Marie Thilliez

Hybrid peer-to-peer model in proximity applications

The recent emergence of handheld devices and wireless networks has implied an exponential increase of terminals users. So, today, service providers have to propose new applications adapted to mobile environments. In this paper, we propose and describe a new class of distributed applications called the proximity applications. In such applications, two or more handheld devices, physically close to each other, can communicate and exchange data in a secure way. Proximity applications rely on the use of both different mobile devices and heterogeneous wireless networks. Thus, these applications need a high degree of flexibility, for an easy and rapid application development. In this context, our purpose is to study the interest of the Hybrid Peer-To-Peer (P2P) architecture model use specially for the extensibility, the fault-tolerance management and the scalability of proximity applications. Moreover, thanks to this model, proximity applications can easily face to the heterogeneity of devices and networks.

Evaluating Location Dependent Queries Using ISLANDS

The recent emergence of handheld devices and wireless networks implies an exponential increase in the numbers of terminals users. Given this increase, today’s service providers have to propose new applications adapted to mobile environments. In this article, we focus on distributed proximity M-services, in which several handheld devices, situated in close physical proximity to one another, can communicate and exchange data. Proximity M-services exploit a combination of mobile devices and heterogeneous mobile and/or fixed networks, and require a high degree of flexibility in order to permit easy and rapid application development. Because these applications are based on the Hybrid Peer-To-Peer (P2P) software architecture, such problems as scalability, deployment, security, reliability and information retrieval in M-services, can be more easily resolved than in other architectures. Within the framework of this software architecture, we focus on the localization problematic. Existing localization solutions are not well adapted to the mobility, dynamicity and heterogeneity of the Proximity M-service environment. Our solution to this lack of adaptation is ISLANDS, a service designed to aid the user in identifying and locating information and/or services within a communication area. ISLANDS is a decentralized service derived from the directory service model and adapted to the management of numerous distributed resources. In addition, ISLANDS provides querying facilities and a query evaluation model adapted to distributed directory management. Since most ISLANDS users are mobile, location dependent queries are also supported, thus enabling users to track down the closest theater or, more importantly in an emergency, the closest general practitioner’s office. Doing so obviously requires pinpointing the location of the query issuer, and one of the original aspects of our approach is the adaptability of the localization process : given that few handheld devices today are equipped with geo-localization features, ISLANDS is able to return approximate solutions by estimating the position of the client terminal, although ISLANDS can also use GPS-like techniques if available.

Multi-scale query processing in vehicular networks

In the last decade, a number of wireless and small-sized devices (e.g., PDAs, smartphones, sensors, laptops, etc.) with increasing computing capabilities have appeared in the market at very affordable costs. These devices have started to be embedded in modern cars in the form of on-board computers, GPS navigators, or even multimedia centers. Thus, the vehicles can carry useful information, acting as data sources for other vehicles. Recently, some works have addressed the problem of processing queries in such highly dynamic vehicular networks in order to share information between drivers. The proposed query processing techniques usually rely on a push model. Hence, each vehicle receives data from its neighbors and decides whether they are relevant enough to be stored in a local data cache. Then, the data may be used by a query processor to retrieve relevant data for the driver. In this paper, we look at the problem from a broader perspective and discuss the interest of multi-scale query processing techniques in such context. The goal of such techniques is to exploit, at the mobile device’s level, different access modes (e.g., push, pull) and various data sources (e.g., data cached locally, data stored by vehicles nearby, remote Web services, etc.) to provide the users with results for their queries. We highlight the most important challenges and outline some possible approaches. We also present a prototype of a first query evaluator developed using the Microsoft LINQ API.

A context aware framework for services management in the transportation domain

The transportation domain brings particular needs dealing with the specificities of the environment (highly mobile, distributed, unstable network connection, …). In this paper, we propose a dynamic adaptable framework responding to the needs of transportations applicative services. Among the most important services we can name positioning, time or communication capabilities. To achieve both this flexibility and automate context adaptation, we rely on a Service Oriented architecture and experiment our proposition on the inter-vehicular communication system, VESPA. This framework is based on the use of context informations for the transportation domain.

A Localization Service for Mobile Users in Peer-to-Peer Environments

The recent emergence of handheld devices and wireless networks has implied an exponential increase of terminals users. So, today, service providers have to propose new applications adapted to mobile environments. In this article, we describe a new class of distributed M-services called proximity applications. In such applications, two or more handheld devices, physically close to each other, can communicate and exchange data in a same communication area. These applications need a high degree of flexibility, for an easy and rapid application development. Based on the Hybrid Peer-To-Peer (P2P) software architecture, different problems such as scalability, deployment, security, reliability and information retrieval in M-services, can be more easily resolved. In this article, we focus on the information localization problematic in proximity applications. Existing localization solutions (naming services, trading services, etc.) are not well adapted to the dynamicity and the heterogeneity imposed in this environment. So, we propose a decentralized localization service relying on the directory service model and adapted to the management of numerous distributed resources. This service allows users to locate and discover information, particularly location based services retrieved in function of users location.

Requêtes dépendantes de la localisation : Evaluation distribuée et optimisation

Today, the evolution of handheld devices and mobile and/or wireless networks allows to propose new services adapted to users profiles and mobility. In this context, we highlighted a new class of mobile distributed applications: the proximity applications. These applications are based both on an already existing and fixed infrastructure and on mobile devices around this infrastructure. In this environment strongly distributed, we are interested in localization of available information in the application and particularly in the evaluation of location dependent queries such as: which is the closest bus stop? . In this paper, we describe an evaluation model to evaluate location dependent queries and the optimization strategies. The purpose of these strategies is to minimize the response times and resources consumption during the query evaluation and to guarantee a good response quality.

Vers l'interopérabilité des modèles de localisation géométrique et sémantique au travers de l'utilisation de métadonnées

This paper describes a metadata model dedicated to location data. These location data may be represented either using a geometric model (coordinates) or a semantic one (i.e. physical elements like city, street etc.). To answer the needs of upcoming Location Dependent Applications (LDA), the proposed metadata model allows to provide a service of correspondences between the two different location representation models to ensure the interoperability between them.