In this paper, we study the stochastic transportation-inventory network design problem involving one supplier and multiple retailers. Each retailer faces some uncertain demand. Due to this uncertainty, some amount of safety stock must be maintained to achieve suitable service levels. However, risk-pooling benefits may be achieved by allowing some retailers to serve as distribution centers (and therefore inventory storage locations) for other retailers. The problem is to determine which retailers should serve as distribution centers and how to allocate the other retailers to the distribution centers. Shen et al. (2000) and Daskin et al. (2001) formulated this problem as a set-covering integer-programming model. The pricing subproblem that arises from the column generation algorithm gives rise to a new class of submodular function minimization problem. They only provided efficient algorithms for two special cases, and assort to ellipsoid method to solve the general pricing problem, which run in O(n⁷ log(n)) time, where n is the number of retailers. In this paper, we show that by exploiting the special structures of the pricing problem, we can solve it in O(n² log n) time. Our approach implicitly utilizes the fact that the set of all lines in 2-D plane has low VC-dimension. Computational results show that moderate size transportation-inventory network design problem can be solved efficiently via this approach. / Singapore-MIT Alliance (SMA)

A scenario where multiple entities interact with a common environment to achieve individual and common goals either co-operatively or competitively can be classified as a Multi-Agent System. In this thesis, we concentrate on the situations where the agents exhibit selfish, competitive and strategic behaviour, giving rise to interesting game theoretic and optimization problems. From a computational point of view, the presence of multiple agents introduces strategic and temporal issues, apart from enhancing the difficulty of optimization. We study the following natural mathematical models of such multi-agent problems faced in practice: a) combinatorial optimization problems with multi-agent submodular cost functions, b) combinatorial auctions with partially public valuations and c) online vertex-weighted bipartite matching and single bid budgeted allocations. We provide approximation algorithms, online algorithms and hardness of approximation results for these problems.

Funções submodulares aparecem naturalmente em diversas áreas, tais como probabilidade, geometria e otimização combinatória. Pode-se dizer que o papel desempenhado por essas funções em otimização discreta é similar ao desempenhado por convexidade em otimização contínua. Com efeito, muitos problemas em otimização combinatória podem ser formulados como um problema de minimizar uma função submodular sobre um conjunto apropriado. Além disso, submodularidade está presente em vários teoremas ou problemas combinatórios e freqüentemente desempenha um papel essencial em uma demonstração ou na eficiência de um algoritmo. Nesta dissertação, estudamos aspectos estruturais e algorítmicos de funções submodulares, com ênfase nos recentes avanços em algoritmos combinatórios para minimização dessas funções. Descrevemos com detalhes os primeiros algoritmos combinatórios e fortemente polinomiais para esse propósito, devidos a Schrijver e Iwata, Fleischer e Fujishige, além de algumas outras extensões. Aplicações de submodularidade em otimização combinatória também estão presentes neste trabalho. / Submodular functions arise naturally in various fields, including probability, geometry and combinatorial optimization. The role assumed by these functions in discrete optimization is similar to that played by convexity in continuous optimization. Indeed, we can state many problems in combinatorial optimization as a problem of minimizing a submodular function over an appropriate set. Moreover, submodularity appears in many combinatorial theorems or problems and frequently plays an essencial role in a proof or an algorithm. In this dissertation, we study structural and algorithmic aspects of submodular functions. In particular, we focus on the recent advances in combinatorial algorithms for submodular function minimization. We describe in detail the first combinatorial strongly polynomial-time algorithms for this purpose, due to Schrijver and Iwata, Fleischer, and Fujishige, as well as some extensions. Some applications of submodularity in combinatorial optimization are also included in this work.

With the advent of Network Function Virtualization (NFV), network services that traditionally run on proprietary dedicated hardware can now be realized using Virtual Network Functions (VNFs) that are hosted on general-purpose commodity hardware. This new network paradigm offers a great flexibility to Internet service providers (ISPs) for efficiently operating their networks (collecting network statistics, enforcing management policies, etc.). However, introducing NFV requires an investment to deploy VNFs at certain network nodes (called VNF-nodes), which has to account for practical constraints such as the deployment budget and the VNF-node limited resources. While gradually transitioning to NFV, ISPs face the problem of where to efficiently introduce NFV; here, we measure the efficiency by the amount of traffic that can be served in an NFV-enabled network. This problem is non-trivial as it is composed of two challenging subproblems: 1) placement of VNF-nodes; 2) allocation of the VNF-nodes' resources to network flows. These two subproblems must be jointly considered to satisfy the objective of serving the maximum amount of traffic. We first consider this problem for the one-dimensional setting, where all network flows require one network function, which requires a unit of resource to process a unit of flow. In contrast to most prior work that often neglects either the budget constraint or the resource allocation constraint, we explicitly consider both of them and prove that accounting for them introduces several new challenges. Specifically, we prove that the studied problem is not only NP-hard but also non-submodular. To address these challenges, we introduce a novel relaxation method such that the objective function of the relaxed placement subproblem becomes submodular. Leveraging this useful submodular property, we propose two algorithms that achieve an approximation ratio of $\frac{1}{2}(1-1/e)$ and $\frac{1}{3}(1-1/e)$ for the original non-relaxed problem, respectively. Next, we consider the multi-dimensional setting, where flows can require multiple network functions, which can also require a different amount of each resource to process a unit of flow. To address the new challenges arising from the multi-dimensional setting, we propose a novel two-level relaxation method that allows us to draw a connection to the sequence submodular theory and utilize the property of sequence submodularity along with the primal-dual technique to design two approximation algorithms. Finally, we perform extensive trace-driven simulations to show the effectiveness of the proposed algorithms. While the NFV paradigm offers great flexibility to network operators for efficient management of their networks, VNF instances are typically more prone to error and more vulnerable to security threats compared with dedicated hardware devices. Therefore, the NFV paradigm also poses new challenges concerning failure resilience. That has motivated us to consider robustness with respect to the class of sequence submodular function maximization problem, which has a wide range of applications, including those in the NFV domain. Submodularity is an important property of set functions and has been extensively studied in the literature. It models set functions that exhibit a diminishing returns property, where the marginal value of adding an element to a set decreases as the set expands. This notion has been generalized to considering sequence functions, where the order of adding elements plays a crucial role and determines the function value; the generalized notion is called sequence (or string) submodularity. In this part of the dissertation, we study a new problem of robust sequence submodular maximization with cardinality constraints. The robustness is against the removal of a subset of elements in the selected sequence (e.g., due to malfunctions or adversarial attacks). Compared to robust submodular maximization for set function, new challenges arise when sequence functions are concerned. Specifically, there are multiple definitions of submodularity for sequence functions, which exhibit subtle yet critical differences. Another challenge comes from two directions of monotonicity: forward monotonicity and backward monotonicity, both of which are important to proving performance guarantees. To address these unique challenges, we design two robust greedy algorithms: while one algorithm achieves a constant approximation ratio but is robust only against the removal of a subset of contiguous elements, the other is robust against the removal of an arbitrary subset of the selected elements but requires a stronger assumption and achieves an approximation ratio that depends on the number of the removed elements. Finally, we consider important problems that arise in the production networks, where packets need to pass through an ordered set of network functions called Service Function Chains (SFC) before reaching the destination. We study the following problems: (1) How to find an SFC-constrained shortest path between any pair of nodes? (2) What is the achievable SFC-constrained maximum flow? We propose a transformation of the network graph to minimize the computational complexity of subsequent applications of any shortest path algorithm. Moreover, we formulate the SFC-constrained maximum flow problem as a fractional multicommodity flow problem and develop a combinatorial algorithm for a special case of practical interest. / Computer and Information Science

Decision-making is an essential topic for multi-robot coordination and collaboration and is also the main topic of this thesis. Examples can be found in autonomous driving, environmental monitoring, intelligent transportation, etc. To study this problem, we first use multiple applications as motivating examples and then construct the general formulation and solution for those applications. Finally, we extend our investigation from the fundamental problem formulation to resilient and decentralized versions. All those problems are studied in the combinatorial optimization domain with the help of submodular and matroid optimization techniques. As a motivating example, we use a multi-robot environmental monitoring problem to extract the general formulation of a multi-robot decision-making problem. Consider the problem of deploying multi-agent teams for environmental monitoring in a precision farming application. We want to answer the question of when and where to deploy our robots. This is a typical task allocation problem in multi-robot systems. Using the above problem as an example, we first focus on this decision-making problem, e.g., intermittent deployment problem, in a centralized scenario. Given a predictable agriculture environment, we want to make decisions for robots for this monitoring task. The problem is formulated as a combinatorial submodular optimization with matroid constraints. By utilizing the properties of submodularity, we aim to develop a solution with performance guarantees. This motivating example demonstrates how to use a submodular function and matroids to model and solve decision-making problems in multi-robot systems. Based on this framework, we continue to explore the fundamental decision-making problem in several other directions in multi-robot systems, including the robust decision-making problem. All those problems and solutions are formulated and considered in a centralized scenario. In the second part of this thesis, we switch our focus from centralized to decentralized scenarios. We first investigate a case where the robots in a distributed multi-robot system need to work together to guard the system against worst-case attacks while making decisions. By worst-case attacks, we refer to the case where the system may have up to $K$ sensor failures. To increase resilience, we propose a fully distributed algorithm to guide each robot's action selection when the system is attacked. The proposed algorithm guarantees performance in a worst-case scenario where up to a portion of the robots malfunction due to attacks. Based on this specific task allocation problem in robotics, we then create a unified framework for a more general case in a decentralized scenario, e.g., asynchronous decentralized decision-making problems with matroid and knapsack constraints. Finally, several applications in decentralized scenarios are used to validate the theoretical guaranteed performance in robotics. / Doctor of Philosophy / Robots have been widely used as mobile sensing agents nowadays in various applications. Especially with the help of multi-robot systems and artificial intelligence, our lives have changed dramatically in the last decades. One of the most fundamental questions is how to utilize multi-robot systems to finish tasks successfully. To answer this, we need first to formulate the problem from applications and then find theoretically guaranteed answers to those questions. Meanwhile, the robustness and resilience of the solution also need to be taken care of, as cyber-attacks or system failures can happen everywhere. Motivated by those two main goals, this thesis will first use multiple applications to introduce the thesis's topic. We then provide solutions to those problems in centralized and decentralized scenarios. Meanwhile, to increase the system's ability to handle failures, we need to answer how to improve the robustness and resilience of the proposed solutions. Therefore, the topic of this thesis spread from problem formulation to failure-proof solutions. The result of this thesis can be widely used in multi-robot decision-making applications, including autonomous driving, intelligent transportation, and other cyber-physical systems.

This work is inspired by robotic search applications where a robot or team of robots is equipped with sensors and tasked to autonomously acquire as much information as possible from a region of interest. To accomplish this task, robots must plan paths through the region of interest that maximize the effectiveness of the sensors they carry. Receding horizon path planning is a popular approach to addressing the computationally expensive task of planning long paths because it allows robotic agents with limited computational resources to iteratively construct a long path by solving for an optimal short path, traversing a portion of the short path, and repeating the process until a receding horizon path of the desired length has been constructed. However, receding horizon paths do not retain the optimality properties of the short paths from which they are constructed and may perform quite poorly in the context of achieving the robotic search objective. The primary contributions of this work address the worst-case performance of receding horizon paths by developing methods of using terminal rewards in the construction of receding horizon paths. We prove that the proposed methods of constructing receding horizon paths provide theoretical worst-case performance guarantees. Our result can be interpreted as ensuring that the receding horizon path performs no worse in expectation than a given sub-optimal search path. This result is especially practical for subsea applications where, due to use of side-scan sonar in search applications, search paths typically consist of parallel straight lines. Thus for subsea search applications, our approach ensures that expected performance is no worse than the usual subsea search path, and it might be much better. The methods proposed in this work provide desirable lower-bound guarantees for a single robot as well as teams of robots. Significantly, we demonstrate that existing planning algorithms may be easily adapted to use our proposed methods. We present our theoretical guarantees in the context of subsea search applications and demonstrate the utility of our proposed methods through simulation experiments and field trials using real autonomous underwater vehicles (AUVs). We show that our worst-case guarantees may be achieved despite non-idealities such as sub-optimal short-paths used to construct the longer receding horizon path and unreliable communication in multi-agent planning. In addition to theoretical guarantees, An important contribution of this work is to describe specific implementation solutions needed to integrate and implement these ideas for real-time operation on AUVs. / Doctor of Philosophy / This work is inspired by robotic search applications where a robot or team of robots is equipped with sensors and tasked to autonomously acquire as much information as possible from a region of interest. To accomplish this task, robots must plan paths through the region of interest that maximize the effectiveness of the sensors they carry. Receding horizon path planning is a popular approach to addressing the computationally expensive task of planning long paths because it allows robotic agents with limited computational resources to iteratively construct a long path by solving for an optimal short path, traversing a portion of the short path, and repeating the process until a receding horizon path of the desired length has been constructed. However, receding horizon paths do not retain the optimality properties of the short paths from which they are constructed and may perform quite poorly in the context of achieving the robotic search objective. The primary contributions of this work address the worst-case performance of receding horizon paths by developing methods of using terminal rewards in the construction of receding horizon paths. The methods proposed in this work provide desirable lower-bound guarantees for a single robot as well as teams of robots. We present our theoretical guarantees in the context of subsea search applications and demonstrate the utility of our proposed methods through simulation experiments and field trials using real autonomous underwater vehicles (AUVs). In addition to theoretical guarantees, An important contribution of this work is to describe specific implementation solutions needed to integrate and implement these ideas for real-time operation on AUVs.

In this dissertation, we study multi-robot coordination in the context of multi-target tracking. Specifically, we are interested in the coordination achieved by means of submodular function optimization. Submodularity encodes the diminishing returns property that arises in multi-robot coordination. For example, the marginal gain of assigning an additional robot to track the same target diminishes as the number of robots assigned increases. The advantage of formulating coordination problems as submodular optimization is that a simple, greedy algorithm is guaranteed to give a good performance. However, often this comes at the expense of unrealistic models and assumptions. For example, the standard formulation does not take into account the fact that robots may fail, either randomly or due to adversarial attacks. When operating in uncertain conditions, we typically seek to optimize the expected performance. However, this does not give any flexibility for a user to seek conservative or aggressive behaviors from the team of robots. Furthermore, most coordination algorithms force robots to communicate at each time step, even though they may not need to. Our goal in this dissertation is to overcome these limitations by devising coordination algorithms that are parsimonious in communication, allow a user to manage the risk of the robot performance, and are resilient to worst-case robot failures and attacks. In the first part of this dissertation, we focus on designing parsimonious communication strategies for target tracking. Specifically, we investigate the problem of determining when to communicate and who to communicate with. When the robots use range sensors, the tracking performance is a function of the relative positions of the robots and the targets. We propose a self-triggered communication strategy in which a robot communicates its own position with its neighbors only when a certain set of conditions are violated. We prove that this strategy converges to the optimal robot positions for tracking a single target and in practice, reduces the number of communication messages by 30%. When tracking multiple targets, we can reduce the communication by forming subsets of robots and assigning one subset to track a target. We investigate a number of measures for tracking quality based on the observability matrix and show which ones are submodular and which ones are not. For non-submodular measures, we show a greedy algorithm gives a 1/(n+1) approximation, if we restrict the subset to n robots. In optimizing submodular functions, a common assumption is that the function value is deterministic, which may not hold in practice. For example, the sensor performance may depend on environmental conditions which are not known exactly. In the second part of the dissertation, we design an algorithm for stochastic submodular optimization. The standard formulation for stochastic optimization optimizes the expected performance. However, the expectation is a risk-neutral measure. Instead, we optimize the Conditional Value-at-Risk (CVaR), which allows the user the flexibility of choosing a risk level. We present an algorithm, based on the greedy algorithm, and prove that its performance has bounded suboptimality and improves with running time. We also present an online version of the algorithm to adapt to real-time scenarios. In the third part of this dissertation, we focus on scenarios where a set of robots may fail naturally or due to adversarial attacks. Our objective is to track as many targets as possible, a submodular measure, assuming worst-case robot failures. We present both centralized and distributed resilient tracking algorithms to cope with centralized and distributed communication settings. We prove these algorithms give a constant-factor approximation of the optimal in polynomial running time. / Doctor of Philosophy / Today, robotics and autonomous systems have been increasingly used in various areas such as manufacturing, military, agriculture, medical sciences, and environmental monitoring. However, most of these systems are fragile and vulnerable to adversarial attacks and uncertain environmental conditions. In most cases, even if a part of the system fails, the entire system performance can be significantly undermined. As robots start to coexist with humans, we need algorithms that can be trusted under real-world, not just ideal conditions. Thus, this dissertation focuses on enabling security, trustworthiness, and long-term autonomy in robotics and autonomous systems. In particular, we devise coordination algorithms that are resilient to attacks, trustworthy in the face of the uncertain conditions, and allow the long-term operation of multi-robot systems. We evaluate our algorithms through extensive simulations and proof-of-concept experiments. Generally speaking, multi-robot systems form the "physical" layer of Cyber-Physical Sytems (CPS), the Internet of Things (IoT), and Smart City. Thus, our research can find applications in the areas of connected and autonomous vehicles, intelligent transportation, communications and sensor networks, and environmental monitoring in smart cities.

L’entropie d’une distribution sur un ensemble de variables aléatoires discrètes est toujours bornée par l’entropie de la distribution factorisée correspondante. Cette propriété est due à la sous-modularité de l’entropie. Par ailleurs, les fonctions sous-modulaires sont une généralisation des fonctions de rang des matroïdes ; ainsi, les fonctions linéaires sur les polytopes associés peuvent être minimisées exactement par un algorithme glouton. Dans ce manuscrit, nous exploitons ces liens entre les structures des modèles graphiques et les fonctions sous-modulaires. Nous utilisons des algorithmes gloutons pour optimiser des fonctions linéaires sur des polytopes liés aux matroïdes graphiques et hypergraphiques pour apprendre la structure de modèles graphiques, tandis que nous utilisons des algorithmes d’inférence sur les graphes pour optimiser des fonctions sous-modulaires. La première contribution de cette thèse consiste à approcher par maximum de vraisemblance une distribution de probabilité par une distribution factorisable et de complexité algorithmique contrôlée. Comme cette complexité est exponentielle dans la largeur arborescente du graphe, notre but est d’apprendre un graphe décomposable avec une largeur arborescente bornée, ce qui est connu pour être NP-difficile. Nous posons ce problème comme un problème d’optimisation combinatoire et nous proposons une relaxation convexe basée sur les matroïdes graphiques et hypergraphiques. Ceci donne lieu à une solution approchée avec une bonne performance pratique. Pour la seconde contribution principale, nous utilisons le fait que l’entropie d’une distribution est toujours bornée par l’entropie de sa distribution factorisée associée, comme conséquence principale de la sous-modularité, permettant une généralisation à toutes les fonctions sous-modulaires de bornes basées sur les concepts de modèles graphiques. Un algorithme est développé pour maximiser les fonctions sous-modulaires, un autre problème NP-difficile, en maximisant ces bornes en utilisant des algorithmes d’inférence vibrationnels sur les graphes. En troisième contribution, nous proposons et analysons des algorithmes visant à minimiser des fonctions sous-modulaires pouvant s’écrire comme somme de fonctions plus simples. Nos algorithmes n’utilisent que des oracles de ces fonctions simple basés sur minimisation sous-modulaires et de variation totale de telle fonctions. / The entropy of a probability distribution on a set of discrete random variables is always bounded by the entropy of its factorisable counterpart. This is due to the submodularity of entropy on the set of discrete random variables. Submodular functions are also generalisation of matroid rank function; therefore, linear functions may be optimised on the associated polytopes exactly using a greedy algorithm. In this manuscript, we exploit these links between the structures of graphical models and submodular functions: we use greedy algorithms to optimise linear functions on the polytopes related to graphic and hypergraphic matroids for learning the structures of graphical models, while we use inference algorithms on graphs to optimise submodular functions.The first main contribution of the thesis aims at approximating a probabilistic distribution with a factorisable tractable distribution under the maximum likelihood framework. Since the tractability of exact inference is exponential in the treewidth of the decomposable graph, our goal is to learn bounded treewidth decomposable graphs, which is known to be NP-hard. We pose this as a combinatorial optimisation problem and provide convex relaxations based on graphic and hypergraphic matroids. This leads to an approximate solution with good empirical performance. In the second main contribution, we use the fact that the entropy of a probability distribution is always bounded by the entropy of its factorisable counterpart mainly as a consequence of submodularity. This property of entropy is generalised to all submodular functions and bounds based on graphical models are proposed. We refer to them as graph-based bounds. An algorithm is developped to maximise submodular functions, which is NPhard, by maximising the graph-based bound using variational inference algorithms on graphs. As third contribution, we propose and analyse algorithms aiming at minimizing submodular functions that can be written as sum of simple functions. Our algorithms only make use of submodular function minimisation and total variation oracles of simple functions.

Recently, Udwani defined a new class of set functions under monotonicity and subadditivity, called submodular order functions, which is a subfamily of submodular functions. Informally, the submodular order function admits a very limited form of submodularity which is defined over a specific permutation of the ground set. His work pointed out the intriguing connection between streaming submodular maximization and submodular order maximization. Inspired by a 0.25-approximation streaming algorithm for maximizing a monotone submodular function subject to a matroid constraint, Udwani gave a 0.25-approximation algorithm for submodular order functions maximization subject to a matroid constraint. Based on the above results, we would like to explore further in which cases it is feasible to generalize from streaming submodular maximization algorithms to submodular order maximization algorithms. As a more general constraint than matroid, p-matchoid is a collection of p matroids with each matroid defined on some subsets of the ground set. Related work gave a 1/4p-approximation streaming algorithm for monotone submodular functions maximization under a p-matchoid constraint. Inspired by the above algorithms and the intriguing connection, we used some techniques to try to generalize several streaming algorithms for submodular functions to the offline algorithms for submodular order functions, including interleaved partitions and incremental values. Assuming that the objective function f is subadditive and non-negative, we gave a 1/4p-approximation algorithm for monotone submodular order maximization to a p-matchoid constraint. In addition, we summarize the failures of other cases. / Nyligen definierade Udwani en ny klass av mängdfunktioner under monotonicitet och subadditivitet, som kallas submodulära ordningsfunktioner och som är en underfamilj av submodulära funktioner. Informellt sett medger den submodulära ordningsfunktionen en mycket begränsad form av submodularitet som är definierad över en specifik permutation av grundmängden. Hans arbete pekade på det spännande sambandet mellan strömmande submodulär maximering och submodulär ordermaximering. Inspirerad av en strömningsalgoritm med 0.25-approximation för maximering av en monoton submodulär funktion som är föremål för en matroidbegränsning, gav Udwani en algoritm med 0.25-approximation för maximering av submodulära ordningsfunktioner som är föremål för en matroidbegränsning. Baserat på ovanstående resultat skulle vi vilja utforska ytterligare i vilka fall det är möjligt att generalisera från algoritmer för strömning av submodulära maximeringsfunktioner till algoritmer för maximering av submodulära orderfunktioner. Som en mer allmän begränsning än matroid är p-matchoid en samling av p matroider där varje matroid definieras på vissa delmängder av grundmängden. Relaterade arbeten gav en strömmingsalgoritm med 1/4p-tillnärmning för monoton submodulär funktionsmaximering under en p-matchoid-begränsning. Inspirerade av ovanstående algoritmer och det spännande sambandet använde vi vissa tekniker för att försöka generalisera flera strömningsalgoritmer för submodulära funktioner till offline-algoritmer för submodulära ordningsfunktioner, inklusive interleaved partitions och inkrementella värden. Under förutsättning att målfunktionen f är subadditiv och icke-negativ gav vi en algoritm för 1/4p-tillnärmning för monoton submodulär ordermaximering till ett p-matchoid-begränsningsvillkor. Dessutom sammanfattar vi misslyckanden i andra fall.

京都大学 / 0048 / 新制・課程博士 / 博士(情報学) / 甲第22588号 / 情博第725号 / 新制||情||124(附属図書館) / 京都大学大学院情報学研究科通信情報システム専攻 / (主査)教授 湊 真一, 教授 五十嵐 淳, 教授 山本 章博 / 学位規則第4条第1項該当 / Doctor of Informatics / Kyoto University / DFAM