Espace membre

Cet espace est dédié aux étudiants, aux enseignants et au personnel administratif de l'école

Valider

Mot de passe oublié?

Actualités de l'école

12/12/2017

Doctorate thesis defense of Emna FAKHFAKH BEN AMAR




Doctorate thesis defense on December 12th 2017 at 109H00 ,in Amphi Ibn Khaldoun, Sup’Com .


Entitled :IMPROVING TRAFFIC OFFLOADING TECHNIQUES IN 4G NETWORKS AND BEYOND

Presented by :Emna FAKHFAKH BEN AMAR





Committee


Chair :

Ridha BOUALLEGUE

Professor at Sup’Com, Tunisia

 

 

 

Examiner :

Fatma ABDELKEFI

Associate professor at Sup’Com, Tunisia

 

Reviewers :

Tijani CHAHED

Professor at Telecom-SudParis, France

   

 

Lamia CHAARI

Associate professor at ISIMS, Tunisia

   

Thesis Director :

Fatma Abdelkefi

Associate Professor, Sup’Com Tunisia

 

Co-supervisor :

Soumaya HAMOUDA

Associate professor at FSB, Tunisia

 

Abstract


Telecom operators have adopted the Traffic Offloading technique for 4G generations and beyond to cope with the growth of mobile data traffic. Heterogeneous networks (HetNet), Device to Device (D2D) communications as well as Wi-Fi technology are among the most attracting technologies to improve radio capability, minimize the outage probability and improve energy efficiency.


In HetNets, new Traffic Offloading strategies are introduced using the eICIC techniques and the CRE technique. We proposed FT-CRE scheme which equally shares the radio resources between the femtocells and the macrocell. The results showed that when FT-CRE scheme is applied, the achieved throughput of the victim macro-users is increased by 6 times as compared to a common application of CRE technique. Then, we proposed an enhanced FT-CRE (eFT-CRE) scheme to utilize the time and frequency resources more efficiently. We have shown that, with our eFT-CRE scheme, the total achieved throughput is improved by 3 times, the outage probability is reduced by 60% and the energy efficiency at the macrocell is more increased.


Using D2D communications in the context of Traffic Offloading, we addressed the problem of communication mode selection. We proposed a D2D mode selection scheme based on the Noise Rise parameter to better control the interference caused by the D2D transmitters at the base station. With this scheme, the system radio capacity is increased by almost twice as compared to the Common Selection Scheme. Then, we proposed a new power control algorithm which further enhances the achieved throughput of D2D users. The results showed that our approach leads to a compromise between the system radio capacity and the Traffic Offloading from the base station.


The last part of the thesis investigates Traffic Offloading through Wi-Fi technology. We have to propose a new Wi-Fi Offloading decision as well as the best Wi-Fi-Access Point selection, using a distributed Q-learning algorithm. The idea is to encourage the mobile to select the Wi-Fi network whenever the latter has sufficient radio resources. Simulation results showed the effectiveness of our proposed Q-learning based scheme as compared to common Wi-Fi Offloading scheme in terms of cellular residence time.


Keywords


LTE-A, 5G, HetNets, eICIC, CRE, Device-to-Device, Noise Rise, Wi-Fi, Q-learning