Dans le contexte des réseaux de capteurs sans fil, la technique de localisation ”range-free” est plus efficiente, par rapport au principe ”range-based”. Par conséquent, nous avons focalisé nos travaux de cette thèse sur les techniques ”range-free”. Afin de permettre à chaque noeud mobile ou normal de choisir son propre algorithme de localisation, nous avons proposé un mécanisme adapté en scindant les noeuds normaux en deux classes: les noeuds de la première classe ont au moins 3 ancres voisines, alors que les noeuds de la deuxième classe ont moins de trois ancres voisines. Pour les noeuds normaux de la classe 1, nous avons proposé un nouvel algorithme ”Mid-perpendicular”. Pour les noeuds normaux de la classe 2, nous avons proposé deux nouveaux algorithmes ”Checkout DV-hop” et ”Selective 3-Anchor DV-hop”.Pour simuler et évaluer la performance de nos trois nouveaux algorithmes dans le contexte protocolaire des réseaux, nous avons pris soin de proposer deux protocoles associés : ”DV-hopprotocol” et ”Classe-1 protocol”. Par la suite, nous avons combiné ces deux protocoles pour obtenir notre ”adaptive range-free localization protocol”. Basé sur nos protocoles, en utilisant le simulateur WSNet, nous avons simulé différents algorithmes ”range-free” dans le contexte des réseaux de capteurs conformes au standard IEEE 802.15.4. Les résultats ont été présentés et analysés en termes de précision de la localisation, charge du réseau, mobilité des noeuds, et synchronisation de ces derniers. / In the context of wireless sensor networks, the range-free localization technique is more costeffective than the range-base scheme. Therefore, in this thesis we focus on the range-free technique.In order to permit each normal node to choose its suitable localization algorithm, we proposed anadaptive mechanism to categorize normal nodes into two classes: the normal nodes having at least 3 neighbor anchors are class-1 nodes, while others are class-2 nodes. For class-1 normal nodes, we proposed a new algorithm named as Mid-perpendicular. For class-2 normal nodes, we proposed twoalgorithms Checkout DV-hop and Selective 3-Anchor DV-hop.In order to simulate and evaluate the performance of our three new algorithms, we proposed two protocols: DV-hop protocol and Class-1 protocol. Then we combined these two protocols into our adaptive range-free localization protocol. Based on our protocols, using the network simulator WSNet,we simulate the concerned range-free localization algorithms in the IEEE 802.15.4 wireless network.The comparative network simulation results are presented and analyzed in terms of localization accuracy, overhead, node mobility, and node synchronization

Das Hauptproblem von Lokalisierungsalgorithmen für WSNs basierend auf Ankerknoten ist die Abhängigkeit von diesen. Mobilität im Netzwerk kann zu Topologien führen, in denen einzelne Knoten oder ganze Teile des Netzwerks temporär von allen Ankerknoten isoliert werden. In diesen Fällen ist keine weitere Lokalisierung möglich. Dies wirkt sich primär auf den Lokalisierungsfehler aus, der in diesen Fällen stark ansteigt. Des weiteren haben Betreiber von Sensornetzwerken Interesse daran, die Anzahl der kosten- und wartungsintensiveren Ankerknoten auf ein Minimum zu reduzieren. Dies verstärkt zusätzlich das Problem von nicht verfügbaren Ankerknoten während des Netzwerkbetriebs. In dieser Arbeit werden zunächst die Vor- und Nachteile der beiden großen Hauptkategorien von Lokalisierungsalgorithmen (range-based und range-free Verfahren) diskutiert und eine Studie eines oft für range-based Lokalisierung genutzten Distanzbestimmungsverfahren mit Hilfe des RSSI vorgestellt. Danach werden zwei neue Varianten für ein bekanntes range-free Lokalisierungsverfahren mit Namen MCL eingeführt. Beide haben zum Ziel das Problem der temporär nicht verfügbaren Ankerknoten zu lösen, bedienen sich dabei aber unterschiedlicher Mittel. SA-MCL nutzt ein dead reckoning Verfahren, um die Positionsschätzung vom letzten bekannten Standort weiter zu führen. Dies geschieht mit Hilfe von zusätzlichen Sensorinformationen, die von einem elektronischen Kompass und einem Beschleunigungsmesser zur Verfügung gestellt werden. PO-MCL hingegen nutzt das Mobilitätsverhalten von einigen Anwendungen in Sensornetzwerken aus, bei denen sich alle Knoten primär auf einer festen Anzahl von Pfaden bewegen, um den Lokalisierungsprozess zu verbessern. Beide Methoden werden durch detaillierte Netzwerksimulationen evaluiert. Im Fall von SA-MCL wird außerdem eine Implementierung auf echter Hardware vorgestellt und eine Feldstudie in einem mobilen Sensornetzwerk durchgeführt. Aus den Ergebnissen ist zu sehen, dass der Lokalisierungsfehler in Situationen mit niedriger Ankerknotendichte im Fall von SA-MCL um bis zu 60% reduziert werden kann, beziehungsweise um bis zu 50% im Fall von PO-MCL.

One of the most essential objectives in WSNs is to determine the spatial coordinates of a source or a sensor node having information. In this study, the problem of range measurement-based localization of a signal source or a sensor is revisited. The main challenge of the problem results from the non-convexity associated with range measurements calculated using the distances from the set of nodes with known positions to a xed sen- sor node. Such measurements corresponding to certain distances are non-convex in two and three dimensions. Attempts recently proposed in the literature to eliminate the non- convexity approach the problem as a non-convex geometric minimization problem, using techniques to handle the non-convexity. This study proposes a new fuzzy range-free sensor localization method. The method suggests using some notions of Euclidean geometry to convert the problem into a convex geometric problem. The convex equivalent problem is built using convex fuzzy sets, thus avoiding multiple stable local minima issues, then a gradient based localization algorithm is chosen to solve the problem. Next, the proposed algorithm is simulated considering various scenarios, including the number of available source nodes, fuzzi cation level, and area coverage. The results are compared with an algorithm having similar fuzzy logic settings. Also, the behaviour of both algorithms with noisy measurements are discussed. Finally, future extensions of the algorithm are suggested, along with some guidelines.

One of the most essential objectives in WSNs is to determine the spatial coordinates of a source or a sensor node having information. In this study, the problem of range measurement-based localization of a signal source or a sensor is revisited. The main challenge of the problem results from the non-convexity associated with range measurements calculated using the distances from the set of nodes with known positions to a xed sen- sor node. Such measurements corresponding to certain distances are non-convex in two and three dimensions. Attempts recently proposed in the literature to eliminate the non- convexity approach the problem as a non-convex geometric minimization problem, using techniques to handle the non-convexity. This study proposes a new fuzzy range-free sensor localization method. The method suggests using some notions of Euclidean geometry to convert the problem into a convex geometric problem. The convex equivalent problem is built using convex fuzzy sets, thus avoiding multiple stable local minima issues, then a gradient based localization algorithm is chosen to solve the problem. Next, the proposed algorithm is simulated considering various scenarios, including the number of available source nodes, fuzzi cation level, and area coverage. The results are compared with an algorithm having similar fuzzy logic settings. Also, the behaviour of both algorithms with noisy measurements are discussed. Finally, future extensions of the algorithm are suggested, along with some guidelines.

Advancement of micro-electro-mechanics and wireless communication have proliferated the deployment of large-scale wireless sensor networks. Due to cost, size and power constraints, at most a few sensor nodes can be equipped with a global positioning system; such nodes (whose positions can be accurately determined) are referred to as anchors. However, one can deter-mine the distance between two nearby sensors using some form of local communication. The problem of computing the positions of the non-anchor nodes from the inter-sensor distances and anchor positions is referred as sensor network localization (SNL). In this dissertation, our aim is to develop an accurate, efficient, and scalable localization algorithm, which can operate both in the presence and absence of anchors. It has been demon-strated in the literature that divide-and-conquer approaches can be used to localize large net-works without compromising the localization accuracy. The core idea with such approaches is to partition the network into overlapping subnetworks, localize each subnetwork using the available distances (and anchor positions), and finally register the subnetworks in a single coordinate system. In this regard, the contributions of this dissertation are as follows: We study the global registration problem and formulate a necessary “rigidity” condition for uniquely recovering the global sensor locations. In particular, we present a method for efficiently testing rigidity, and a heuristic for augmenting the partitioned network to enforce rigidity. We present a mechanism for partitioning the network into smaller subnetworks using cliques. Each clique is efficiently localized using multidimensional scaling. Finally, we use a recently proposed semidefinite program (SDP) to register the localized subnetworks. We develop a scalable ADMM solver for the SDP in question. We present simulation results on random and structured networks to demonstrate the pro-posed methods perform better than state-of-the-art methods in terms of run-time, accuracy, and scalability.

No description available.