This research proposes a methodology to solve distributed optimization problems where quasi-autonomous decision entities directly interact with each other for partial information sharing. In the distributed system we study the quasi-autonomy arising from the assumption that each decision entity has complete and unique responsibility for a subset of decision variables. However, when solving a decision problem locally, consideration is given to how the local decisions affect overall system performance such that close-to-optimal solutions are obtained among all participating decision entities. Partial information sharing refers to the fact that no entity has the complete information access needed to solve the optimization problem globally. This condition hinders the direct application of traditional optimization solution methods. In this research, it is further assumed that direct interaction among the decision entities is allowed. This compensates for the lack of complete information access with the interactive exchange of non-private information. The methodology is tested in different application contexts: manufacturing capacity allocation, single machine scheduling, and jobshop scheduling. The experimental results show that the proposed method generates close-to optimal solutions in the tested problem settings.

Networked multi-agent systems have attracted lots of researchers to develop algorithms, techniques, and applications.A multi-agent networked system consists of more than one subsystem (agent) to cooperately solve a global problem with only local computations and communications in a fully distributed manner. These networked systems have been investigated in various different areas including signal processing, control system, and machine learning. We can see massive applications using networked systems in reality, for example, persistent surveillance, healthcare, factory manufacturing, data mining, machine learning, power system, transportation system, and many other areas. Considering the nature of those mentioned applications, traditional centralized control and optimization algorithms which require both higher communication and computational capacities are not suitable. Additionally, compared to distributed control and optimization approaches, centralized control, and optimization algorithms cannot be scaled into systems with a large number of agents, or guarantee performance and security. All of the limitations of centralized control and optimization algorithms motivate us to investigate and develop new distributed control and optimization algorithms in networked systems. Moreover, convergence rate and analysis are crucial in control and optimization literature, which motivates us to investigate how to analyze and accerlate the convergence of distributed optimization algorithms.

Cloud-radio access network (CRAN) is expected to be the core network architecture for next generation mobile radio system. In CRANs, joint signal processing is performed at multiple cloud computing centers (clouds) that are connected to several base stations (BSs) via high capacity backhaul links. As a result, large-scale interference management and network power consumption reduction can be effectively achieved. Unlike recent works on CRANs which consider a single cloud processing and treat inter-cloud interference as background noise, the first part of this thesis focuses on the more practical scenario of the downlink of a multi-cloud radio access network where BSs are connected to each cloud through wireline backhaul links. Assume that each cloud serves a set of pre-known single-antenna mobile users (MUs). This part focuses on minimizing the network total power consumption subject to practical constraints. The problems are solved using sophisticated techniques from optimization theory (e.g. Dual Decomposition-based algorithm and the alternating direction method of multipliers (ADMM)-based algorithm). One highlight of this part is that the proposed solutions can be implemented in a distributed fashion by allowing a reasonable information exchange between the coupled clouds. Additionally, feasible solutions of the considered optimization problems can be estimated locally at each iteration. Simulation results show that the proposed distributed algorithms converge to the centralized algorithms in a reasonable number of iterations. To further account of the backhaul congestion due to densification in CRANs, the second part of this thesis considers the downlink of a cache-enabled CRAN where each BS is equipped with a local-cache with limited size used to store the popular files without the need for backhauling. Further, each cache-enabled BS is connected to the cloud via limited capacity backhaul link and can serve a set of pre-known single antenna MUs. This part assumes that only imperfect channel state information (CSI) is available at the cloud. This part focuses on jointly minimizing the network total power consumption as well as backhaul cost. It then suggests solving this optimization problem using the majorization-minimization (MM) approach. Simulation results show that the proposed algorithm converges in a reasonable number of iterations.

<div>In recent years, the collective behaviors in nature have motivated rapidly expanding research efforts in the control of multi-agent systems. A multi-agent system is composed of multiple interacting subsystems (agents). In order to seek approaches that respect the network nature of multi-agent systems, distributed algorithms has recently received a significant amount of research attention, the goal of which is allowing multi-agent systems to accomplish global objectives through only local coordination. </div><div> Under this scope, we consider three major problems in this dissertation, namely, distributed computation, distributed optimization, and the resilience of distributed algorithms. First, for distributed computation, we devise distributed algorithms for solving linear equations, which can eliminate the initialization step for agents; converge to the minimum $l_1$ and $l_2$ solutions of under-determined linear equations; achieve ultimate scalability inters of agents' local storage and local states. Second, for distributed optimization, we introduce a new method for algorithm discretization so that the agents no longer have to carefully choose their step-size. We also introduce a new distributed optimization approach that can achieve better convergence rate with lower bandwidth requirement. Finally, for the resilience of distributed algorithms, we propose a new approach that allow normal agents in the multi-agent system to automatically isolate any false information from malicious agents without identification process. Though out the dissertation, all mentioned results are theoretically guaranteed and numerically validated.</div>

The feedback control system is defined as the sampling of an output signal and feeding it back to the input, resulting in an error signal that drives the overall system. This dissertation focuses on the stabilization and performance of state feedback control systems. Chapters 3 and 4 focus on the feedback control protocol approaching in the multi-agents system. In particular, the global regulation of distributed optimization problems has been considered. Firstly, we propose a distributed optimization algorithm based on the proportional-integral control strategy and the exponential convergence rate has been delivered. Moreover, a decentralized mechanism has been equipped to the proposed optimization algorithm, which enables an arbitrarily chosen agent in the system can compute the value of the optimal solution by only using the successive local states. After this, we consider the cost function follows the restricted secant inequality. A dynamic event-triggered mechanism design has been proposed. By ensuring the global regulation of the distributed proportional-integral optimization algorithm, the dynamic event-triggered mechanism efficiently reduces the communication frequency among agents. Chapter 5 focuses on the feedback control protocol approaching the single-agent system. Specifically, we investigate the truncated predictor feedback control of the regulation of linear input-delayed systems. For the purpose of improving the closed-loop performance, we propose a design of the truncated predictor feedback method with time-varying feedback parameters and give the potential range of choosing the time-varying feedback parameters to replace the traditional constant low gain parameters.

In this dissertation, we propose decentralized and scalable algorithms for Gaussian process (GP) training and prediction in multi-agent systems. The first challenge is to compute a spatial field that represents underwater acoustic communication performance from a set of measurements. We compare kriging to cokriging with vehicle range as a secondary variable using a simple approximate linear-log model of the communication performance. Next, we propose a model-based learning methodology for the prediction of underwater acoustic performance using a realistic propagation model. The methodology consists of two steps: i) estimation of the covariance matrix by evaluating candidate functions with estimated parameters; and ii) prediction of communication performance. Covariance estimation is addressed with a multi-stage iterative training method that produces unbiased and robust results with nested models. The efficiency of the framework is validated with simulations and experimental data from field trials. The second challenge is to perform predictions at unvisited locations with a team of agents and limited inter-agent information exchange. To decentralize the implementation of GP training, we employ the alternating direction method of multipliers (ADMM). A closed-form solution of the decentralized proximal ADMM is provided for the case of GP hyper-parameter training with maximum likelihood estimation. Multiple aggregation techniques for GP prediction are decentralized with the use of iterative and consensus methods. In addition, we propose a covariance-based nearest neighbor selection strategy that enables a subset of agents to perform predictions. Empirical evaluations illustrate the efficiency of the proposed methods / Doctor of Philosophy / In this dissertation, we propose decentralized and scalable algorithms for collaborative multiagent learning. Mobile robots, such as autonomous underwater vehicles (AUVs), can use predictions of communication performance to anticipate where they are likely to be connected to the communication network. The first challenge is to predict the acoustic communication performance of AUVs from a set of measurements. We compare two methodologies using a simple model of communication performance. Next, we propose a model-based learning methodology for the prediction of underwater acoustic performance using a realistic model. The methodology first estimates the covariance matrix and then predicts the communication performance. The efficiency of the framework is validated with simulations and experimental data from field trials. The second challenge regards the efficient execution of Gaussian processes using multiple agents and communicating as little as possible. We propose decentralized algorithms that facilitate local computations at the expense of inter-agent communications. Moreover, we propose a nearest neighbor selection strategy that enables a subset of agents to participate in the prediction. Illustrative examples with real world data are provided to validate the efficiency of the algorithms.

Le secteur industriel de l'énergie, et les réseaux électriques en particulier, rendent à nos sociétés modernes d'immenses services dont nous ne pouvons plus nous passer. Ils présentent aussi, hélas, un certain nombre de graves inconvénients, notamment en matière d'impact environnemental. Ces inconvénients apparaissent aujourd'hui comme inacceptables; le secteur de l'énergie s'efforce donc actuellement de les amoindrir autant que possible, dans le cadre de ce qu'on appelle la transition énergétique.Outre d'indispensables efforts en matière de sobriété et d'efficacité énergétique, deux grands axes d'amélioration se dessinent: d'une part, le remplacement progressif de certains moyens de production d'électricité conventionnels par des moyens de production renouvelables; et d'autre part, le transfert de certains usages aujourd'hui non-électriques vers l'électricité --- en particulier en matière de mobilité.L'intégration au réseau électrique de ces nouveaux types de dispositifs pose cependant des difficultés techniques considérables, qui motivent depuis le début des années 2000 de nombreux travaux sur le thème de ce que l'on appelle aujourd'hui les "smart grids": des réseaux électriques compatibles avec les exigences de la transition énergétique, c'est-à-dire capables d'accueillir massivement les nouveaux types d'usages comme la production photovoltaïque et les bornes de recharge des véhicules électriques, ceci notamment grâce à l'utilisation accrue des nouvelles technologies de l'information et de la communication. Parmi les difficultés susmentionnées, qui limitent la capacité d'accueil du réseau, figurent les congestions, c'est-à-dire les limites physiques à la puissance que l'on peut faire transiter d'un point à un autre sur une infrastructure donnée. C'est à la gestion des congestions que nos travaux sont consacrés. À ce sujet, la question fondamentale est de déterminer par quel enchaînement de mesures, de calcul, de communications et in fine d'actions, on peut passer d'une situation de contrainte sur un réseau de distribution d'électricité, à une situation où cette contrainte a été éliminée par l'action des flexibilités environnantes; autrement dit, en augmentant ou en réduisant judicieusement la production et/ou la consommation locales, et éventuellement en jouant sur d'autres types de leviers.L'objet de cette thèse est de participer à l'élaboration des outils conceptuels et informatiques qui nous permettront de répondre à la question fondamentale ci-dessus. Nos travaux portent ainsi sur la question de la modélisation des réseaux de distribution d'électricité "flexibles", et sur l'implémentation concrète des modèles retenus sous forme d'un logiciel de simulation ad hoc, parfaitement adapté à l'étude de ce type de réseaux. / The energy sector and the electrical networks in particular, provide great and indispensable services to our modern societies. Unfortunately, they also bring some serious drawbacks, especially with regard to the environment. These drawbacks are becoming more and more unacceptable; that is why the energy sector is trying to reduce them as much as possible, in the framework of the so-called energy transition.In addition to mandatory efforts in terms of energy efficiency and sobriety, two major directions of improvement have been identified: on the one hand, the progressive replacement of some conventional power plants with renewable production units; and on the other hand, the transfer of several non-electrical usages towards electricity --- in particular in the area of mobility.The integration of these new devices into electrical networks raise new technical challenges which, since the early 2000s, have been driving a lot of work about so-called "smart grids": electrical networks compatible with the requirements of the energy transition, ie. able to host new devices like photovoltaic solar panels and charging stations for electric vehicles, notably through the increasing usage of new information and communications technologies.Among the difficulties mentioned above, which limit the hosting capacity of the network, there are congestions ie. physical constraints limiting the amount of power that may be transmitted through a given infrastructure. Our work is devoted to the management of congestions. The fundamental issue thereon is to define a sequence of decisions, computations, communications and in fine actionsthat allows to move from a constrained situation on the electrical distribution network, to a situation in which the action of local flexibilities has lifted the constraint; in other words, to a situation where increasing or decreasing local generation and/or consumption, or taking some other control action, relieved the network.The aim of this thesis is to contribute to the development of conceptual and computing tools that will allow us to answer the fundamental aforementioned issue. Our work thus deals with the modelling of flexible electrical distribution networks, and with the tangible implementation of selected models in the form of ad hoc simulation software, specifically designed for the study of such networks.

As global climate change concerns, technological advancements, and economic shifts increase the adoption of electric vehicles, it is vital to study how best to integrate these into our existing energy systems. Electric vehicles (EVs) are on track to quickly become a large factor in the energy grid. If left uncoordinated, the charging of EVs will become a burden on the grid by increasing peak demand and overloading transformers. However, with proper charging control strategies, the problems can be mitigated without the need for expensive capital investments. Distributed control methods are a powerful tool to coordinate the charging, but it will be important to assess the trade-offs between performance, information privacy, and computational speed between different control strategies. This work presents a comprehensive comparison between four distributed control algorithms simulating two case studies constrained by dynamic transformer temperature and current limits. The transformer temperature dynamics are inherently nonlinear and this implementation is contrasted with a piece-wise linear convex relaxation. The more commonly distributed control methods of Dual Decomposition and Alternating Direction Method of Multipliers (ADMM) are compared against a relatively new algorithm, Augmented Lagrangian based Alternating Direction Inexact Newton (ALADIN), as well as against a low-information packetized energy management control scheme (PEM). These algorithms are implemented with a receding horizon in two distinct case studies: a local neighborhood scenario with EVs at each network node and a hub scenario where each node represents a collection of EVs. Finally, these simulation results are compared and analyzed to assess the methods’ performance, privacy, and processing metrics for each case study as no algorithm is found to be optimal for all applications.

Numerous control and decision problems in networked systems can be posed as optimization problems. Examples include the framework of network utility maximization for resource allocation in communication networks, multi-agent coordination in robotics, and collaborative estimation in wireless sensor networks (WSNs). In contrast to classical distributed optimization, which focuses on improving computational efficiency and scalability, these new applications require simple mechanisms that can operate under limited communication. In this thesis, we develop several novel mechanisms for distributed optimization under communication constraints, and apply these to several challenging engineering problems. In particular, we devise three tailored optimization algorithms relying only on nearest neighbor, also known as peer-to-peer, communication. Two of the algorithms are designed to minimize a non-smooth convex additive objective function, in which each term corresponds to a node in a network. The first method is an extension of the randomized incremental subgradient method where the update order is given by a random walk on the underlying communication graph, resulting in a randomized peer-to-peer algorithm with guaranteed convergence properties. The second method combines local subgradient iterations with consensus steps to average local update directions. The resulting optimization method can be executed in a peer-to-peer fashion and analyzed using epsilon-subgradient methods. The third algorithm is a center-free algorithm, which solves a non-smooth resource allocation problem with a separable additive convex objective function subject to a constant sum constraint. Then we consider cross-layer optimization of communication networks, and demonstrate how optimization techniques allow us to engineer protocols that mimic the operation of distributed optimization algorithms to obtain an optimal resource allocation. We describe a novel use of decomposition methods for cross-layer optimization, and present a flowchart that can be used to categorize and visualize a large part of the current literature on this topic. In addition, we devise protocols that optimize the resource allocation in frequency-division multiple access (FDMA) networks and spatial reuse time-division multiple access (TDMA) networks, respectively. Next we investigate some variants of the consensus problem for multi-robot coordination, for which it is usually standard to assume that agents should meet at the barycenter of the initial states. We propose a negotiation strategy to find an optimal meeting point in the sense that the agents' trajectories to the meeting point minimize a quadratic cost criterion. Furthermore, we also demonstrate how an augmented state vector can be used to boost the convergence rate of the standard linear distributed averaging iterations, and we present necessary and sufficient convergence conditions for a general version of these iterations. Finally, we devise a generic optimization software component for WSNs. To this end, we implement some of the most promising optimization algorithms developed by ourselves and others in our WSN testbed, and present experimental results, which show that the proposed algorithms work surprisingly well. / QC 20100813

Tradicionalment, les tècniques d'accés múltiple en sistemes de comunicacions multi-usuari han estat desenvolupades o bé orientades a la connexió o bé orientades al tràfic. En el primer cas, l'objectiu és establir tants canals ortogonals com sigui possible per tal d'assignar-los als usuaris. Aquesta idea va motivar el disseny de les estratègies més conegudes, com són FDMA, TDMA i CDMA. Per altra banda, però, els mètodes d'accés aleatori que s'iniciaren amb el conegut ALOHA pretenen compartir estadísticament un mateix medi de comunicació aprofitant la necessitat de transmetre la informació a ràfegues que s'origina en les xarxes de dades. Així, molts dels actuals sistemes es poden encabir dins d'aquest esquema si a més a més, tenim en compte combinacions d'aquestes. No obstant, sistemes moderns com el DVB-RCS en l'entorn de comunicacions digitals per satèl·lit o el WiMAX en l'accés terrestre de banda ampla implementen mecanismes de petició i assignació de recursos, els quals requereixen una gestió dinàmica d'aquests en el sistema (és el que s'anomena distribució dinàmica de l'amplada de banda en un sentit ampli).L'anterior concepte inclou múltiples variables, configuracions i protocols tant de capa física com de capa d'enllaç. En aquesta tesi s'exploren en primer lloc les bases matemàtiques que permeten coordinar les diferents capes de la divisió OSI dels sistemes i els distints nodes dins la xarxa. Ens referim a les tècniques de descomposició focalitzades en problemes d'optimització convexa, els quals han aportat, durant els últims anys, solucions elegants a molts problemes dins dels camps del processament del senyal i les comunicacions. Revisarem els esquemes coneguts i proposarem una nova metodologia. Acte seguit, es comparen les diferents possibilitats de descomposició, cadascuna de les quals implica diferents maneres d'establir la senyalització. A la pràctica, són aquestes diverses opcions de descomposició les que infereixen les diferents interaccions entre capes o els protocols de control entre elements de la xarxa. Els resultats en quant a nombre d'iteracions requerides per a convergir a la solució òptima són favorables al nou mètode proposat, la qual cosa obra noves línies d'investigació.Finalment, es contribueix també amb dos exemples d'aplicació, en DVB-RCS i en WiMAX. Formulem el problema de gestió de recursos resultant de l'accés múltiple dissenyat per cadascun dels sistemes com un problema de maximització d'utilitat de xarxa (conegut com a NUM en la bibliografia) i el solucionarem aplicant les tècniques anteriors. L'objectiu serà garantir l'equitativitat entre els usuaris i preservar, al mateix temps, la seva qualitat de servei. Per aconseguir-ho, cal seleccionar funcions d'utilitat adequades que permetin balancejar l'assignació de recursos cap als serveis més prioritaris. Mostrarem que en els escenaris considerats, l'ús del mètode proposat comporta guanys significatius ja que requereix menys iteracions en el procés (i per tant, menys senyalització) o bé menys temps de càlcul en un enfoc centralitzat (que es tradueix en la possibilitat d'incloure més usuaris). També es mostren els avantatges de considerar interaccions entre capes, ja que es poden ajustar els paràmetres de capa física per tal d'afavorir els tràfics més prioritaris o bé extreure els requeriments de servei de valors típicament disponibles en capes superiors.En general, la implementació eficient de tècniques de gestió dinàmica de recursos treballant en l'accés múltiple dels sistemes pot aportar guanys significatius però implica establir una bona coordinació entre capes i elements de xarxa. L'eina matemàtica que ho possibilita són les tècniques de descomposició. Cada nou escenari i sistema introdueix un nou repte d'optimització i la capacitat que tinguem de coordinar totes les variables del sistema cap al punt òptim en determinarà el rendiment global. / Traditionally, multiple access schemes in multi-user communications systems have been designed either connection-oriented or traffic-oriented. In the first ones, the goal was to provide as many orthogonal channels as possible, each one serving a different connection. That is the motivation of the so-called FDMA, TDMA and CDMA solutions. On the other hand, random access techniques, which started with the so-called ALOHA protocol, aim to statistically multiplex a shared communication medium by means of exploiting the random and bursty nature of transmission needs in data networks. Most of the multiple access solutions can be interpreted according to that classification or as a combination of those approaches. Notwithstanding, modern systems, such as the digital satellite communications standard DVB-RCS or the broadband wireless access WiMAX, have implemented a multiple access technique where users request for transmission opportunities and receive grants from the network, therefore requiring dynamic bandwidth allocation techniques. The concept of dynamic bandwidth allocation is wide and involves a number of physical and link layer variables, configurations and protocols. In this Ph.D. dissertation we first explore the mathematical foundation that is required to coordinate the distinct layers of the OSI protocol stack and the distinct nodes within the network. We talk about decomposition techniques focused on the resolution of convex programs, which have elegantly solved many problems in the signal processing and communications fields during the last years. Known schemes are reviewed and a novel decomposition methodology is proposed. Thereafter, we compare the four resulting strategies, each one having its own particular signalling needs, which results in distinct cross-layer interactions or signalling protocols at implementation level. The results in terms of iterations required to converge are favourable to the proposed method, thus opening a new line of research.Finally, we contribute with two practical application examples in the DVB-RCS and WiMAX systems. First, we formulate the dynamic bandwidth allocation problem that is derived from the multiple access schemes of both systems. Thereafter, the resulting Network Utility Maximization (NUM) based problem is solved by means of the previous decomposition mechanisms. The goal is to guarantee fairness among the users at the same time that Quality of Service (QoS) is preserved. In order to achieve that, we choose adequate utility functions that allow to balance the allocation towards the most priority traffic flows under a common fairness framework. We show that in the scenarios considered, the novel proposed coupled-decomposition method reports significant gains since it reduces significantly the iterations required (less iterations implies less signalling) or it reduces the time needed to obtain the optimal allocation when it is centrally computed (more users can be managed). We further show the advantages of cross-layer interactions with the physical and upper layers, which allow to benefit from more favourable adjustments of the transmission parameters and to consider the QoS requirements at upper layers. In general, an efficient implementation of dynamic bandwidth allocation techniques in Demand Assignment Multiple Access (DAMA) schemes may report significant performance gains but it requires proper coordination among system layers and network nodes, which is attained thanks to decomposition techniques. Each new scenario and system adds another optimization challenge and, as far as we are able to coordinate all the variables in the system towards that optimal point, the highest will be the revenue.