Publications Scientifiques

Air transportation networks have been extensively studied in the network science literature. Researchers focus on airlines networks, national, regional, continental and worldwide networks using monoplex or multiplex approaches. Inspired by recent results on community-aware centrality measures [1], in this work, an extensive analysis of the macroscopic, mesoscopic and microscopic topological properties of the world air transportation network is performed. Based on the community structure uncovered by the Louvain algorithm, the original network is split into local components and global components. The local components are made of the communities by removing the intercommunity links. The global components are formed by removing all the intracommunity links and subsequent isolated nodes. Results reported in figure 1 shows that the main local components are geographically distributed over seven areas (North America, Europe, East & Southeast Asia, Africa-West and South Asia Oceania, South America, Russia). The main global component is distributed among all these areas (figure 2). A comparative analysis of the macroscopic, mesoscopic and microscopic properties of these various main networks reveal some consistencies. Indeed, density of all the components is low. However local components are all denser than the overall network. Transitivity is low and more particularly for the global component. All the networks are disassortative. Except for the Russia component, the hub dominance is relatively small. The study also reveals some dissimilarities. For example, the diameter of the local components ranges from 4 to 9 as compared to a value of 13 for the original network. The highest k-core subnetworks (k=31 and k=29) of the original network are mainly localized in the European component, and in the North America components. The maximum k-truss subnetwork (k=24) of the original network is in the North America component. It is followed by the European component (k=22). The remaining components exhibit low values of max k-truss. Comparative evaluation of degree and betweenness centrality measures allows to clearly identify the reasons behind the anomalies reported in several studies [2,3]. One of the main advantages of this approach is to get a better understanding of the similarity and differences of the various regional air transportation systems. Results indicate that it is more informative than considering the world air transportation network as a single entity. Indeed, if one considers the centrality measures, it allows to clearly distinguish airports having a regional influence from those with an international influence.

Wireless Sensor Networks (WSN) consists of a set of small and resources-constrained devices. They have received much attention over the last few years for the study and development of a plethora of potential applications. However, the common denominator of all applications of these sensor networks is the vulnerability of micro-sensors because of their limited material resources the most constraining of which is energy. In fact, wireless sensors are limited in terms of calculation, storage, battery, etc. Therefore, every possible solution that aims to conserve these resources is extensively sought. Thus, a great deal of researches has been conducted leading to an effective technique answering the established problem. This solution concerns the aggregation of data which is one of the techniques that is actually considered as an essential paradigm for sensor networks since it tends to save computation and communication resources. Data aggregation allows innetwork processing which leads to lesser packet transmissions and reduces redundancy, and therefore, helps in increasing network’s overall lifetime. However, sensor networks are usually deployed in unattended and hostile environments. Thus, the designer should not only consider the limited resources of the sensor nodes but also the security threats that can occur in an easily accessible network to the attacker. In this paper, we will give an overview study of the existing data aggregation solutions in wireless sensor networks then, we will propose a new efficient and secure approach to aggregation which can be evaluated with specific criteria.

In wireless sensor networks (WSN), grouping of small group nodes, called clusters, is an effective technique for facilitating scalability, self-organization, energy saving, access channels, routing, data aggre-gation, etc. Several clustering protocols have been proposed for WSN. Depending on the order in which the formation of clusters and the election of cluster leaders are performed, these protocols can be subdivided into two categories: the Leader-First (LF) and the Cluster-First (CF) approaches. In LF approaches, clusterheads are first elected according to certain metrics (eg, degree of connectivity, remaining energy), and then agree to assign the other nodes to different clusters. While in CF tech-niques, all sensors first form clusters, and then each cluster chooses its leader. In previous work, we proposed LQI-DCP, which is a non-secure distributed protocol for multihop clustering with an LF approach. Dur-ing the clustering process, one or more malicious nodes could attempt to disrupt the cluster leader election by electing illegitimate nodes. In this paper we propose LQI-DCPSec which is a secure version of LQI-DCP that we will compare, in terms of energy consumption, to SEFA which is another secure clustering algorithm using an LF approach. Simulation results show that LQI-DCPSec is more energy-efficient than SEFA.

In wireless sensor networks (WSN), grouping of small group nodes, called clusters, is an effective technique for facilitating scalability, self-organization, energy saving, access channels, routing, data aggregation, etc. Several clustering protocols have been proposed for WSN. Depending on the order in which the formation of clusters and the election of cluster leaders are performed, these protocols can be subdivided into two categories: the Leader-First (LF) and the Cluster-First (CF) approaches. In LF approaches, clusterheads are first elected according to certain metrics (e.g., degree of connectivity, remaining energy), and then agree to assign the other nodes to different clusters. While in CF techniques, all sensors first form clusters, and then each cluster chooses its leader. In previous work, we proposed LQI-DCP, which is a non-secure distributed protocol for multihop clustering with an LF approach. During the clustering process, one or more malicious nodes could attempt to disrupt the cluster leader election by electing illegitimate nodes. In this paper we propose LQI-DCPSec which is a secure version of LQI-DCP that we will compare, in terms of energy consumption, to SEFA which is another secure clustering algorithm using an LF approach. Simulation results show that LQI-DCPSec is more energy-efficient than SEFA.

In a wireless sensor network (WSN), sensors are typically deployed in an area in order to collect events related to one or more sensing features such as temperature, humidity, light, pressure, magnetic, accelerometer, acoustic activity, seismic movements, etc. The sensors communicate with each other to relay events to a Sink node or a Base Station that communicates with the network operator. Since the most critical feature in sensor networks is the hardship of its energy resources because each sensor node has initial low resources in terms of energy and it is not often possible to replace batteries during network operations. Moreover, equipped with a small processor generally associated to a small storage unit, sensors often operate without human assistance, and could be deployed in open and hostile environments. Therefore, wireless sensor networks are subject to various types of threats, vulnerabilities and attacks which are carried out in order to disclosure sensitive information, to disrupt network operations and to exhaust the batteries of the nodes and then to reduce their lifetimes, and consequently that of the network. In this paper, we will try to give an overview study of security issues and solutions in WSN.

In Wireless Sensor Networks (WSN), the nodes are often deployed to collect environmental data related to some particular relevant events. The collected data are then sent to a sink node or a base station following operations aimed at optimizing critical network resources such as storage, computing capacities and bandwidth. However, sensors have low processing and storage capabilities, small bandwidth and relatively low energetic resources. Thus, data aggregation is one of the most powerful techniques used to optimize sensor resources. Instead of sending data directly to the base station, the sensors send it first to an aggregation node that will compute a summary avoiding then redundant data and reducing the number of transmissions. During the data aggregation process, several attacks exploiting network vulnerabilities and weaknesses threaten sensor applications. In this paper we will give an overview study of the data aggregation process security issues and solutions in WSNs.

Wireless sensor networks are designed to collect information related to some events in hostile environments where human presence is not often possible. That is why we consider that once they are deployed, the sensors are autonomous. Their lifetime tightly rely on the manner of their batteries are utilized during network operations. Then, the most critical feature in sensor networks is the hardship of its energy resources because each sensor node has initial low resources in terms of energy and it's not often possible to replace batteries during network operations. Furthermore, Wireless Sensor Networks (WSN) are vulnerable, and subject to various passive and active attacks. So securing WSN lead up to add cryptography keys which make longer (and then more energy consumption) the packets exchanged during network running. Naturally, it becomes thrilling challenge to work to improve WSN security while wanting to optimize energy consumption. However, several works address WSN security without caring about energy. So in this paper we will study over-consumption of energy related to securing packets exchanged during WSN operations.

Le Génie Logiciel concerne tous les aspects de la production de logiciels à partir des premières étapes de la spécification du système jusqu’à la maintenance du système après qu’il soit entré en utilisation (Laurie Williams 2004). Il permet d’adopter une pratique d’un point de vue technique, une approche systématique et organisée pour développer des systèmes. Le cours expose les étudiants sur la façon d’utiliser plusieurs pratiques spécifiques (ou techniques) pour le développement de logiciels. Il est nécessaire de familiariser les apprenants à des outils de modélisation industriels dans le but de les exposer à des pratiques de l’état de l’art en matière de développement de logiciels. Ce faisant, l’étudiant aura une meilleure compréhension de la complexité ainsi que les subtilités des différentes activités de développement de logiciels qui incluent la collaboration dans une équipe ou un groupe. Les étudiants apprendront à construire un logiciel efficace, appliquer les bonnes pratiques, les techniques de conception efficaces et des outils de développement. Le Génie Logiciel est nécessaire pour développer toutes sortes de projets de logiciels, y compris des projets de logiciels complexes.

La programmation système est un type de programmation qui vise au développement de programmes qui font partie du système d’exploitation d’un ordinateur ou qui en réalisent les fonctions. Elle se distingue de la programmation des applications en ce qu’elle s’intéresse non pas au traitement des données, mais à la résolution des problèmes pour les humains, aux interfaces, aux protocoles et à la gestion des ressources, telles que le temps et l’espace. Donc, en réalité seuls les programmes d'application sont réellement utilisés par les utilisateurs. Les programmes système le sont implicitement. Elle inclut, en outre, l’accès aux fichiers, la programmation du clavier, de l’écran, des modems, la programmation réseau, et, en général, la programmation de tous les périphériques qui font entrer ou sortir de l’information d’un ordinateur, de la mémoire vive et des processeurs. Ce module est composé des unités suivantes : Unité 0 : Évaluation diagnostique Unité 1 : Librairie C et appels systèmes d’E/S Unité 2: Programmation shell et Code assembleur en ligne Unité 3: Processus, Threads et Gestion de la mémoire Unité 4: Communications interprocessus

Complex networks are distinguished to have a particular structure due to its large dimension and its many interactions, which play an important role in its characteristic. In fact, these networks are the locus of many dynamical phenomena such as birth of community, opinion formation, information diffusion, and rumors or epidemic spreading and so on. User mobility is of critical important when you want to know how disease spreads in network. The synthetic models represents a good alternative to describe how humans move. In this paper, we study the dynamics of epidemic spreading in temporal network generated by synthetic mobility model such as Random Waypoint (RWP), Gauss Markov (GM) and Truncated Levy Walk (TLW). Then, we show how proximity in terms in distance could impact a spreading process like droplet or airborne modes of pathogen spreading. Finally, we will statistically evaluate the number …

Wireless sensor networks face significant design challenges due to limited computing and storage capacities and, most importantly, dependence on limited battery power. Energy is a critical resource and is often an important issue to the deployment of sensor applications that claim to be omnipresent in the world of future. Thus optimizing the deployment of sensors becomes a major constraint in the design and implementation of a WSN in order to ensure better network operations. In wireless networking, clustering techniques add scalability, reduce the computation complexity of routing protocols, allow data aggregation and then enhance the network performance. The well-known MaxMin clustering algorithm was previously generalized, corrected and validated. Then, in a previous work we have improved MaxMin by proposing a Single-node Cluster Reduction (SNCR) mechanism which eliminates single-node clusters and then improve energy efficiency. In this paper, we show that MaxMin, because of its original pathological case, does not support the grid deployment topology, which is frequently used in WSN architectures. The unreliability feature of the wireless links could have negative impacts on Link Quality Indicator (LQI) based clustering protocols. So, in the second part of this paper we show how our distributed Link Quality based d-Clustering Protocol (LQI-DCP) has good performance in both stable and high unreliable link environments. Finally, performance evaluation results also show that LQI-DCP fully supports the grid deployment topology and is more energy efficient than MaxMin.

In wireless networking, clustering techniques add scal-ability, reduce the computation complexity of routing protocols, allow data aggregation and then enhance the network performance. The well known MaxMin clustering algorithm was previously generalized, corrected and validated. In [1] we improve MaxMin by proposing a Single-node Cluster Reduction (SNCR) mechanism which eliminates single-node clusters and then improve energy efficiency. In this paper, we show that MaxMin, because of its original pathological case, does not support the grid deployment topology, which is frequently used in WSN architectures. The unreliability feature of the wireless links could have negative impacts on Link Quality Indicator (LQI) based clustering protocols. So, in the second part of this paper we show how our distributed Link Quality based d-Clustering Protocol (LQI-DCP) has good performance in high unreliable link environments. Performance evaluation results also show that LQI-DCP fully supports the grid deployment topology.

In the process of epidemic spreading, the most important question is: how a healthy individual does catch a disease? To answer this question, researchers have developed plenty models depending on the assumption that said the contamination factor is a β coefficient, which is defined probabilistically, representing the infection rate. However, in the real life, the contagion process could depend on many factors like direct contact, indirect contact or proximity. In this paper, we use an epidemic model, SIS (Suspected, Infected, Suspected), that simulates in the framework of synthetic mobility models such as Random Waypoint (RWP), Gauss Markov (GM) and Truncated Levy Walk (TLW) and we show the impact of distance-proximity in the process of contamination.

We encounter in real world large networks as diverse as neural networks, power grid, financial networks, friendship networks, Internet, WWW. These networks are commonly characterized by a large number of highly interconnected dynamical entities. They are so complex that they may give birth to many dynamic phenomena. Understanding at the same time the topological and dynamical aspect of complex networks is an important challenge. In this paper, we review on the fundamental elements of complex networks study. We include definitions, measurements, models used to analyse topology and dynamics of these systems. Moreover, we discuss some questions related to mobility and dynamic community detection in complex systems.

In wireless sensor networks (WSN) energy efficiency of routing protocols is of primary importance. Embedded with local load balancing mechanisms, the proposed L2RP protocol is a link reliability based routing protocol which aims to help source nodes to exploit the potential capabilities of their respective neighbors. As it is a reliability-oriented protocol, L2RP discards unreliable links to avoid the substantial energy cost of packet losses. Simulation results show major efficiency benefits that stem from load balancing which helps in lengthening the network lifetime while minimizing packet losses. In WSN, the choice of a routing protocol and its key parameters depends on the nature of the application and on its primary mission. Lot of works addressed routing issues with more or less effectiveness, some of which pointed out the use of the link quality indicator (LQI) as a route selection criterion (metric). In a previous work, following an experimental study, we have shown, under some conditions, the inefficiency of the LQI based routing. In this paper, we propose through L2RP a simple way to improve reliability and efficiency of the LQI based routing in WSN. We also give a comparative study of several metrics including new definitions of LQI based metrics. Simulation results show that our adaptation of the LQI metric is among the best route selection criteria regardless of the performance criterion under consideration.

This article presents the real difficulties and topological challenges while deploying the senors networks for cold chain. At the highest level of deployment of a network, not only do we have to decide about the routing, but also have to come up with realistic scenarios. Link Quality based routing algorithms are used most of the time. In this context, routing parameters work well in simulation or in favorable conditions, but have limitations when we deploy them in the real network, especially when we have asymmetric links between the sensors. One of the answers is to add a few nodes with a high transmission power level which can act as the clusterhead and indeed these nodes really stabilize the system.

The choice of a routing protocol in a wireless sensor network (WSN) depends on the nature of the application and on its primary mission. Lot of works addressed the problem of routing mechanisms with more or less effectiveness, some of which pointed out the use of the link quality indicator (LQI) as a route selection criterion (metric). In a previous work, following an experimental study, we have shown [1], under some conditions, the inefficiency of the LQI based routing. In this paper, we propose a simple way, to improve reliability and efficiency of the LQI based routing in WSN. We also evaluate other route selection criteria including the "remaining energy level" and the "sensor proximity with respect to the Base Station (BS)". The proposed local load balancing routing protocol is aimed to help source nodes to exploit potential capabilities of their respective neighbors without having any knowledge of their properties such as the position or the energy level. Simulation results show that our adaptation of the LQI metric is among the best route selection criteria regardless of the performance criterion under consideration, and that the load balancing significantly improves the routing efficiency by lengthening the network lifetime while minimizing packet losses.

A large Ad Hoc network can be represented as several sets of clusters. Each cluster contains one or more nodes and has its clusterhead (or caryomme) chosen following an election based on an appropriate criterion. The clustering in wireless networking adds scalability, reduces the computation complexity of routing protocols, allows data aggregation and then enhances the network performance. Several studies have focused on clustering algorithms, providing some mechanisms well suited to wireless sensor networks (WSN). The MaxMin clustering algorithm proposed by [1] was generalized, corrected and validated in [2],[3] which added the use of criteria functions to choose appropriate clusterheads. We take part in this work by examining comparative results of different criteria that reveal single-node cluster phenomena. A single-node cluster is a cluster of which the only member is its clusterhead itself. In WSN clustering, the density of single-node clusters is a performance criterion. We use MaxMin in a cold chain monitoring application which shows that single-node clusters have negative impacts on the energy consumption. In this paper we compare several criteria including the "remaining energy" and the "link quality indicator (LQI)". We propose an effective manner of using the LQI which is among the best performant criteria. We also propose a simple mechanism to reduce single-node cluster phenomena which highly enhances the energy efficiency.