One of the trickiest challenges introduced by cellular communications networks is mobility prediction for Location Based-Services (LBSs). Hence, an accurate and efficient mobility prediction technique is particularly needed for these networks. The mobility prediction technique incurs overheads on the transmission process. These overheads affect properties of the cellular communications network such as delay, denial of services, manual filtering and bandwidth. The main goal of this research is to enhance a mobility prediction scheme in cellular communications networks through three phases. Firstly, current mobility prediction techniques will be investigated. Secondly, innovation and examination of new mobility prediction techniques will be based on three hypothesises that are suitable for cellular communications network and mobile user (MU) resources with low computation cost and high prediction success rate without using MU resources in the prediction process. Thirdly, a new mobility prediction scheme will be generated that is based on different levels of mobility prediction. In this thesis, a new mobility prediction scheme for LBSs is proposed. It could be considered as a combination of the cell and routing area (RA) prediction levels. For cell level prediction, most of the current location prediction research is focused on generalized location models, where the geographic extent is divided into regular-shape cells. These models are not suitable for certain LBSs where the objectives are to compute and present on-road services. Such techniques are the New Markov-Based Mobility Prediction (NMMP) and Prediction Location Model (PLM) that deal with inner cell structure and different levels of prediction, respectively. The NMMP and PLM techniques suffer from complex computation, accuracy rate regression and insufficient accuracy. In this thesis, Location Prediction based on a Sector Snapshot (LPSS) is introduced, which is based on a Novel Cell Splitting Algorithm (NCPA). This algorithm is implemented in a micro cell in parallel with the new prediction technique. The LPSS technique, compared with two classic prediction techniques and the experimental results, shows the effectiveness and robustness of the new splitting algorithm and prediction technique. In the cell side, the proposed approach reduces the complexity cost and prevents the cell level prediction technique from performing in time slots that are too close. For these reasons, the RA avoids cell-side problems. This research discusses a New Routing Area Displacement Prediction for Location-Based Services (NRADP) which is based on developed Ant Colony Optimization (ACO). The NRADP, compared with Mobility Prediction based on an Ant System (MPAS) and the experimental results, shows the effectiveness, higher prediction rate, reduced search stagnation ratio, and reduced computation cost of the new prediction technique.

Mobility prediction is defined as guessing the next access point(s) a mobile terminal will join so as to connect to a (wired or wireless) network. Knowing in advance where a terminal is heading for allows taking proactive measures so as to mitigate the impact of handovers and, hence, improve the network QoS. This thesis analyzes this topic from different points of view. It is divided into three parts. The first part evaluates the feasibility of mobility prediction in a real environment. It thus analyzes a mobility trace captured from a real network to measure the intrinsic entropy of the nodes motion and to measure the effectiveness of a simple prediction method. The second part investigates how to perform mobility prediction. Firstly, it examines a generic prediction scheme based on a simple machine learning method; this scheme is evaluated under various conditions. Secondly, it shows how the pieces of information that are most useful for the prediction algorithm can be obtained. The third part studies how knowing the probable next access point of a mobile terminal allows one to improve the QoS of the network considered. We deal with two situations. We first show how the handover blocking rate of a cellular network can be decreased thanks to resource reservation. We then propose a new routing protocol for delay tolerant networks (i.e. an ad hoc network where packets must be delayed in the absence of an end-to-end path) that assumes that the contacts between the nodes can be (imperfectly) predicted.

Behavioral patterns prediction in the context of Vehicular Ad hoc Networks (VANETs)has been receiving increasing attention due to enabling on-demand, intelligent traffic analysis and response to real-time traffic issues. One of these patterns, sequential patterns, are a type of behavioral patterns that describe the occurence of events in a timely-ordered fashion. In the context of VANETs, these events are defined as an ordered list of road segments traversed by vehicles during their trips from a starting point to their final intended destination, forming a vehicular path. Due to their predictable nature, undertaken vehicular paths can be exploited to extract the paths that are considered frequent. From the extracted frequent paths through data mining, the probability that a vehicular path will take a certain direction is obtained. However, in order to achieve this, samples of vehicular paths need to be initially collected over periods of time in order to be data-mined accordingly. In this thesis, a new set of formal definitions depicting vehicular paths as sequential patterns is described. Also, five novel communication schemes have been designed and implemented under a simulated environment to collect vehicular paths; such schemes are classified under two categories: Road Side Unit-Triggered (RSU-Triggered) and Vehicle-Triggered. After collection, extracted frequent paths are obtained through data mining, and the probability of these frequent paths is measured. In order to evaluate the e ciency and e ectiveness of the proposed schemes, extensive experimental analysis has been realized. From the results, two of the Vehicle-Triggered schemes, VTB-FP and VTRD-FP, have improved the vehicular path collection operation in terms of communication cost and latency over others. In terms of reliability, the Vehicle-Triggered schemes achieved a higher success rate than the RSU-Triggered scheme. Finally, frequent vehicular movement patterns have been effectively extracted from the collected vehicular paths according to a user-de ned threshold and the confidence of generated movement rules have been measured. From the analysis, it was clear that the user-de ned threshold needs to be set accordingly in order to not discard important vehicular movement patterns.

Behavioral patterns prediction in the context of Vehicular Ad hoc Networks (VANETs)has been receiving increasing attention due to enabling on-demand, intelligent traffic analysis and response to real-time traffic issues. One of these patterns, sequential patterns, are a type of behavioral patterns that describe the occurence of events in a timely-ordered fashion. In the context of VANETs, these events are defined as an ordered list of road segments traversed by vehicles during their trips from a starting point to their final intended destination, forming a vehicular path. Due to their predictable nature, undertaken vehicular paths can be exploited to extract the paths that are considered frequent. From the extracted frequent paths through data mining, the probability that a vehicular path will take a certain direction is obtained. However, in order to achieve this, samples of vehicular paths need to be initially collected over periods of time in order to be data-mined accordingly. In this thesis, a new set of formal definitions depicting vehicular paths as sequential patterns is described. Also, five novel communication schemes have been designed and implemented under a simulated environment to collect vehicular paths; such schemes are classified under two categories: Road Side Unit-Triggered (RSU-Triggered) and Vehicle-Triggered. After collection, extracted frequent paths are obtained through data mining, and the probability of these frequent paths is measured. In order to evaluate the e ciency and e ectiveness of the proposed schemes, extensive experimental analysis has been realized. From the results, two of the Vehicle-Triggered schemes, VTB-FP and VTRD-FP, have improved the vehicular path collection operation in terms of communication cost and latency over others. In terms of reliability, the Vehicle-Triggered schemes achieved a higher success rate than the RSU-Triggered scheme. Finally, frequent vehicular movement patterns have been effectively extracted from the collected vehicular paths according to a user-de ned threshold and the confidence of generated movement rules have been measured. From the analysis, it was clear that the user-de ned threshold needs to be set accordingly in order to not discard important vehicular movement patterns.

Behavioral patterns prediction in the context of Vehicular Ad hoc Networks (VANETs)has been receiving increasing attention due to enabling on-demand, intelligent traffic analysis and response to real-time traffic issues. One of these patterns, sequential patterns, are a type of behavioral patterns that describe the occurence of events in a timely-ordered fashion. In the context of VANETs, these events are defined as an ordered list of road segments traversed by vehicles during their trips from a starting point to their final intended destination, forming a vehicular path. Due to their predictable nature, undertaken vehicular paths can be exploited to extract the paths that are considered frequent. From the extracted frequent paths through data mining, the probability that a vehicular path will take a certain direction is obtained. However, in order to achieve this, samples of vehicular paths need to be initially collected over periods of time in order to be data-mined accordingly. In this thesis, a new set of formal definitions depicting vehicular paths as sequential patterns is described. Also, five novel communication schemes have been designed and implemented under a simulated environment to collect vehicular paths; such schemes are classified under two categories: Road Side Unit-Triggered (RSU-Triggered) and Vehicle-Triggered. After collection, extracted frequent paths are obtained through data mining, and the probability of these frequent paths is measured. In order to evaluate the e ciency and e ectiveness of the proposed schemes, extensive experimental analysis has been realized. From the results, two of the Vehicle-Triggered schemes, VTB-FP and VTRD-FP, have improved the vehicular path collection operation in terms of communication cost and latency over others. In terms of reliability, the Vehicle-Triggered schemes achieved a higher success rate than the RSU-Triggered scheme. Finally, frequent vehicular movement patterns have been effectively extracted from the collected vehicular paths according to a user-de ned threshold and the confidence of generated movement rules have been measured. From the analysis, it was clear that the user-de ned threshold needs to be set accordingly in order to not discard important vehicular movement patterns.

Context: The context of this thesis to predict user mobility using deep learning algorithms which can increase the quality of service for the users and reduce the cost of paging for telecom carriers. Objectives: This study first investigates to find the suitable deep learning algorithms that can be used to predict user mobility and then an experiment is performed with the chosen algorithms as a global model and individual model then evaluate the performance of algorithms. Methods: Firstly, a Literature review is used to find suitable deep learning algorithms and then based on finding an experiment is performed to evaluate the chosen deep learning algorithms. Results: Results from the literature review show that the RNN, LSTM, and variants of the LSTM are the suitable deep learning algorithms. The models are evaluated with metrics accuracy. The results from the experiment showed that the individual model gives better performance in predicting user mobility when compared to the global model. Conclusions: From the results obtained from the experiment, it can be concluded that the individual model is the technique of choice in predicting user mobility.

With the recent growth and wide availability of heterogeneous wireless access technologies, inter-vehicle communication systems are expected to culminate in integrating various wireless standards for the next generation of connected and autonomous vehicles. The role of 5G-enabled vehicular networks has become increasingly important, as current Internet clients and providers have urged robustness and effectiveness in digital services over wireless networks to cope with the latest advances in wireless mobile communication. However, to enable 5G wireless technologies' dense diversity, seamless and reliable wireless communication protocols need to be thoroughly investigated in vehicular networks. 5G-enabled vehicular networks applications and services such as routing, mobility management, and service discovery protocols can integrate mobility-based prediction techniques to elevate those applications' performance with various vehicles, applications, and network measurements. In this thesis, we propose a novel suite of 5G-enabled smart mobility prediction and management schemes and design a roadmap guide to mobility-based predictions for intelligent vehicular network applications and protocols. We present a thorough review and classification of vehicular network architectures and components, in addition to mobility management schemes, benchmarks advantages, and drawbacks. Moreover, multiple mobility-based schemes are proposed, in which vehicles' mobility is managed through the utilization of machine learning prediction and probability analysis techniques. We propose a novel predictive mobility management protocol that incorporates a new networks' infrastructure discovery and selection scheme. Next, we design an efficient handover trigger scheme based on time-series prediction and a novel online neural network-based next roadside unit prediction protocol for smart vehicular networks. Then, we propose an original adaptive predictive location management technique that utilizes vehicle movement projections to estimate the link lifetime between vehicles and infrastructure units, followed by an efficient movement-based collision detection scheme and infrastructure units localization strategy. Last but not least, the proposed techniques have been extensively evaluated and compared to several benchmark schemes with various networks' parameters and environments. Results showed the high potentials of empowering vehicular networks' mobility-based protocols with the vehicles' future projections and the prediction of the network's status.

Les femtocellules sont déployées par des FAPs dans la couverture des macrocellules afin d'offrir aux utilisateurs un service continu aussi bien à l'intérieur qu'à l'extérieur.Elles sont caractérisées par une courte portée,faible puissance et ne peuvent couvrir qu'un nombre limité des utilisateurs.Ces caractéristiques rendent la gestion de la mobilité l'un des plus importants défis à résoudre.Dans cette thèse,nous proposons des nouveaux algorithmes de handover.En premier lieu,nous considérons la direction du mobile comme un paramètre clé pour la prise de décision de Handover.Nous proposons un algorithme de handover nommé OHMF basé sur l'optimisation de la liste de FAPs candidats tout en considérant la qualité de signal ainsi que la direction de mouvement de mobile.Ensuite,nous proposons un processus de prédiction de direction basé sur la régression linéaire.L'idée est de prédire la position future du mobile tout en tenant compte des positions actuelle et précédente.Cet algorithme est intitulé OHDP. En deuxième lieu,nous nous intéressons au problème de prédiction de mobilité pour être plus rigoureux lors de prise de décision de handover.Pour cela,nous utilisons les chaînes de markov cachées comme prédicteur du prochain FAP et nous proposons un algorithme de handover nommé OHMP. Afin d'adapter notre solution à toutes les contraintes du réseau femtocellules,nous proposons un algorithme de handover intitulé OHMP-CAC qui intègre un CAC approprié au réseau étudié et une différenciation de service avec et sans contraintes de QoS.Des études de performances basées sur des simulations et des traces de mobilité réelles ont été réalisées pour évaluer l'efficacité de nos propositions. / Femtocell network are deployed in the macrocell’s coverage to provide extended services with better performances. Femtocells have a short-range and low power transmissions.Each FAP supports a few number of connected users.Owing to these inherent features, one of the most challenging issues for the femtocellular network deployment remains the mobility management.In this thesis, we propose new handovers algorithms adapted to the characteristics of femtocells network.As a first part,we consider the direction of mobile user as a key parameter for the handover decision.To do so,we propose a new handover algorithm called OHMF. Its main purpose is the optimization of the list of FAPs candidates based on signal quality as well as the mobile direction to better choose the FAP target.After that, we propose an algorithm called OHDP based on the direction prediction using the linear regression.The idea behind this is to predict the future position of mobile based on its current and previous position. As a second part, we focus on mobility prediction problem to make an efficient handover decision.We propose a novel handoff decision algorithm called OHMP that uses HMM as a predictor to accurately estimate the next FAP that a mobile UE would visit,given its current and historical movement information.In order to adapt our solution to the characteristics of femtocells network,we propose a handover algorithm called OHMP-CAC based on HMM tool as a predictor, a proposed CAC and the availability of resources of the predicted FAP,SINR and the traffic type.In order to assess the efficiency of our proposals,all underlying algorithms are evaluated through simulations and real mobility traces.

L'objectif de cette thèse est de développer les méthodes nécessaires à une gestion ciblée et efficace de la mobilité des utilisateurs dans un réseau mobile hétérogène. Ces réseaux sont caractérisés par le déploiement de différents types de cellules (macro, micro, pico et/ou femto). Le déploiement massif des petites cellules (pico et femto) a permis d'offrir une capacité et une qualité de couverture accrue au réseau, notamment dans les zones à forte densité. Cependant, les contraintes temps réel engendrées limitent la QoS offerte aux utilisateurs. De plus, pour des raisons commerciales et/ou environnementales, la nécessité de réduire la consommation énergétique des réseaux mobiles est devenue une réalité. Ainsi, les opérateurs mobiles doivent trouver le bon compromis entre d'une part, la garantie de la QoS offerte aux utilisateurs et la vitesse de mobilité de ces derniers, et d'une autre part, le coût énergétique engendré pour le déploiement du réseau. Pour cela, dans le cadre de la gestion de la mobilité des utilisateurs, nous proposons des modèles pour la gestion des ressources des stations de base ainsi que pour la gestion de leur consommation énergétique. Le premier modèle proposé vise à gérer le partage des ressources entre les clients de l'opérateur mobile. Basé sur la prédiction de la mobilité des utilisateurs, ce modèle permet d'anticiper la gestion des ressources d'une station de base. Le deuxième modèle gère la consommation énergétique du réseau en se basant sur un contrôle d'affectation des utilisateurs mobiles. Cela permet de contrôler en continu la consommation énergétique des stations de base et la QoS qu'elles offrent aux utilisateurs mobiles. Par simulation, en utilisant une topologie réelle d'un réseau mobile, les performances des méthodes proposées sont évaluées en considérant différents scénarios possibles. Leurs performances sont comparées à celles de l'approche adoptée par des opérateurs mobiles actuels, ainsi qu'à celles de certaines approches proposées dans la littérature. / The objective of this thesis is to develop methods for a targeted and efficient management of users mobility in heterogeneous mobile networks. This network is characterized by the deployment of different types of cells (macro, micro, pico and/or femto). The massive deployment of small cells (pico and femto) provides a supplementary coverage and capacity to mobile networks, specially in dense areas. However, the resulting real-time constraints limit the offered QoS. Furthermore, for commercial and/or environmental reasons, the needs to reduce the energy consumed by mobile networks became reality. Thus, mobile operators have to find a good compromise between, on the one hand, the users velocity and the guaranteed QoS, and on the other hand, the cost of deployment of such networks. For that, in the context of users mobility management, we propose models for resource and energy consumption management of base stations. The first model aims at controlling resource sharing between clients of the mobile operators. Based on a mobility prediction of users, this model anticipates the resource management of a base station. The second model aims at reducing energy consumption of the network by managing mobile users assignment to detected cells. This allows a continuous control of consumed energy of base stations while offered QoS is guaranteed. Based on simulation of a real mobile network topology, the performances of proposed models are evaluated while considering different possible scenarios. They are compared to the performances of different strategies as the ones proposed in literature or adopted by current mobile operators.

Récemment, l'utilisation des réseaux cellulaires a radicalement changé avec l’émergence de la quatrième génération (4G) de systèmes de télécommunications mobiles LTE/LTE-A (Long Term Evolution-Advanced). Les réseaux de générations précédentes (3G), initialement conçus pour le transport de la voix et les données à faible et moyen débits, ont du mal à faire face à l’augmentation accrue du trafic de données multimédia tout en répondant à leurs fortes exigences et contraintes en termes de qualité de service (QdS). Pour mieux répondre à ces besoins, les réseaux 4G ont introduit le paradigme des Réseaux Hétérogènes (HetNet).Les réseaux HetNet introduisent une nouvelle notion d’hétérogénéité pour les réseaux cellulaires en introduisant le concept des smalls cells (petites cellules) qui met en place des antennes à faible puissance d’émission. Ainsi, le réseau est composé de plusieurs couches (tiers) qui se chevauchent incluant la couverture traditionnelle macro-cellulaire, les pico-cellules, les femto-cellules, et les relais. Outre les améliorations des couvertures radio en environnements intérieurs, les smalls cells permettent d’augmenter la capacité du système par une meilleure utilisation du spectre et en rapprochant l’utilisateur de son point d’accès au réseau. Une des conséquences directes de cette densification cellulaire est l’interférence générée entre les différentes cellules des diverses couches quand ces dernières réutilisent les mêmes fréquences. Aussi, la définition de solutions efficaces de gestion des interférences dans ce type de systèmes constitue un de leurs défis majeurs. Cette thèse s’intéresse au problème de gestion des interférences dans les systèmes hétérogènes LTE-A. Notre objectif est d’apporter des solutions efficaces et originales au problème d’interférence dans ce contexte via des mécanismes d’ajustement de puissance des petites cellules. Nous avons pour cela distingués deux cas d’étude à savoir un déploiement à deux couches macro-femtocellules et macro-picocellules. Dans la première partie dédiée à un déploiement femtocellule et macrocellule, nous concevons une stratégie d'ajustement de puissance des femtocellules assisté par la macrocellule et qui prend en compte les performances des utilisateurs des femtocells tout en atténuant l'interférence causée aux utilisateurs des macrocellules sur leurs liens montants. Cette solution offre l’avantage de la prise en compte de paramètres contextuels locaux aux femtocellules (tels que le nombre d’utilisateurs en situation de outage) tout en considérant des scénarios de mobilité réalistes. Nous avons montré par simulation que les interférences sur les utilisateurs des macrocellules sont sensiblement réduites et que les femtocellules sont en mesure de dynamiquement ajuster leur puissance d'émission pour atteindre les objectifs fixés en termes d’équilibre entre performance des utilisateurs des macrocellules et celle de leurs propres utilisateurs. Dans la seconde partie de la thèse, nous considérons le déploiement de picocellules sous l'égide de la macrocellule. Nous nous sommes intéressés ici aux solutions d’extension de l’aire picocellulaire qui permettent une meilleure association utilisateur/cellule permettant de réduire l’interférence mais aussi offrir une meilleure efficacité spectrale. Nous proposons donc une approche basée sur un modèle de prédiction de la mobilité des utilisateurs qui permet de mieux ajuster la proportion de bande passante à partager entre la macrocellule et la picocellule en fonction de la durée de séjour estimée de ces utilisateurs ainsi que de leur demandes en bande passante. Notre solution a permis d’offrir un bon compromis entre les débits réalisables de la Macro et des picocellules. / Recently the use of modern cellular networks has drastically changed with the emerging Long Term Evolution Advanced (LTE-A) technology. Homogeneous networks which were initially designed for voice-centric and low data rates face unprecedented challenges for meeting the increasing traffic demands of high data-driven applications and their important quality of service requirements. Therefore, these networks are moving towards the so called Heterogeneous Networks (HetNets). HetNets represent a new paradigm for cellular networks as their nodes have different characteristics such as transmission power and radio frequency coverage area. Consequently, a HetNet shows completely different interference characteristics compared to homogeneous deployment and attention must be paid to these disparities when different tiers are collocated together. This is mostly due to the potential spectrum frequency reuse by the involved tiers in the HetNets. Hence, efficient inter-cell interference mitigation solutions in co-channel deployments of HetNets remain a challenge for both industry and academic researchers. This thesis focuses on LTE-A HetNet systems which are based on Orthogonal Frequency Division Multiplexing Access (OFDMA) modulation. Our aim is to investigate the aggressive interference issue that appears when different types of base stations are jointly deployed together and especially in two cases, namely Macro-Femtocells and Macro-Picocells co-existence. We propose new practical power adjustment solutions for managing inter-cell interference dynamically for both cases. In the first part dedicated to Femtocells and Macrocell coexistence, we design a MBS-assisted femtocell power adjustment strategy which takes into account femtocells users performance while mitigating the inter-cell interference on victim macrocell users. Further, we propose a new cooperative and context-aware interference mitigation method which is derived for realistic scenarios involving mobility of users and their varying locations. We proved numerically that the Femtocells are able to maintain their interference under a desirable threshold by adjusting their transmission power. Our strategies provide an efficient means for achieving the desired level of macrocell/femtocell throughput trade-off. In the second part of the studies where Picocells are deployed under the umbrella of the Macrocell, we paid a special attention and efforts to the interference management in the situation where Picocells are configured to set up a cell range expansion. We suggest a MBS-assisted collaborative scheme powered by an analytical model to predict the mobility of Macrocell users passing through the cell range expansion area of the picocell. Our goal is to adapt the muting ratio ruling the frequency resource partitioning between both tiers according to the mobility behavior of the range-expanded users, thereby providing an efficient trade-off between Macrocell and Picocell achievable throughputs.