Habilitation À Diriger Des Recherches Year : 2022

Towards Data Driven Intelligent Networks

Vers les réseaux intelligents guidés par les données

Abstract

Data driven techniques are transforming the world of Information and Communication Technology. Data driven techniques are starting to enhance the computer networks by providing solutions for the problems of detection as well as for the problems of optimisation and decision making. The exploitation of such techniques for enhancing networks is becoming more and more possible due to the increased flexibility offered by the arrival of virtualisation, programmable network devices, software defined networking. They are the enablers which bring agility to the networks, allowing techniques such as machine learning to solve networking problems.

Such transformations in networks are supported by the long-term vision of self-driving networks where automation and intelligence are two essential ingredients. At the same time, users are always asking for better Quality of Experience (QoE). New services and paradigms such as the Internet of things (IoT) are growing. Thus, there is a need for agile, flexible and fully autonomous networks to accommodate a plethora of new smart services. Costs (OPEX/CAPEX) also need to be minimised to stay competitive. Networks should largely self-manage themselves and automatically deal with issues such as QoE, network optimisation, management of heterogeneous services including IoT. They should be able to automatically detect important context and quality information and should be able to automatically optimise the network and services for the users.

This work focuses on AI and optimization approaches for improving network performance and advancing towards intelligent networks. Different contributions are grouped together and organized in 3 parts. The direction of the research is oriented towards the design of intelligent networks. The first part uses machine learning based approaches to automatically detect and estimate QoE. The second part uses the estimation and detection results and combines them with optimisation techniques. This is done to design QoS and QoE aware multicast, routing as well as virtual network function (VNF) placement schemes. The last part of the work focuses on AI techniques as well as integrating them more and more with networking to go towards automation as well as autonomous networks.

Fichier principal
Vignette du fichier
hdr.pdf (3.81 Mo) Télécharger le fichier
Origin Files produced by the author(s)
Licence

Dates and versions

tel-05024042 , version 1 (07-04-2025)

Licence

Identifiers

  • HAL Id : tel-05024042 , version 1

Cite

Kamal Singh. Towards Data Driven Intelligent Networks. Computer Science [cs]. Université jean Monnet - Saint-Etienne, 2022. ⟨tel-05024042⟩
292 View
299 Download

Share

  • More