Cette thèse a pour l’objectif principal d’identifier et de comprendre les facteurs clés qui contribuent à fidéliser le consommateur à un Drive dans un contexte de distribution cross-canal. Elle se compose de deux parties. Dans la première partie, nous avons d’abord mené une revue de littérature tournant autour de quatre champs de recherche : l’innovation commerciale, la stratégie de distribution cross-canal, le comportement cross-canal et la fidélité du consommateur. Elle a été combinée à une étude qualitative exploratoire. Ces deux étapes nous ont permis de proposer un modèle conceptuel de la qualité de service perçue du Drive et de ses effets sur les représentations mentales du consommateur ainsi que sur sa fidélité au Drive évalué. La deuxième partie est consacrée au teste les hypothèses, le modèle de recherche et à l’analyse des résultats. Pour cela, une enquête quantitative a été réalisée auprès de 500 utilisateurs du Drive. Les répondants ont été subdivisés en deux sous-échantillons (N1 = 150 et N2 = 350) pour réaliser les analyses factorielles exploratoires et confirmatoires. Le test du modèle général ainsi que des hypothèses de recherche a été conduit en recourant à la méthode des équations structurelles, plus précisément à la modélisation PLS-PM au moyen du logiciel XLSTAT 2017. Nos résultats confirment que dans un contexte de distribution cross-canal, l’évaluation globale du consommateur ne dépend pas d’un canal mais de l’intégration des canaux utilisés qui permet au consommateur de vivre une expérience à la fois originale, plaisants et intéressante. Cette dernière contribue à la création de la fidélité du consommateur à un Drive. / The main objective of this thesis is to identify and understand the key factors that contribute to consumer loyalty to a Drive-in in a cross-channel distribution context. It consists of two parts.In the first part, we present a literature review around four fields of research: commercial innovation, cross-channel distribution strategy, cross-channel behavior and consumer loyalty. We complete this part with an exploratory qualitative study (in-depth interviews with consumers). These two steps allowed us to propose a conceptual model of the perceived quality of service of the Drive-in and its effects on the mental representations of the consumer (perceived overall value and satisfaction) as well as on its fidelity to the evaluated Drive-in.The second part is devoted to testing the hypotheses, the research model, and the analysis of the results. For this, a quantitative survey was conducted among 500 consumers of Drive-in. Respondents were divided into two sub-samples (N1 = 150 and N2 = 350) to perform exploratory and confirmatory factor analyzes. The general model test and research hypotheses were conducted using the structural equations method. In detail, to run the PLS-PM (Partial least squares) modeling, we employed XLSTAT 2017 software. Our results confirm that in a context of cross-channel distribution, the overall evaluation of a consumer does not depend on a channel, but it is related to the integration of the used channels which creates an original, pleasant and interesting experience for consumers. With the integration of channels, Drive-in consumers have an original, pleasant and interesting experience which contributes to the creation of consumer loyalty.

<p>Communicating over a wireless channel poses many unique challenges not found in wired communication because of the special characteristics of the wireless channel. The capacity in a wireless network is typically scarce as a result of the limited bandwidth and many distinct phenomenons, like attenuation and interference, that work destructively on the received signals.</p><p>The Medium Access Control (MAC) layer is responsible for sharing this limited resource among the users. This allocation problem should be handled by considering the Quality of Service (QoS) requirements of each user as to maximize the utility. Efficient MAC algorithms are crucial in minimizing collisions between transmissions and thus achieving high utilization of the channel.</p><p>This thesis focuses on conflict resolution and service differentiation algorithms for wireless local area networks, where there is no central control of the channel and each sender independently contends for access.</p><p>In part I, we study three approaches to improve the IEEE 802.11(e) standards with focus on QoS. In the first approach, utility functions are considered, that model application preferences, to achieve service differentiation and maximize the aggregated utility. We provide algorithms for two subsidiary problems that arise from the maximization problem, and show that a near--optimal solution is found. In the second approach a collision detection algorithm for multicast transmissions is proposed, that increases the reliability for multicast compared to the protected unicast traffic. The third approach is an improved MAC algorithm for the QoS standard IEEE 802.11e. The improved algorithm outperforms the standard and achieves close to optimal performance for large number of scenarios, which significantly reduces the need of adjusting the contention parameters.</p><p>In part II, we focus on channel bursting protocols that use noise bursts to resolve channel conflicts. These protocols is capable of achieving very low collision probability. We propose two new bursting protocols, that achieve very high channel utilization, and show that the bursting technique has good fairness properties and provides efficient support for service differentiation. We also show that it is possible to reduce the number of bursts without loosing performance.</p><p>In part III, the optimal backoff distribution that minimizes the collision probability is derived. We then propose a heuristic backoff distribution with similar properties that yields high channel utilization. An extension for service differentiation is provided where the sizes of the backoff windows are adjusted.</p>

No description available.

A design of anautomatic network capacity markets, oftenreferred to as a bandwidth market, is presented. Three topicsare investigated. First, a network model is proposed. Theproposed model is based upon a trisection of the participantroles into network users, network owners, and market middlemen.The network capacity is defined in a way that allows it to betraded, and to have a well defined price. The network devicesare modeled as core nodes, access nodes, and border nodes.Requirements on these are given. It is shown how theirfunctionalities can be implemented in a network. Second, asimulated capacity market is presented, and a statisticalmethod for estimating the price dynamics in the market isproposed. A method for pricing network services based on sharedcapacity is proposed, in which the price of a service isequivalent to that of a financial derivative contract on anumber of simple capacity shares.Third, protocols for theinteraction between the participants are proposed. The marketparticipants need to commit to contracts with an auditableprotocol with a small overhead. The proposed protocol is basedon a public key infrastructure and on known protocols for multiparty contract signing. The proposed model allows networkcapacity to be traded in a manner that utilizes the networkeciently. A new feature of this market model, compared to othernetwork capacity markets, is that the prices are not controlledby the network owners. It is the end-users who, by middlemen,trade capacity among each-other. Therefore, financial, ratherthan control theoretic, methods are used for the pricing ofcapacity. <b>Keywords:</b>Computer network architecture, bandwidthtrading, inter-domain Quality-of-Service, pricing,combinatorial allocation, financial derivative pricing,stochastic modeling

Communicating over a wireless channel poses many unique challenges not found in wired communication because of the special characteristics of the wireless channel. The capacity in a wireless network is typically scarce as a result of the limited bandwidth and many distinct phenomenons, like attenuation and interference, that work destructively on the received signals. The Medium Access Control (MAC) layer is responsible for sharing this limited resource among the users. This allocation problem should be handled by considering the Quality of Service (QoS) requirements of each user as to maximize the utility. Efficient MAC algorithms are crucial in minimizing collisions between transmissions and thus achieving high utilization of the channel. This thesis focuses on conflict resolution and service differentiation algorithms for wireless local area networks, where there is no central control of the channel and each sender independently contends for access. In part I, we study three approaches to improve the IEEE 802.11(e) standards with focus on QoS. In the first approach, utility functions are considered, that model application preferences, to achieve service differentiation and maximize the aggregated utility. We provide algorithms for two subsidiary problems that arise from the maximization problem, and show that a near--optimal solution is found. In the second approach a collision detection algorithm for multicast transmissions is proposed, that increases the reliability for multicast compared to the protected unicast traffic. The third approach is an improved MAC algorithm for the QoS standard IEEE 802.11e. The improved algorithm outperforms the standard and achieves close to optimal performance for large number of scenarios, which significantly reduces the need of adjusting the contention parameters. In part II, we focus on channel bursting protocols that use noise bursts to resolve channel conflicts. These protocols is capable of achieving very low collision probability. We propose two new bursting protocols, that achieve very high channel utilization, and show that the bursting technique has good fairness properties and provides efficient support for service differentiation. We also show that it is possible to reduce the number of bursts without loosing performance. In part III, the optimal backoff distribution that minimizes the collision probability is derived. We then propose a heuristic backoff distribution with similar properties that yields high channel utilization. An extension for service differentiation is provided where the sizes of the backoff windows are adjusted.

We design, implement and evaluate statistical methods for managing uncertainty when consuming and provisioning resources in a federated computational market. To enable efficient allocation of resources in this environment, providers need to know consumers' risk preferences, and the expected future demand. The guarantee levels to offer thus depend on techniques to forecast future usage and to accurately capture and model uncertainties. Our main contribution in this thesis is threefold; first, we evaluate a set of techniques to forecast demand in computational markets; second, we design a scalable method which captures a succinct summary of usage statistics and allows consumers to express risk preferences; and finally we propose a method for providers to set resource prices and determine guarantee levels to offer. The methods employed are based on fundamental concepts in probability theory, and are thus easy to implement, as well as to analyze and evaluate. The key component of our solution is a predictor that dynamically constructs approximations of the price probability density and quantile functions for arbitrary resources in a computational market. Because highly fluctuating and skewed demand is common in these markets, it is difficult to accurately and automatically construct representations of arbitrary demand distributions. We discovered that a technique based on the Chebyshev inequality and empirical prediction bounds, which estimates worst case bounds on deviations from the mean given a variance, provided the most reliable forecasts for a set of representative high performance and shared cluster workload traces. We further show how these forecasts can help the consumers determine how much to spend given a risk preference and how providers can offer admission control services with different guarantee levels given a recent history of resource prices. / QC 20100909

No description available.

An ad hoc network is a collection of wireless mobile nodes dynamically forming a temporary network without the use of any fixed network infrastructure or centralized administration. Due to the limitations in the wireless environment, it may be necessary for one mobile host to enlist the aid of other hosts in forwarding a packet to its destination. In order to enable communication within the network, a routing protocol is needed to discover routes between nodes. The primary goal of ad hoc network routing protocols is to establish routes between node pairs so that messages may be delivered reliably and in a timely manner. The basic access method in IEEE 802.11 ad hoc networks is the Distributed Coordination Function (DCF), which provides a fair medium access. Enhanced Distributed Coordination Function (EDCF) has been developed to provide service differentiation among different traffic flows. In this thesis, we investigate the performance of the EDCF with routing protocols / Direct Sequenced Distance Vector (DSDV) and Dynamic Source Routing (DSR) by simulations.

de la TesisLas redes de hoy en día presentan un gran crecimiento, una alta complejidad de gestión y los nuevos servicios tienen requerimientos cada vez más estrictos. Es por ello que las plataformas de gestión de la década pasada resultan inadecuadas en los nuevos escenarios de red. Esta Tesis es una contribución a los nuevos esquemas de gestión para redes de gran escala, en especial es una contribución a los sistemas de gestión basados en políticas, sin perder por ello, la compatibilidad con los sistemas de gestión que se ocupan actualmente como por ejemplo SNMP, la gestión basada en agentes, etc. Las investigaciones relacionadas hasta ahora con los sistemas de gestión basados en políticas se enfocan principalmente en la gestión de recursos locales y en el control de admisión. La Tesis que se sustenta en este trabajo ofrece una perspectiva de la utilización de las políticas en un contexto más amplio, se propone una arquitectura para la gestión de red utilizando directorios y roles de políticas, analizando las políticas desde su fase de diseño hasta su configuración en los elementos físicos de la redSe considera que la creación de políticas pueden llevarla a cabo diferentes entidades, por ejemplo cuando las crea el administrador de la red, cuando los propios usuarios crean sus políticas (políticas personalizadas), o bien cuando la red basándose en un conjunto de políticas previamente definidas crea a partir de ellas nuevas políticas (metapolíticas). En esta Tesis la representación de las políticas de alto nivel se basa en los modelos propuestos por el IETF y DMTF, el Policy Core Information Model (PCIM) y sus extensiones (PCIMe). Se propone un esquema de clases orientadas a objetos para el almacenamiento de las políticas en un directorio LDAP (Lightweight Directory Access Protocol). Este esquema es una de las contribuciones que está Tesis realiza, la cual se ve reflejada en un draft realizado en conjunción con el grupo de trabajo de políticas del IETF.Debido a que no es posible implementar directamente las políticas de alto nivel en los elementos físicos de la red es necesario establecer un conjunto de parámetros de configuración de red que definan la política que debe aplicarse. Para resolver este mapeo se crearon perfiles SLS (Service Level Specification) basados en la especificación de nivel de servicio que el usuario acuerda con el proveedor de servicio Internet. En la implementación realizada se decidió utilizar cuatro perfiles, sin embargo la granularidad que se elija en la creación de perfiles SLS se deja abierta para que el administrador de la red cree los perfiles necesarios de acuerdo con las características topológicas de la red, los objetivos empresariales, etc.El directorio LDAP que se utiliza como repositorio de políticas almacena cientos o miles de políticas que son necesarias para resolver las diferentes tareas de gestión involucradas en un sistema de redes heterogéneas, esto puede afectar la ejecución del sistema, por lo tanto, se diseñaron métodos basados en roles de políticas para seleccionar la política o el conjunto de políticas adecuado que debe implementarse en la red en un momento especifico. Para resolver los conflictos que puedan ocurrir entre las políticas seleccionadas y evitar inconsistencias en la red, se crearon diversos módulos para la prevención y resolución de conflictos entre políticas. El primer proceso interviene en la creación de las políticas detectando conflictos sintácticos, es decir, se analiza que la política este correctamente diseñada y que pueda ser interpretada sin problemas por la red, posteriormente se verifica que la política pueda implementarse en los elementos de la topología de red que se utilice y que cubra los objetivos empresariales existentes.Para el caso de conflictos que puedan ocurrir en tiempo de ejecución se diseñó un método basado en espacios hiper-geométricos que permiten identificar un conflicto potencial e indicar la política adecuada que debe implementarse en la red. Dicho método está basado en una serie de métricas propuestas para definir cada servicio. Se realiza en la Tesis una aplicación de dicho método para el encaminamiento basado en restricciones de Calidad de Servicio en una red con Servicios Diferenciados y MPLS. / In today's telecommunications world the networks offer several new services involving higher and higher requirements, it means an increment of management complexity that cannot be adequately solved with the management platforms of previous years. This thesis is a contribution to new management schemes for big-scale networks; especially it is a set of contributions to the Policy-Based Management Systems (PBMS) without loosing compatibility with the current management systems such as SNMP, agent-based management, etc.Current research mainly proposes the use of policies to configure network local devices and admission control. This thesis works on a wide perspective about the use of policies. An efficiently architecture for network management on the basis of directories and policy roles is proposed, also there is a full analysis of policies from its design to its implementation in the network elements. The creation of policies can be carried out by different entities, for example network administrators, users (personalized policies) and when the network itself creates own policies based on a previous set of policies (metapolicies). In this thesis the representation of high-level policies is based on the Policy Core Information Model (PCIM) and their extensions (PCIMe) from the DMTF and the IETF. Policies are stored in a directory using the Lightweight Directory Access Protocol (LDAP) via an object oriented classes model designed in this thesis. These results led to an Internet draft for the policy-working group of the IETF.Because of direct implementation of high-level policies in the network elements it is not possible, it is necessary to establish a set of configuration parameters that define the policy that has to be enforced in the network. The methodology to map high-level policies to low-level policies is detailed in this thesis. Mapping processes involve the use of policy roles and profiles that come from Service Level Specifications (SLS) that users agree with the network. The implementation of the management system uses four SLS profiles but it is scalable to allow increasing profiles according to different aspects such as new services offered by the network, topology of the network, business goals, etc.The policy architecture manages heterogeneous interconnected networks, for this reason policy repositories have to be able of storing hundreds or thousands of policies in order to get the desired behavior in the entire network. Due to the fact that policy decision points have to choose adequate policies to apply in the network from a very big set of policies, the network performance could be affected. This thesis proposes an efficient selection and evaluation process on the basis of both, policy roles and the network status in a specific time.To solve possible conflicts that can occur between selected policies and avoid system inconsistencies, a set of models for the prevention and resolution of conflicts between policies is proposed. Prevention process has an algorithm to avoid syntactic conflicts and edition of new policies that produce potential conflicts with previous defined policies. Prevention process also considers congruency among policies, business goals and network topology.Conflict resolution process solves conflicts occurring during the performance of the system, this method is based on hyper geometrical spaces and policy roles to select the adequate policy, from the conflicting policies. These methods are presented in the Thesis with an application in a routing system with Quality of Service (QoS) restrictions into a network scenario based on Differentiated Services and the Multi Protocol Label Switching (MPLS).

The explosive growth in networked systems and applications and the increase in device capabilities (as evidenced by the availability of inexpensive multimedia devices) enable novel complex distributed applications, including video conferencing, on-demand computing services, and virtual environments. These applications' need for high performance, real-time, or reliability requires the provision of Quality of Service (QoS) guarantees along the path of information exchange between two or more communicating systems. Execution environments that are prone to dynamic variability and uncertainty make QoS provision a challenging task, e.g., changes in user behavior, resource requirements, resource availabilities, or system failures are difficult or even impossible to predict. Further, with the coexistence of multiple adaptation techniques and resource management mechanisms, it becomes increasingly important to provide an integrated or cooperative approach to distributed QoS management. This work's goals are the provision of system-level tools needed for the efficient integration of multiple adaptation approaches available at different layers of a system (e.g., application-level, operating system, or network) and the use of these tools such that distributed QoS management is performed efficiently with predictable results. These goals are addressed constructively and experimentally with the Q-Fabric architecture, which provides the required system-level mechanisms to efficiently integrate multiple adaptation techniques. The foundation of this integration is the event-based communication implemented by it, realizing a loosely-coupled group communication approach frequently found in multi-peer applications. Experimental evaluations are performed in the context of a mobile multimedia application, where the focus is directed toward efficient energy consumption on battery-operated devices. Here, integration is particularly important to prevent multiple energy management techniques found on modern mobile devices to negate the energy savings of each other.