Utility companies (electricity, gas, and water suppliers), governments, and researchers recognize an urgent need to deploy communication-based systems to automate data collection from smart meters and sensors, known as Smart Metering Infrastructure (SMI) or Automatic Meter Reading (AMR). A smart metering system is envisaged to bring tremendous benefits to customers, utilities, and governments. The advantages include reducing peak demand for energy, supporting the time-of-use concept for billing, enabling customers to make informed decisions, and performing effective load management, to name a few. A key element in an SMI is communications between meters and utility servers. However, the mass deployment of metering devices in the grid calls for studying the scalability of communication protocols. SMI is characterized by the deployment of a large number of small Internet Protocol (IP) devices sending small packets at a low rate to a central server. Although the individual devices generate data at a low rate, the collective traffic produced is significant and is disruptive to network communication functionality. This research work focuses on the scalability of the transport layer functionalities. The TCP congestion control mechanism, in particular, would be ineffective for the traffic of smart meters because a large volume of data comes from a large number of individual sources. This situation makes the TCP congestion control mechanism unable to lower the transmission rate even when congestion occurs. The consequences are a high loss rate for metered data and degraded throughput for competing traffic in the smart metering network. To enhance the performance of TCP in a smart metering infrastructure (SMI), we introduce a novel TCP-based scheme, called Split- and Aggregated-TCP (SA-TCP). This scheme is based on the idea of upgrading intermediate devices in SMI (known in the industry as regional collectors) to offer the service of aggregating the TCP connections. An SA-TCP aggregator collects data packets from the smart meters of its region over separate TCP connections; then it reliably forwards the data over another TCP connection to the utility server. The proposed split and aggregated scheme provides a better response to traffic conditions and, most importantly, makes the TCP congestion control and flow control mechanisms effective. Supported by extensive ns-2 simulations, we show the effectiveness of the SA-TCP approach to mitigating the problems in terms of the throughput and packet loss rate performance metrics. A full mathematical model of SA-TCP is provided. The model is highly accurate and flexible in predicting the behaviour of the two stages, separately and combined, of the SA-TCP scheme in terms of throughput, packet loss rate and end-to-end delay. Considering the two stages of the scheme, the modelling approach uses Markovian models to represent smart meters in the first stage and SA-TCP aggregators in the second. Then, the approach studies the interaction of smart meters and SA-TCP aggregators with the network by means of standard queuing models. The ns-2 simulations validate the math model results. A comprehensive performance analysis of the SA-TCP scheme is performed. It studies the impact of varying various parameters on the scheme, including the impact of network link capacity, buffering capacity of those RCs that act as SA-TCP aggregators, propagation delay between the meters and the utility server, and finally, the number of SA-TCP aggregators. The performance results show that adjusting those parameters makes it possible to further enhance congestion control in SMI. Therefore, this thesis also formulates an optimization model to achieve better TCP performance and ensures satisfactory performance results, such as a minimal loss rate and acceptable end-to-end delay. The optimization model also considers minimizing the SA-TCP scheme deployment cost by balancing the number of SA-TCP aggregators and the link bandwidth, while still satisfying performance requirements.

Utility companies (electricity, gas, and water suppliers), governments, and researchers recognize an urgent need to deploy communication-based systems to automate data collection from smart meters and sensors, known as Smart Metering Infrastructure (SMI) or Automatic Meter Reading (AMR). A smart metering system is envisaged to bring tremendous benefits to customers, utilities, and governments. The advantages include reducing peak demand for energy, supporting the time-of-use concept for billing, enabling customers to make informed decisions, and performing effective load management, to name a few. A key element in an SMI is communications between meters and utility servers. However, the mass deployment of metering devices in the grid calls for studying the scalability of communication protocols. SMI is characterized by the deployment of a large number of small Internet Protocol (IP) devices sending small packets at a low rate to a central server. Although the individual devices generate data at a low rate, the collective traffic produced is significant and is disruptive to network communication functionality. This research work focuses on the scalability of the transport layer functionalities. The TCP congestion control mechanism, in particular, would be ineffective for the traffic of smart meters because a large volume of data comes from a large number of individual sources. This situation makes the TCP congestion control mechanism unable to lower the transmission rate even when congestion occurs. The consequences are a high loss rate for metered data and degraded throughput for competing traffic in the smart metering network. To enhance the performance of TCP in a smart metering infrastructure (SMI), we introduce a novel TCP-based scheme, called Split- and Aggregated-TCP (SA-TCP). This scheme is based on the idea of upgrading intermediate devices in SMI (known in the industry as regional collectors) to offer the service of aggregating the TCP connections. An SA-TCP aggregator collects data packets from the smart meters of its region over separate TCP connections; then it reliably forwards the data over another TCP connection to the utility server. The proposed split and aggregated scheme provides a better response to traffic conditions and, most importantly, makes the TCP congestion control and flow control mechanisms effective. Supported by extensive ns-2 simulations, we show the effectiveness of the SA-TCP approach to mitigating the problems in terms of the throughput and packet loss rate performance metrics. A full mathematical model of SA-TCP is provided. The model is highly accurate and flexible in predicting the behaviour of the two stages, separately and combined, of the SA-TCP scheme in terms of throughput, packet loss rate and end-to-end delay. Considering the two stages of the scheme, the modelling approach uses Markovian models to represent smart meters in the first stage and SA-TCP aggregators in the second. Then, the approach studies the interaction of smart meters and SA-TCP aggregators with the network by means of standard queuing models. The ns-2 simulations validate the math model results. A comprehensive performance analysis of the SA-TCP scheme is performed. It studies the impact of varying various parameters on the scheme, including the impact of network link capacity, buffering capacity of those RCs that act as SA-TCP aggregators, propagation delay between the meters and the utility server, and finally, the number of SA-TCP aggregators. The performance results show that adjusting those parameters makes it possible to further enhance congestion control in SMI. Therefore, this thesis also formulates an optimization model to achieve better TCP performance and ensures satisfactory performance results, such as a minimal loss rate and acceptable end-to-end delay. The optimization model also considers minimizing the SA-TCP scheme deployment cost by balancing the number of SA-TCP aggregators and the link bandwidth, while still satisfying performance requirements.

De façon à réduire l'impact du réchauffement climatique, la Commission Européenne a décidé dans le prolongement du protocole de Kyoto de réduire drastiquement les émissions de gaz à effet de serre. En la matière, le remplacement progressif des véhicules thermiques par des véhicules électriques est considéré comme un objectif clé. Dans ce but, plusieurs prérequis doivent être pris en compte pour le déploiement rapide des véhicules électriques sur le marché européen. Le premier de ces prérequis consiste en la conception d'une nouvelle génération de batteries avec une efficacité énergétique accrue, une plus grande capacité et une robustesse améliorée. Le second prérequis est la conception et le déploiement rapide et économique d'infrastructures de recharge. Enfin, l'émergence de normes européennes dans tous ces domaines est une impérieuse nécessité pour la viabilité économique du marché européen des véhicules électriques. Cette thèse se focalise essentiellement sur le second de ces trois prérequis. Il est largement admis qu'à court terme, l'usage des véhicules électriques sera essentiellement limité aux zones urbaines avant d'être étendu à tout le territoire. Tel que cela est pris en compte par le projet de recherche national TELEWATT dans lequel nous avons été impliqués, les réseaux d'éclairage publics en zones urbaines peuvent être utilisés judicieusement pour déployer à court terme des infrastructures de recharge bon marché. Le principe de base de ce projet consiste à connecter à chaque candélabre une ou deux stations de recharge pour véhicule électrique si une ou deux places de parking sont disponibles dans l'environnement proche. Diverses contraintes électriques spécifiées dans le projet TELEWATT doivent être satisfaites pour permettre la recharge de véhicules électriques sans perturber la qualité de l’éclairage. Dans cette thèse, nous commençons par démontrer la faisabilité d’une telle opération. Pour cela, nous avons développé un simulateur permettant de décrire le comportement dynamique du système global. La force et l’originalité de ce simulateur réside dans sa capacité à déterminer en temps-réel et avec laprécision nécessaire si un véhicule électrique peut effectivement être connecté à une borne. La réponse à cette question dépend des caractéristiques statiques et dynamiques du réseau d’éclairage et de l’état de charge des véhicules déjà connectés comme des véhicules candidats. La dynamique du processus de recharge de chaque véhicule électrique dépend fortement de la puissance instantanée consommée par l’infrastructure globale. Le second objectif original de cette thèse consiste en la conception de politiques d’ordonnancement d’activation des diverses bornes de recharge. Notre objectif est de proposer, via ces politiquesd’ordonnancement, divers types de qualité de service (QoS) garanties aux usagers. De telles garanties peuvent par exemple se décliner en termes d’état de charge attendu pour une certaine durée de stationnement. / In order to reduce the impact of global warming, the European Commission has decided in the continuation of the Kyoto protocol to reduce drastically greenhouse gas emissions. In this matter, the progressive replacement of thermal vehicles by electric ones is considered as a key objective. For that purpose, several prerequisites must be considered for the rapid deployment of Electrical Vehicles (EV) in the European market. The first of these prerequisites consists in the design of a new generation of batteries with higher energy efficiency, larger capacity and improved robustness. The second one is the design and rapid deployment of cost effective charging infrastructures. At last, the emergence of European standards in all these matters is an imperious necessity for the viability of the EVs European market. This thesis is mainly focused on the second of these three prerequisites. It is widely admitted that in the short term, EVs usage will be essentially limited to urban areas before being extended to the whole territory. As it is considered by the French National TELEWATT research project to which we have contributed, street lighting networks can be used judiciously to deploy in the short term low cost charging infrastructures. The basic principle of this project consists in coupling to each street light one or two EV's charging stations if one or two parking spots are located nearby. Various electrical constraints specified in the TELEWATT project must be satisfied to enable EVs' charging without degrading the quality of the lighting system. In this thesis, we begin to demonstrate the feasibility of such an operation. For that purpose, we have developed a simulator enablingto depict the dynamic behavior of the global system. The strength and originality of this simulator resides in its capacity to determine in real-time and with the necessary precision if an EV can effectively be connected to a charging station. The response to this question depends on static and dynamic characteristics of the street lighting network and on the state of charge of the batteries of both the connected and candidate EVs. The dynamism of the charging process of each EV strongly depends on the instantaneous power consumed by the global infrastructure. The second original result of this thesis consists in the design of scheduling policies to activate the various charging stations. Our objective is to propose via these scheduling policies, various types of guaranteed Quality of Service (QoS) to the end-users. Such guarantees can be declined for instance in terms of expected state of charge for a given parking term

Over the last several decades, electricity consumption and generation have continually grown. Investment in the Transmission and Distribution (T&D) infrastructure has been minimal and it has become increasingly difficult and expensive to permit and build new power lines. At the same time, a growing increase in the penetration of renewable energy resources is causing an unprecedented level of dynamics on the grid. Consequently, the power grid is congested and under stress. To compound the situation, the utilities do not possess detailed information on the status and operating margins on their assets in order to use them optimally. The task of monitoring asset status and optimizing asset utilization for the electric power industry seems particularly challenging, given millions of assets and hundreds of thousands of miles of power lines distributed geographically over millions of square miles. The lack of situational awareness compromises system reliability, and raises the possibility of power outages and even cascading blackouts. To address this problem, a conceptual Power Line Sensor Network (PLSN) is proposed in this research. The main objective of this research is to develop a distributed PLSN to provide continuous on-line monitoring of the geographically dispersed power grid by using hundreds of thousands of low-cost, autonomous, smart, and communication-enabled Power Line Sensor (PLS) modules thus to improve the utilization and reliability of the existing power system. The proposed PLSN specifically targets the use of passive sensing techniques, focusing on monitoring the real-time dynamic capacity of a specific span of a power line under present weather conditions by using computational intelligence technologies. An ancillary function is to detect the presence of incipient failures along overhead power lines via monitoring and characterizing the electromagnetic fields around overhead conductors. This research integrates detailed modeling of the power lines and the physical manifestations of the parameters being sensed, with pattern recognition technologies. Key issues of this research also include design of a prototype PLS module with integrated sensing, power and communication functions, and validation of the Wireless Sensor Network (WSN) technology integrated to this proposed PLSN.