[Defence of thesis] 15/07/2024 - Sahand KHODAPARAS TALATAPEH : " Orchestration et optimisation du cache dans les réseaux IoT " (UPR 4128 - LIA - Laboratoire d'Informatique d'Avignon)

Sahand KHODAPARAS TALATAPEH will defend his thesis on 15 July 2024 at Avignon University on the subject of "Orchestration and optimisation of caching in IoT networks".

Date and place

Oral defense scheduled on Monday 15 July 2024 at 15:00
Location: 339 Chem. des Meinajaries, 84000 Avignon
Room: CERI meeting room

Discipline

Computer Science

Laboratory

UPR 4128 - LIA - Avignon Computer Science Laboratory

Management

Mr Abderrahim BENSLIMANE	Avignon University	Thesis supervisor
Mr Saleh YOUSEFI	Urmia university	Thesis co-director

Composition of the jury

Mr Abderrahim BENSLIMANE	Avignon University	Thesis supervisor
Mr Antoine GALLAIS	INSA Hauts-de-France	Rapporteur
Mr Saleh YOUSEFI	Urmia university	Thesis co-director
Mr Jamshid BAGHERZADEH	Urmia university	Examiner
Ms Leila SHARIFI	Urmia university	Examiner
Mr Vahid SOLOUK	Urmia University of Technology	Examiner
Yezekael HAYEL	Avignon University	Examiner
Ms Anna Maria VEGNI	Roma Tre University	Rapporteur

Summary

This thesis explores the improvement of caching mechanisms within Internet of Things (IoT) and Internet of Vehicles (IoV) networks in order to mitigate backhaul network congestion, extend the lifetime of IoT nodes and improve overall network quality of service. It begins by introducing a clustering method within IoT infrastructures, using multi-criteria decision making techniques such as TOPSIS and AHP to optimise cache placement among nodes and leveraging Software Defined Networking (SDN) for efficient routing. The work then looks at IoV networks, differentiating between security and entertainment content caching strategies. Security content caching is localised and dynamic, based on incident severity and data freshness, while entertainment content caching employs federated learning to predict content popularity without compromising user privacy. The final aspect of the research addresses the challenges of network coverage and load distribution through the deployment of Unmanned Aerial Vehicles (UAVs). An optimisation model for UAV placement is developed, taking into account energy constraints, user satisfaction, required data rates and resource efficiency, with the application of reinforcement learning to solve this complex problem. Together, these strategies form a coherent approach to reducing latency and improving Quality of Service in vehicular networks.

Key words : Caching, IoT, IoV, UAV, Federated learning, Blockchain