resources. 5G further increases the need for evolution and for innovation in network management. It introduces strong constraints on latency, mobility, throughput, energy efficiency, security, and above all QoS. The infrastructure of mobile networks is also evolving, but this must be coupled with a change in the network management approach. The great heterogeneity of use cases and of their requirements led to the introduction of network slicing. This new paradigm involves the partitioning of the physical network into multiple independent virtual networks, each of which provides a specific service to meet precise requirements. Of course, with the new paradigm come new algorithms that require further development.

• Keywords: mrežno raslojavanje; 5G; strojno učenje; duboko učenje; kvaliteta usluge
• Keywords (english): network slicing; 5G; machine learning; deep learning; Quality of Services
• Language: croatian
• URN:NBN: urn:nbn:hr:168:837832
• Study programme: Title: Computing; Study programme type: university; Study level: graduate; Academic / professional title: magistar/magistra inženjer/inženjerka računarstva (magistar/magistra inženjer/inženjerka računarstva)
• Type of resource: Text
• File origin: Born digital
• Access conditions: Closed access
• Created on: 2023-01-13 14:59:23