Papers by Philippe Canalda
2022 4th International Conference on Emerging Trends in Electrical, Electronic and Communications Engineering (ELECOM), 2022
Indoor localization has increased in popularity in recent years. Different technology has been pr... more Indoor localization has increased in popularity in recent years. Different technology has been proposed to provide indoor localization but many of them don’t give a good accuracy without expensive infrastructure. However, Fingerprinting based on Wi-Fi Received Signal Strength Indicator (RSSI) is widely used for indoor localization due to its simplicity and low hardware requirements. But, the RSSI values are affected by random fluctuation due to the fading and multi-path phenomena which lead to a decrease in the accuracy. In this paper, we propose an indoor localization using Machine learning algorithms such as Random Forest (RF), Extreme Gradient Boosting (XGboost), K-Nearest Neighbors (KNN), and Support-Vector Machine (SVM) combined with a Recursive Least Squares (RLS) filter. The method first uses Machine learning algorithms to establish an indoor positioning model, an implementation that achieves indoor positioning. Secondly, RLS is used to enhance the accuracy when the environment changes by reducing the noise in data. The performance of these methods is evaluated through extensive experiments in real-time indoor areas. The results are compared to the state-of-the-art and recently published papers. Our proposal has a good accuracy than other algorithms. The results are obtained from data collected in a real environment.
Le louage tunisien : une alternative sociétale opérationnelle aux VTC et à la ubérisation à tout-va
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
Ce travail s'inscrit dans la conception et la réalisation d'algorithmes de positionnement pair-à-... more Ce travail s'inscrit dans la conception et la réalisation d'algorithmes de positionnement pair-à-pair pour un environnement d'intérieur et pour des techniques de positionnement basées sur les réseaux radios 802.11.x. Dans les travaux de cette Thèse, nous avons souhaité étudier 3 techniques de base afin de montrer leur complémentarité pour améliorer la qualité (précision et la fréquence) de positionnement. L'étude rigoureuse s'appuie sur une méthodologie de tests construites sur plusieurs années, en l'occurrence, un jeu de 21 scénarios de tests permettant d'évaluer des solutions de positionnement centrées mobiles ou bien infrastructure, en tenant compte de l'orientation des antennes, de dispositions 4D (hauteur + modifications temporelles), de la présence humaine, d'obstacles matériel inopinés, des variations de températures ... Nous avons donc réalisé une étude de trois techniques de base de positionnement en intérieur basées sur les réseaux sans fils. Ces techniques sont une approche barycentrique fondée sur la N-latération et le modèle de propagation de signaux radio FRIIS, deux approches de fingerprint statique et dynamique. Les tests menées montrent qu'aucune des approches est fondée à procurer le meilleur positionnement, qu'une approche combinant ces 3 techniques est à même de proposer une amélioration substantielle du positionnement.
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific re... more HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
La mobilite intelligente dans le contexte urbain connait depuis ces dernieres annees une expansio... more La mobilite intelligente dans le contexte urbain connait depuis ces dernieres annees une expansion fulgurante. L’augmentation simultanee des flux de transport de personnes et de marchandises pousse les decideurs a trouver des solutions pour lutter contre plusieurs phenomenes propres au contexte urbain : congestion reguliere, stationnement indisponible, pollution environnementale et sonore, cout important du dernier kilometre. Les collectivites urbaines doivent reorganiser les infrastructures pour s’adapter en consequence aux besoins. On assiste ainsi a l’emergence de nombreux services urbains comme le co-voiturage dynamique, le transport a la demande, les vehicules « libre-service », le stationnement intelligent ou encore les centres de distribution urbains. Mais les challenges sont encore nombreux pour exploiter au mieux les divers modes disponibles afin de repondre aux besoins des usagers. L’UTBM (equipe OPERA) et l’UFC (Institut FEMTO-ST) ont decide de mutualiser leurs competence...
HAL (Le Centre pour la Communication Scientifique Directe), Sep 5, 2017
HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific re... more HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Open Source OwlPS 1.3: Towards a Reactive Wi-Fi Positioning System Sensitive to Dynamic Changes
Springer eBooks, 2013
ABSTRACT
This paper reports on asynchronous coordination of a virtual team in a virtual enterprise. It con... more This paper reports on asynchronous coordination of a virtual team in a virtual enterprise. It confronts two approaches : explicit coordination based on explicit process modeling, and implicit coordination, based on group awareness, to finally conclude that a good coordination is a subtle mixture of both approaches. For each approach and for the combination of both, requirements and design criteria are given and a study of the state of the art is done.
An adapted machine learning algorithm based-Fingerprints using RLS to improve indoor Wi-fi localization systems
Système Pair-à-Pair du Positionnement intérieur basé sur le réseaux wifi
HAL (Le Centre pour la Communication Scientifique Directe), Mar 9, 2017
HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific re... more HAL is a multidisciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not. The documents may come from teaching and research institutions in France or abroad, or from public or private research centers. L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de documents scientifiques de niveau recherche, publiés ou non, émanant des établissements d'enseignement et de recherche français ou étrangers, des laboratoires publics ou privés.
Impact of Environmental Factors for Wireless Network Based in Indoor Positioning Algorithms
The purpose of this paper is to show the impact of physical parameters (antenna gain, transmit an... more The purpose of this paper is to show the impact of physical parameters (antenna gain, transmit and receive frequency, transmit and receive power, etc.) and atmospheric conditions (temperature, test time, humidity,) in the wireless network based indoor positioning algorithms. This study focuses on the accuracy and quality of the calculated position relative to the actual position. We first present the OWLPS (Owl Positioning System) which is a positioning system based on IEEE 802.11 signal strength, designed, developed and released by the FEMTO-ST/DISC/OMNI team of University of Franche-comteé at Montbéliard. The data produced by the aggregator server are used to the positioning server in order to appreciate our approach. Finally, we will understand the results produced by our method with those produced by other algorithms.
IGI Global eBooks, May 25, 2010
Advances in technology have enabled a proliferation of mobile devices and a broad spectrum of nov... more Advances in technology have enabled a proliferation of mobile devices and a broad spectrum of novel and outbreaking solutions for new applications and services. Presently more and more people and companies are demanding mobile access to multimedia services such as real-time rich media. Today, it is necessary to be able to predict adaptation behavior that concerns and addresses not only the mobile usage or the infrastructure availability, but also the service quality, especially the continuity of service. Our chapter provides insight to new challenges of mobile multimedia services and applications: wifi indoor positioning system adapted to heterogeneous building, static and learning mobility prediction, predictive handover policy for multimedia cache management, mobile multimedia guide (e.g., museums), and network scalability.
Cimo: An efficient 2-phases calculator of multimodal itineraries for real trans-territories based on a dynamic programming
In this work we propose an exact solution for calculating multimodal itinerary. This solution is ... more In this work we propose an exact solution for calculating multimodal itinerary. This solution is named Cimo (Calculateur d'Itineraires Multimodaux Ordonnes). Cimo is an exact optimal itineraries' calculator wherein itineraries are sorted, multimodal, and trans-territorial. The solution is based on a dynamic programming algorithm "cut", "price" and "share". This solution is multi-objectives and multi-constraints. Several versions of this algorithm are proposed following a methodological approach that enables evaluation of efficiency and complexity's gain : through theoretical calculus and benchmarks. In the first version of realistic problem, we propose a solution with itineraries calculated exhaustively and satisfying all the constraints (version. 1.0), then a solution with incremental display of itineraries, including the optimal, following the in-depth course strategy and a smart sorted list of stations and the best itinerary previously calculated (version 1.2), another one's with the calculation of the best solution by implementing the cut on a subset of objectives, as well as the impossibility that a itinerary passes 2 times in the same station (version 2.0). One of the original features of Cimo is that it makes a preliminary phase of calculation of optimal itineraries, modulo the combinatorial instantiation of tri-modalities and also the very first satisfying transition schedule. During the second phase, all other objectives are valued following a certain hierarchy. The second novelty is that Cimo produces the best solution and it is speed-up by a smart sorted list of accessible stations from a reference station according to the degree of correspondence as well as proximity to the target station (version 3.0).
In this work, we propose a modeling and an implementation of an algorithm for the computerization... more In this work, we propose a modeling and an implementation of an algorithm for the computerization of the Nintra-modal carpooling with transhipment. It's a new type of carpooling that allows modal shifts and subsequently a variability between the offer and the demand. Another advantage of our calculator is that it checks the possibility of proposing a return itinerary for a driver who has already carried out a modal shift on his outward journey. Also, the calculator guarantees the good operation for the classic carpooling without modal shifts. Some dynamic constraints are managed by our proposed solution. A system modeling work and managed data, multiple constraints formalization and system multiple objectives as well as implementing a matching motor will be presented in this paper. This motor is based on a greedy combinatorial optimization algorithm. Various tests are run, that prove the possibility of solving this problem based on an exact approach in a reasonable time.
Enhanced Convex Hull based Clustering for High Population Density Avoidance under D2D Enabled Network
2021 IEEE 94th Vehicular Technology Conference (VTC2021-Fall), Sep 1, 2021
Global pandemics such as Covid-19 have led to massive loss of human lives and strict lockdown mea... more Global pandemics such as Covid-19 have led to massive loss of human lives and strict lockdown measures worldwide. To return to a certain level of normalcy, community awareness on avoiding high population density areas is significantly important for infection prevention and control. With the availability of new telecommunication technologies, it is possible to provide highly informative population clustering data back to people using wireless aerial agents (WAAs) placed in a local area. Hence, a service architecture that allows users to access the localization of population clusters is proposed. Further, a convex hull-based clustering method, enhanced population clustering (E-PC), is proposed. This method refined the result of conventional clustering methods such as K-means and Gaussian mixture model (GMM). Moreover, the potential in E-PC to achieve the same or higher results compared to the original K-means and GMM, while consuming lesser data points, is demonstrated. On average, E-PC improved the cluster detection performance in both K-means and GMM by 18.93% under different environments such as remote, rural, suburban, and urban in terms of silhouette score. Further, E-PC allows a 15% data reduction which results in decreasing the computational cost and energy consumption of the WAAs.
A Very First Geometric Dilution of Precision Proposal for Wireless Access Mobile Networks
... Philippe Canalda LIFC - Laboratoire d Informatique de lUniversite de Franche-Comte Montbeliar... more ... Philippe Canalda LIFC - Laboratoire d Informatique de lUniversite de Franche-Comte Montbeliard, France philippe.canalda@univ-fcomte.fr Francois Spies LIFC - Laboratoire d Informatique de lUniversite de Franche-Comte Montbeliard, France francois.spies@univ-fcomte.fr ...
Iterative Algorithms on Heterogeneous Network Computing: Parallel Polynomial Root Extracting
Lecture Notes in Computer Science, 2002
Page 307. Iterative Algorithms on Heterogeneous Network Computing: Parallel Polynomial Root Extra... more Page 307. Iterative Algorithms on Heterogeneous Network Computing: Parallel Polynomial Root Extracting Raphael Couturier, Philippe Canalda, and Francois Spies Laboratoire d'Informatique LIFC, University of Pranche-Comte, France Abstract. ...
Uploads
Papers by Philippe Canalda