Work

Αξιολόγηση μεθόδων που χρησιμοποιούν βαθιά μάθηση για την βελτιστοποίηση του Caching

Year: 2024
Type: dissertation
Abstract: The rapid increase in internet traffic creates challenges in the current network infrastructure.An emerging solution to the traffic congestion problem is edge caching, by storing popular content near the edge of the network it reduces the traffic in the core network.A good caching strategy is crucial for increasing system performance.Deep learning emerges as a suitable solution, with its ability to learn user preferences and predict what content will be frequently requested.This work presents an extensive literature review focused on the exploration of various deep learning architectures for caching optimization in edge networks.Initially, an overview of edge computing and edge caching is provided, highlighting their benefits and various design possibilities, along with their respective impact on caching performance.Then, a collection of machine learning algorithms is presented including supervised, unsupervised, reinforcement and federated learning.Moreover, this paper conducts an analysis of recent research publications, with a focus on factors such as deep learning architectures and the objectives guiding their use.Closing with, exploring implementation challenges and future research paths of utilizing deep learning in edge caching, Institutional Repository
Authors: Μπεκαριάν, Χρήστος Μ., Μπεκαριάν, Χρήστος Μ.
Language: Greek
Open Access status: green
OpenAlex ID: W7016005188
Cites: 83
Cited by:
Related to: 10