Rozprawy doktorskie na temat „Network planning”

Renewable energy generation is becoming a major part of energy supply, often in the form of distributed generation (DG) connected to distribution networks. While growth has been rapid, there is awareness that limitations on spare capacity within distribution (and transmission) networks is holding back development. Developments are being shelved until new network reinforcements can be built, which may make some projects non-viable. Reinforcements are costly and often underutilised, typically only loaded to their limits for a few occasions during the year. In order to accommodate new DG without the high costs or delays, active network management (ANM) is being promoted in which generation and other network assets are controlled within the limits of the existing network. There is a great deal of complexity and uncertainty associated with developing ANM and devising coherent plans to accommodate new DG is challenging for Distribution Network Operators (DNOs). As such, there is a need for robust network planning tools that can explicitly handle ANM and which can be trusted and implemented easily. This thesis describes the need for and the development of a new distribution expansion planning framework that provides DNOs with a better understanding of the impacts created by renewable DG and the value of ANM. This revolves around a heuristic planning framework which schedules necessary upgrades in power lines and transformers associated with changes in demand as well as those driven by the connection of DG. Within this framework a form of decentralised, adaptive control of DG output has been introduced to allow estimation of the impact of managing voltage and power flow constraints on the timing and need for network upgrades. The framework is initially deployed using simple scenarios but a further advance is the explicit use of time series to provide substantially improved estimates of the levels of curtailment implied by ANM. In addition, a simplified approach to incorporating demand side management has been deployed to facilitate understanding of the scope and role this may play in facilitating DG connections.

Este trabalho apresenta uma metodologia e correspondente programa computacional para o planejamento da expansão de uma rede distribuição que permite determinar, por exemplo, a localização de banco de capacitores e de reguladores de tensão e/ou auto-boosters, o ganho de tensão e redução de perdas elétricas com a troca de condutores. A metodologia proposta é baseada na solução do problema de fluxo de potência desenvolvido especificamente para a distribuição. O trabalho automatiza procedimentos rotineiros do planejamento dos sistemas de distribuição, aumentando sua eficiência na análise correta de alternativas, auxiliando o planejador na tomada de decisões para determinar a expansão dos sistemas de distribuição. Testes com sistema da COELBA são apresentados.
This work presentes a methodology and a digital computer program for the expansion planning of a distribuition network that allows to determine, for exemple, the location of capacitors bank and voltage regulators and/or automatic-boosters, voltage gain and redution of electrical losses with the changing of conductors. The proposed methodology is based on the solution of the power flow problem developd espcifically to distribution systems. The work automatizes common procedures for distribuition systems planing, increasing its efficiency in choosing correct alternatives, aidding the planer to make decisions to determine the distribution systems expansion. Several tests with COELBA system s are presented.

This dissertation addresses the development and analysis of optimization models for evacuation planning. Specifically we consider the cases of large-scale regional evacuation using household vehicles and hospital evacuation. Since it is difficult to estimate the exact number of people evacuating, we first consider the case where the population size is uncertain. We review the methods studied in the literature, mainly the strategy of using a deterministic counterpart, i.e., a single deterministic parameter to represent the uncertain population, and we show that these methods are not very effective in generating a good traffic management strategy. We provide alternatives, where we describe some networks where an optimal policy exist independent of the demand realization and we propose some simple heuristics for more complex ones. Next we consider the traffic management tools that can be generated from an evacuation plan. We start by introducing the cell transmission model with flow reduction. This model captures the flow reduction after the onset of congestion. We then discuss the management tools that can be extracted from this model. We also propose some simplification to the model formulation to enhance its tractability. A heuristic for generating a solution is also proposed, and its solution quality is analyzed. Finally, we discuss the hospital evacuation problem where we develop an integer programming model that integrates the building evacuation with the transportation of patients. The impact of building evacuation capabilities on the transportation plan is investigated through the case of a large regional hospital case study. We also propose a decomposition scheme to improve the tractability of the integer program.
Ph. D.

This thesis discusses an algorithm and two heuristics for solving a particular network optimization problem: The node-disjoint paths problem. The goal of this optimization problem is to find two node-disjoint paths between a given origin-destination pair whose total cost is minimum. This problem is shown to be NP-Hard. Two heuristics are investigated in this thesis. The sequential shortest paths heuristic, is the faster of the two methods, but the quality of the solution may be sacrificed. On the other hand, the simultaneous shortest paths heuristic, which yields very good solutions, has higher complexity. We also discuss an implicit enumeration algorithm that is used to verify the quality of the solution obtained from the heuristics.

Thesis (M.S. in Information Technology Management) Naval Postgraduate School, September 1996.
Thesis advisor(s): S. Sridhar, H.K. Bhargava. "September 1996." Includes bibliographical references (p. 125). Also available online.

Next-generation internet services such as live TV and video on demand require high bandwidth and ultra-low latency. The ever-increasing volume, dynamicity and stringent requirements of these services’ demands are generating new challenges to nowadays telecom networks. To decrease expenses, service-layer content providers are delivering their content near the end users, thus allowing a low latency and tailored content delivery. As a consequence of this, unseen metro and even core traffic dynamicity is arising with changes in the volume and direction of the traffic along the day. A tremendous effort to efficiently manage networks is currently ongoing towards the realisation of 5G networks. This translates in looking for network architectures supporting dynamic resource allocation, fulfilling strict service requirements and minimising the total cost of ownership (TCO). In this regard, in-operation network planning was recently proven to successfully support various network reconfiguration use cases in prospective scenarios. Nevertheless, additional research to extend in-operation planning capabilities from typical reactive optimization schemes to proactive and predictive schemes based on the analysis of network monitoring data is required. A hot topic raising increasing attention is cognitive networking, where an elevated knowledge about the network could be obtained as a result of introducing data analytics in the telecom operator’s infrastructure. By using predictive knowledge about the network traffic, in-operation network planning mechanisms could be enhanced to efficiently adapt the network by means of future traffic prediction, thus achieving cognitive in-operation network planning. In this thesis, we focus on studying mechanisms to enable cognitive in-operation network planning in core networks. In particular, we focus on dynamically reconfiguring virtual network topologies (VNT) at the MPLS layer, covering a number of detailed objectives. First, we start studying mechanisms to allow network traffic flow modelling, from monitoring and data transformation to the estimation of predictive traffic model based on this data. By means of these traffic models, then we tackle a cognitive approach to periodically adapt the core VNT to current and future traffic, using predicted traffic matrices based on origin-destination (OD) predictive models. This optimization approach, named VENTURE, is efficiently solved using dedicated heuristic algorithms and its feasibility is demonstrated in an experimental in-operation network planning environment. Finally, we extend VENTURE to consider core flows dynamicity as a result of metro flows re-routing, which represents a meaningful dynamic traffic scenario. This extension, which entails enhancements to coordinate metro and core network controllers with the aim of allowing fast adaption of core OD traffic models, is evaluated and validated in terms of traffic models accuracy and experimental feasibility.
Els serveis d’internet de nova generació tals com la televisió en viu o el vídeo sota demanda requereixen d’un gran ample de banda i d’ultra-baixa latència. L’increment continu del volum, dinamicitat i requeriments d’aquests serveis està generant nous reptes pels teleoperadors de xarxa. Per reduir costs, els proveïdors de contingut estan disposant aquests més a prop dels usuaris finals, aconseguint així una entrega de contingut feta a mida. Conseqüentment, estem presenciant una dinamicitat mai vista en el tràfic de xarxes de metro amb canvis en la direcció i el volum del tràfic al llarg del dia. Actualment, s’està duent a terme un gran esforç cap a la realització de xarxes 5G. Aquest esforç es tradueix en cercar noves arquitectures de xarxa que suportin l’assignació dinàmica de recursos, complint requeriments de servei estrictes i minimitzant el cost total de la propietat. En aquest sentit, recentment s’ha demostrat com l’aplicació de “in-operation network planning” permet exitosament suportar diversos casos d’ús de reconfiguració de xarxa en escenaris prospectius. No obstant, és necessari dur a terme més recerca per tal d’estendre “in-operation network planning” des d’un esquema reactiu d’optimització cap a un nou esquema proactiu basat en l’analítica de dades provinents del monitoritzat de la xarxa. El concepte de xarxes cognitives es també troba al centre d’atenció, on un elevat coneixement de la xarxa s’obtindria com a resultat d’introduir analítica de dades en la infraestructura del teleoperador. Mitjançant un coneixement predictiu sobre el tràfic de xarxa, els mecanismes de in-operation network planning es podrien millorar per adaptar la xarxa eficientment basant-se en predicció de tràfic, assolint així el que anomenem com a “cognitive in-operation network Planning”. En aquesta tesi ens centrem en l’estudi de mecanismes que permetin establir “el cognitive in-operation network Planning” en xarxes de core. En particular, ens centrem en reconfigurar dinàmicament topologies de xarxa virtual (VNT) a la capa MPLS, cobrint una sèrie d’objectius detallats. Primer comencem estudiant mecanismes pel modelat de fluxos de tràfic de xarxa, des del seu monitoritzat i transformació fins a l’estimació de models predictius de tràfic. Posteriorment, i mitjançant aquests models predictius, tractem un esquema cognitiu per adaptar periòdicament la VNT utilitzant matrius de tràfic basades en predicció de parells origen-destí (OD). Aquesta optimització, anomenada VENTURE, és resolta eficientment fent servir heurístiques dedicades i és posteriorment avaluada sota escenaris de tràfic de xarxa dinàmics. A continuació, estenem VENTURE considerant la dinamicitat dels fluxos de tràfic de xarxes de metro, el qual representa un escenari rellevant de dinamicitat de tràfic. Aquesta extensió involucra millores per coordinar els operadors de metro i core amb l’objectiu d’aconseguir una ràpida adaptació de models de tràfic OD. Finalment, proposem dues arquitectures de xarxa necessàries per aplicar els mecanismes anteriors en entorns experimentals, emprant protocols estat-de-l’art com són OpenFlow i IPFIX. La metodologia emprada per avaluar el treball anterior consisteix en una primera avaluació numèrica fent servir un simulador de xarxes íntegrament dissenyat i desenvolupat per a aquesta tesi. Després d’aquesta validació basada en simulació, la factibilitat experimental de les arquitectures de xarxa proposades és avaluada en un entorn de proves distribuït.

In this project, coverage planning in GSM networks as well as capacity and frequency planning has been studied. Various signal interruptions and the necessary steps to remove those interruptions in order to maintain signal quality in mobile communication have been studied. Precautions that should be taken for reducing the effects of interruptions have also been discussed. A drive test has been performed as a part of the improvement process. Guidelines for key performance indicators (KPI) pave the way for radio network quality, coverage and the smooth functioning of the GSM system.

Thesis (M.Eng.)--Massachusetts Institute of Technology, Dept. of Electrical Engineering and Computer Science, 2002.
Includes bibliographical references (leaves 93-96).
Attack scenario graphs provide a concise way of displaying all possible sequences of attacks a malicious user can execute to obtain a desired goal, such as remotely achieving root undetected on a critical host machine. NETSPA, the Network Security Planning Architecture, is a C++ system that quickly generates worst-case attack graphs using a forward-chaining depth-first search of the possible attack space using actions modeled with REM, a simple attack description language. NETSPA accepts network configuration information from a database that includes host and network software types and versions, intrusion detection system placement and types, network connectivity, and firewall rulesets. It is controlled by command line inputs that determine a critical goal state, trust relationships between hosts, and maximum recursive depth. NETSPA was shown to efficiently provide easily understood attack graphs that revealed non-obvious security problems against a realistic sample network of 17 representative hosts using 23 REM defined actions. The largest useful graph was generated within 1.5 minutes of execution. NETSPA-executes faster and handles larger networks than any existing graph generation system. This allows NETSPA to be practically used in combination with other security components to develop and analyze secure networks.
by Michael Lyle Artz.
M.Eng.

The rapid expansion of mobile services and the emerging demand for multimedia applications have led to an impressive traffic growth. To this end, Mobile Network Operators (MNOs) seek to extend their infrastructure by installing more Base Stations (BSs), in an effort to increase the network capacity and meet the pressing traffic demands. Furthermore, to fulfill the escalated demands, Heterogeneous Networks (HetNets), which consist of Small Cells (SCs) and the traditional BSs, constitute the new trend of next generation networks. The deployed infrastructure implies a rise in the Capital Expenditures and has a direct impact on the network energy consumption, thus resulting in higher Operational Expenditures. Hence, the investigation of energy efficient solutions will bring down the energy consumption and the network cost. Since the BS is the most power hungry component, the research community has shifted towards the investigation of BS deactivation schemes. These schemes propose that part of the infrastructure can be temporarily switched off, when the traffic is low, while the active BSs extend their coverage to serve the network. Based on a comprehensive review of the state-of-the-art, a set of research opportunities were identified. This thesis provides contributions to the field of BS switching off strategies for wireless macro BSs networks and HetNets of single and multiple MNOs by proposing mechanisms that enhance different aspects of the network performance. The BSs deactivation, the innovative trend of infrastructure sharing and the financially driven collaboration among the involved parties of the current and future networks promise significant improvements in terms of energy and cost savings. The main thesis contributions are divided into three parts, described next. The first part of the thesis introduces innovative BS switching off approaches in single-operator environments, where only macro BSs are deployed. The proposed strategies exploit the inherit characteristics of the traffic load pattern (e.g., distribution of the users, traffic volume, etc.) and the distinctive features of the wireless cellular networks (e.g., BSs position, topology, etc.). Theoretical analysis and computer-based simulations show the performance improvement offered by the switching off strategies with respect to energy efficiency. The second part of the thesis explores a different challenge in network planning. The coexistence of multiple MNOs in the same geographical area has motivated a new business model, known as infrastructure sharing. A roaming-based deactivation scheme is proposed, by taking into account the rationality and the conflicting interests of the MNOs. The proposed game theoretic framework enables the MNOs to take individual switching off decisions, thus bypassing potential complicated agreements. The theoretical and simulation results show that our proposal significantly improves the energy efficiency, guaranteeing at the same time the throughput in realistic scenarios. Moreover, the proposed scheme provides higher cost efficiency and fairness compared to the state-of-the-art algorithms, motivating the MNOs to adopt game theoretic strategies. The third part of the thesis focuses on the exploitation of HetNets and the proposal of energy and cost effective strategies in SC networks with multiple MNOs. We effectively address the cost sharing by proposing accurate cost models for the SCs to share the network cost. Taking into account the impact of the traffic on the cost, we propose novel cost sharing policies that provide a fair outcome. In continuation, innovative auction-based schemes within multiobjective optimization framework are introduced for data offloading from the BSs, owned by the MNOs, to the third-party SC networks. The proposed solution captures the conflicting interests of the MNOs and the third-party companies and the obtained results show that the benefit of proposing switching off approaches for HetNets.
La rápida expansión de los servicios móviles y de la demanda emergente de aplicaciones multimedia han dado lugar a un impresionante crecimiento del tráfico. Operadores de redes móviles (MNOs) tratará de extender su infraestructura mediante la instalación de más estaciones base (BSs), en un esfuerzo por aumentar la capacidad de la red y satisfacer las apremiantes demandas de tráfico. Además, para cumplir con las exigencias escalada, redes heterogéneas (HetNets), constituyen la nueva tendencia de las redes de próxima generación. La infraestructura implica un aumento en los gastos de capital y tiene un impacto directo en el consumo de energía a la red, lo que resulta en un aumento de los gastos operacionales. La investigación de soluciones de eficiencia energética hará bajar el consumo de energía y el coste de la red. La comunidad científica se ha desplazado hacia la investigación de los sistemas de desactivación de BSs. Estos esquemas proponen que parte de la infraestructura se puede cdesectivarse, cuando el tráfico es bajo, mientras que los BSs activas extender su cobertura al servicio de la red. Esta tesis ofrece contribuciones al campo de la BSs desconexión para las redes y HetNets con uno o multiples MNOs, proponiendo mecanismos que mejoran diferentes aspectos del rendimiento de la red. La desactivación de BS,s la tendencia innovadora de compartir infraestructura y la colaboración impulsada financieramente entre las partes implicadas de las redes actuales y futuras prometen mejoras significativas en términos de ahorro energético y económico. Las principales contribuciones de tesis se dividen en tres partes, que se describen a continuación. La primera parte de la tesis introduce innovadora apagar enfoques en entornos de un solo operador, donde se despliegan sólo macro BSs. Las estrategias propuestas explotan las características de tráfico (por ejemplo, la distribución de los usuarios, el volumen de tráfico, etc.) y las características distintivas de las redes (por ejemplo, la posición BS, topología, etc.). Análisis teórico y simulaciones muestran la mejora del rendimiento ofrecido por las estrategias de conmutación con respecto a la eficiencia energética. La segunda parte de la tesis explora un reto diferente en la planificación de la red. La coexistencia de múltiples operadores en la misma zona geográfica ha motivado un nuevo modelo de negocio, conocida como la compartición de infraestructura. Se propone un esquema de desactivación basada en itinerancia, teniendo en cuenta la racionalidad y los intereses en conflicto de los operadores de redes móviles. Los resultados teóricos y de simulación muestran que nuestra propuesta mejora significativamente la eficiencia energética, garantizando al mismo tiempo el rendimiento en escenarios realistas. Por otra parte, el esquema propuesto proporciona una mayor eficiencia de costes y la equidad en comparación con los algoritmos del estado de la técnica, motivar al mnos de adoptar estrategias de teoría de juegos. La tercera parte de la tesis se centra en la explotación de HetNets y la propuesta de estrategias eficaces de energía y costes en las redes con múltiples operadores. Nos dirigimos efectivamente la participación en los costos, proponiendo modelos de costos precisos para para compartir el costo de la red. Teniendo en cuenta el impacto del tráfico en el coste, proponemos políticas costos compartidos novedosas que proporcionan un resultado justo. En la continuación, los esquemas basados en subastas innovadoras dentro de marco de optimización multiobjetivo se introducen los datos que descargan de la BS. La solución propuesta recoge los intereses en conflicto de los operadores de redes móviles y las compañías de terceros y los resultados obtenidos muestran que el beneficio de proponer la desconexión se acerca para HetNets

This paper explores how an all pair shortest path can be obtained in a wireless sensor network when sensors fail. Sensors are randomly deployed in a predefined geographical area, simulating the deployment of sensors from an airplane, and finding shortest path between all the sensors deployed based on distance. A major problem to address in wireless sensor networks is the impact of sensor failures on existing shortest paths in the network. An application is developed to simulate a network and find shortest paths affected by a sensor failure and find alternative shortest path. When a sensor fails, all the shortest paths and all the remaining sensors in the network are checked to see if the sensor failure has any impact on the network. Alternative shortest path is calculated for those paths affected by sensor failures.

One of EU’s actions against climate change is to meet 20% of our energy needs from renewable resources. Given that the renewable resources are becoming more economical to extract electricity from, this will result in that more and more distributed generation (DG) will be connected to power distribution. The increasing share of DG in the electricity networks implies both increased costs and benefits for distribution system operators (DSOs), customers and DG producers. How the costs and benefits will be allocated among the actors will depend on the established regulation. Distribution networks are traditionally not designed to accommodate generation. Hence, increasing DG penetration is causing profound changes for DSOs in planning, operation and maintenance of distribution networks. Due to the unbundling between DSOs and electricity production, DSOs can not determine either the location or the size of DG. This new power distribution environment brings new challenges for the DSOs and the electric power system regulator. The DSOs are obliged to enable connection of DG meanwhile fulfilling requirements on power quality and adequate reliability. Moreover, regulatory implications can make potential DG less attractive. Therefore regulation should be able to send out incentives for the DSOs to efficiently plan the network to accommodate the increasing levels of DG. To analyze the effects of regulatory polices on network investments, risk analysis methods for integrating the DG considering uncertainties are therefore needed. In this work, regulation impact on network planning methods and network tariff designs in unbundled electricity network is firstly analyzed in order to formulate a realistic long-term network planning model considering DG. Photovoltaic (PV) power and wind power plants are used to demonstrate DG. Secondly, this work develops a deterministic model for low-voltage (LV) networks mainly considering PV connections which is based on the worst-case scenario. Dimension the network using worst-case scenario is the convention in the long-term electricity distribution network planning for the reliability and security reason. This model is then further developed into a probabilistic model in order to consider the uncertainties from DG production and load. Therefore more realistic operation conditions are considered and probabilistic constrains on voltage variation can be applied. Thirdly, this work develops a distribution medium-voltage (MV) network planning model considering wind power plant connections. The model obtains the optimal network expansion and reinforcement plan of the target network considering the uncertainties from DG production and load. The model is flexible to modify the constraints. The technical constraints are respected in any scenario and violated in few scenarios are implemented into the model separately. In LV networks only PV connections are demonstrated and in MV networks only wind power connections are demonstrated. The planning model for LV networks is proposed as a practical guideline for PV connections. It has been shown that it is simple to be implemented and flexible to adjust the planning constraints. The proposed planning model for MV networks takes reinforcement on existing lines, new connection lines to DG, alternatives for conductor sizes and substation upgrade into account, and considers non-linear power flow constraints as an iterative linear optimization process. The planning model applies conservative limits and probabilistic limits for increasing utilization of the network, and the different results are compared in case studies. The model’s efficiency, flexibility and accuracy in long-term distribution network planning problems are shown in the case studies.

QC 20140217

Elforsk Risknanlys II

Technologies like IEEE 802.16j wireless mesh networks are drawing increasing attention of the research community. Mesh networks are economically viable and may extend services such as Internet to remote locations. This thesis takes interest into a planning problem in IEEE 802.16j networks, where we need to establish minimum cost relay and base stations to cover the bandwidth demand of wireless clients. A special feature of this planning problem is that any node in this network can send data to at most one node towards the next hop, thus traffic flow is unsplittable from source to destination. We study different integer programming formulations of the problem. We propose four types of heuristic algorithms that uses greedy, local search, variable neighborhood search and Lagrangian relaxation based approaches for the problem. We evaluate the algorithms on database of network instances of 500-5000 nodes, some of which are randomly generated network instances, while other network instances are generated over geometric distribution. Our experiments show that the proposed algorithms produce satisfactory result compared to benchmarks produced by generalized optimization problem solver software.
x, 131 leaves : ill. ; 29 cm

With the increasing demand for high speed wireless network services, the next generation wireless networks are proposed to use advanced wireless communication technologies. These technologies include massive MIMO, mmWave and distributed MIMO. In order to deploy wireless networks equipped with these technologies, channel models capturing the channel features and characteristics of these wireless technologies are essential in the planning and optimisation of networks. However, conventional channel models lack the capability to support these next generation network technologies. In this PhD thesis, I investigated the channel models for the next generation wireless technologies, including massive MIMO, mmWave communications and distributed MIMO. I developed channel models for network planning and optimisation based on conventional ray launching algorithms for these wireless technologies. The models have been validated and applied to optimise network performance. The existing challenge in wireless channel modelling is the improvement of modelling accuracy without increasing modelling complexity. In order to achieve this goal, a new calibration method is developed to improve the accuracy of the predication model when measurements are available. Moreover, in order to use the channel models as an effective tool in wireless network planning and optimisation, a new wireless capacity definition from radio propagation perspective is also investigated. It provides insight to the physical limit of wireless channel capacity from a radio propagation perspective.

Au cours de ces dernières années, la quantité de trafic que les utilisateurs Internet produisent sur une base quotidienne a augmenté de façon exponentielle, principalement en raison du succès des services de streaming vidéo, tels que Netflix et YouTube. Alors que les réseaux de diffusion de contenu (Content-Delivery Networks, CDN) sont la technique standard utilisée actuellement pour servir les demandes des utilisateurs, la communauté scientifique a formulé des propositions connues sous le nom de Content-Centric Networks (CCN) pour changer la pile de protocoles réseau afin de transformer Internet en une infrastructure de distribution de contenu. Dans ce contexte, cette thèse de doctorat étudie des techniques efficaces pour la distribution de contenu numérique en tenant compte de trois problèmes complémentaires : 1) Nous considérons le scénario d’un réseau hétérogène sans fil, et nous formulons un mécanisme pour motiver les propriétaires des points d’accès à partager leur capacité WiFi et stockage cache inutilisés, en échange d’une contribution économique.2) Nous étudions le problème centralisé de planification du réseau en présence de caches distribuées et (I) nous analysons la migration optimale du réseau à CCN; (II) nous comparons les bornes de performance d’un réseau CDN avec ceux d’un CCN, et (III) nous considérons un réseau CDN virtualisé et étudions le problème stochastique de planification d’une telle infrastructure.3) Nous considérons les implications de sécurité sur le contrôle d’accès et la traçabilité, et nous formulons ConfTrack-CCN, une extension deCCN utilisée pour garantir la confidentialité, traçabilité et l’évolution de la politique d’accès, en présence de caches distribuées
In recent years, the amount of traffic requests that Internet users generate on a daily basis has increased exponentially, mostly due to the worldwide success of video streaming services, such as Netflix and YouTube. While Content-Delivery Networks (CDNs) are the de-facto standard used nowadays to serve the ever increasing users’ demands, the scientific community has formulated proposals known under the name of Content-Centric Networks (CCN) to change the network protocol stack in order to turn the network into a content distribution infrastructure. In this context this Ph.D. thesis studies efficient techniques to foster content distribution taking into account three complementary problems:1) We consider the scenario of a wireless heterogeneous network, and we formulate a novel mechanism to motivate wireless access point owners to lease their unexploited bandwidth and cache storage, in exchange for an economic incentive.2) We study the centralized network planning problem and (I) we analyze the migration to CCN; (II) we compare the performance bounds for a CDN with those of a CCN, and (III) we take into account a virtualized CDN and study the stochastic planning problem for one such architecture.3) We investigate the security properties on access control and trackability and formulate ConfTrack-CCN: a CCN extension to enforce confidentiality, trackability and access policy evolution in the presence of distributed caches

The ever demanding bandwidth requirements for supporting emerging telecom services such as ultra-high-definition video streaming, cloud computing, connected car, virtual/augmented reality, etc., bring to the fore the necessity to upgrade continuously the technology behind transport networks in order to keep pace with this exponential traffic growth. Thus, everything seems to indicate that fixed-grid Wavelength-Division Multiplexed (WDM) networks will be upgraded by adopting a flexible-grid, thus providing finer bandwidth allocation granularities, and therefore, increasing the Grade-of-Service by packing more information in the same spectral band of standard Single-Mode Fibers (SMFs). Nevertheless, unfortunately, the fundamental Shannon’s limit of SMFs is rapidly approaching, and, then, the research efforts to increase the SMFs' capacity will be useless. One solution to overcome this capacity crunch effect is to enable one extra dimension in addition to the frequency one, namely, the spatial dimension, thus deploying S parallel paths in order to multiply, in the best case, by S the capacity of SMF-based networks. However, additionally, it is necessary to decrease the cost and energy per bit in order to provide economically attractive solutions. For this purpose, a smooth upgrade path has to be carried out as new integrated devices and system components are developed for Space Division Multiplexing (SDM). This thesis is concentrated on the planning and operation of the combined flexible WDM and SDM networks (i.e., Flex-Grid/SDM networks) proposing several strategies aimed at optimizing network resources usage with hardware complexity analysis. For this purpose, firstly, network problems are carefully studied and stated, and then, mathematical and/or heuristic algorithms are designed and implemented in an optical network simulator. Specifically, after an introduction to the thesis, chapter 2 presents the background and related work. Next, chapter 3 concentrates on the study of spatially fixed Flex-Grid/SDM networks, i.e., when a rigid number of spatial channels are reserved per allocated traffic demand. In its turn, chapter 4 studies the case of Spectrally-Spatially Flexible Optical Networks (SS-FONs), as the ones providing the upper-bound network capacity. Costs and hardware requirements implied on providing this flexibility are analyzed. Network nodes aimed at reducing the cost of SS-FONs are presented and evaluated in chapter 5. Finally, this thesis ends with the presentation of the main contributions and future research work in chapter 6.
La demanda de ancho de banda cada vez más exigente para soportar servicios de telecomunicación emergentes tales como la transmisión de video de alta calidad, computación en la nube, vehículo conectado, realidad virtual/aumentada, etc.…, ha puesto de manifiesto la necesidad de actualizar constantemente la tecnología detrás de las redes de transporte óptico con la finalidad de ir a la par de este incremento exponencial del tráfico. De esta manera, todo parece indicar que las redes basadas en la multiplexación por division de longitud de onda (Wavelength Division Multiplexing, WDM) de ancho espectral fijo serán actualizadas adoptando un ancho de banda espectral flexible, que ofrece asignaciones de ancho de banda con granularidad más fina acorde a las demandas de tráfico; y por lo tanto, incremanta el Grado de Servicio de la red, ya que se permite acomodar mayor información en la misma banda espectral de las fibras monomodo (Single Mode Fibers, SMFs). Sin embargo, desafortunadamente, el límite de Shannon de las fibras monomodo se está aproximando cada vez más, y cuando esto ocurra las investigaciones para incrementar la capacidad de las fibras monomodo serán infructuosas. Una posible solución para superar este colapso de las fibras monomodo es habilitar la dimensión espacial a más de la frecuencial, desplegando 𝑆� caminos paralelos con la finalidad de multiplicar por 𝑆� (en el mejor de los casos) la capacidad de las fibras monomodo. No obstante, es necesario disminuir el costo y la energía por bit con la finalidad de proveer soluciones comerciales atractivas. Para tal propósito debe llevarse a cabo una actualización moderada conforme nuevos dispositivos y componentes integrados son desarrollados para la implementación de la tecnología basada en la multiplexación por división de espacio (Space Division Multiplexing, SDM). Esta tesis se concentra en la planificación y operación de la combinación de las redes WDM flexibles y SDM (es decir, de las redes Flex-Grid/SDM) proponiendo varias estrategias dirigidas a optimizar el uso de los recursos de red junto con el análisis de la complejidad del hardware que viene acompañada. Para este fin, primeramente, los problemas de red son cuidadosamente estudiados y descritos. A continuación, se han diseñado e implementado algoritmos basados en programación lineal entera o heurísticas en un simulador de redes ópticas. Después de una introducción inicial, el capítulo 2 de esta tesis presenta el marco teórico sobre los conceptos tratados y los trabajos publicados anteriormente. A continuación, el capítulo 3 se concentra en el estudio de las redes Flex-Grid/SDM con la dimensión espacial rígida; es decir, cuando un número fijo de canales espaciales son reservados por cada demanda de tráfico establecida. Por su parte, el capítulo 4 estudia las redes Flex-Grid/SDM considerando flexibilidad tanto en el dominio espacial como espectral (Spectrally and Spatially Flexible Optical Networks, SS-FONs), las cuales proveerían la capacidad máxima de las redes SDM. Adicionalmente, los costos y requerimientos de hardware implicados en la provisión de esta flexibilidad son analizados. El capítulo 5 presenta la evaluación de nodos orientados a reducir los costos de las SS-FONs. Finalmente, el capítulo 6 expone las principales contribuciones y las posibles líneas de trabajo futuro
Els requisits incessants d’ample de banda per al suport de nous serveis de telecomunicació, com poden ser la difusió en directe de vídeo de molt alta definició, la informàtica en el núvol, els cotxes intel·ligents connectats a la xarxa, la realitat virtual/augmentada, etc…, han exigit una millora contínua de les tecnologíes de les actuals xarxes de transport de dades. Tot sembla indicar que les xarxes de transport òptiques actuals, basades en la tecnologia de multiplexació per divisió de longitud d’ona (Wavelength Division Multiplexing, WDM) sobre un grid espectral rígid, hauran de ser reemplaçades per tecnologies òptiques més flexibles, amb una granularitat més fina a l’hora de suportar noves connexions, incrementat el grau de servei de les xarxes gràcies a aprofitament major de l’ample de banda espectral proporcionat per les fibres òptiques monomode (Single Mode Fibers, SMFs). Tanmateix, estem exhaurint ja la capacitat màxima de les fibres òptiques SMF segons ens indica el límit fonamental de Shannon. Per tant, qualsevol esforç enfocat a millorar la capacitat d’aquestes xarxes basades en SMFs pot acabar sent infructuós. Una possible solució per superar aquestes limitacions de capacitat és explorar la dimensió espacial, a més de l’espectral, desplegant 𝑆 camins en paral·lel per tal de multiplicar per 𝑆, en el millor cas, la capacitat de les SMFs. Tot i això, és necessari reduir el cost i el consum energètic per bit transmès, per tal de proporcionar solucions econòmicament viables. Amb aquest propòsit, pot ser necessària una migració progressiva, a mesura que es desenvolupen nous dispositius i components per aquesta nova tecnologia de multiplexació per divisió espacial (Spatial Division Multiplexing, SDM). La present tesi es centra en la planificació i operació de xarxes òptiques de nova generació que combinin tecnologies de xarxa WDM flexible i SDM (és a dir, xarxes Flex-Grid/SDM), proposant estratègies per a l’optimització de l’ús dels recursos de xarxa i, en definitiva, el seu cost (CapEx). Amb aquest propòsit, s’analitzen en primer moment els problemes adreçats. Tot seguit, es dissenyen algorismes per tal de solucionar-los, basats en tècniques de programació matemàtica i heurístiques, els quals s’implementen i es proven en un simulador de xarxa òptica. Després d’una introducció inicial, el capítol 2 d’aquesta tesi presenta tots els conceptes tractats i treballs relacionats publicats amb anterioritat. Tot seguit, el capítol 3 es centra en l’estudi de les xarxes Flex-Grid/SDM fixes en el domini espai, és a dir, on sempre es reserva un nombre rígid de canals espacials per qualsevol demanda suportada. El capítol 4 estudia les xarxes flexibles en els dominis espectrals i espacials (Spectrally-Spatially Flexible Optical Nextworks, SS-FONs), com aquelles que poden proporcionar una capacitat de xarxa màxima. En aquest context, s’analitzen els requeriments en termes de cost i hardware per tal de proporcionar aquesta flexibilitat. Llavors, en el capítol 6 es presenten opcions de node de xarxa capaces de reduir els costos de les xarxes SS-FONs. Finalment, en el capítol 7 es repassen totes les contribucions de la tesi, així com posibles línies de treball futur

The widespread introduction of Local Area Network (LAN) media has had profound implications for communications protocols. It is required that the campus network at Hatfield, which since 1981 has been based on the cambridge Ring, should take advantage of the properties of their LAN, should meet the demand of both new and traditional applications and should allow network interconnections. To show how these requirements might be met a review is given of the communications support provided elsewhere by data transportation protocols. Expansion is also required, but it may not be achieved unless appropriate planning decisions are made. Measurements which provide knowledge of typical traffic characteristics and quantities and of constraints or erroneous behaviour which may affect the decisions made are required. This information will allow modellers and planners to make predictions and estimates so that future demands can be met. A monitor tool has therefore been developed. It allows the traffic of the network to be monitored and measurements to be retrieved, displayed and analysed. A decentralized approach which provides an integrated measurement facility has been adopted. The design, and the decisions and constraints which influenced that design, are desribed. From the measurements gathered a comprehensive traffic characterization is provided. It relates traffic characteristics of different grains to applications, to system characteristics and to constraints. The measurement analysis therefore establishes a firm base from which predictions and estimates may be made. It also provides a base for comparisons, one which allows the effects of hardware and software changes to be observed and which also provides valuable information to both planners and modellers of similar and different environments. Implementation errors and erroneous behaviour are demonstrated and their cause is established. A bottleneck is identified, performance thresholds are determined and protocol modifications are suggested. Specific traffic characteristics are identified for particular applications and user groups, effects of different buffer sizes are considered and relationships between protocol efficiency and traffic patterns are discussed

In order to meet demands in mobile broadband and to bridge the digital divide a new technology, namely WiMAX, was introduced in 2004. However, in order to increase the financial return on the investment inWiMAX, service operators need to make every effort in designing and deploying the most cost-effective networks. This thesis presents a novel dimensioning technique for WiMAX technology which takes the dimensioning problem to a new level and produces more accurate results in comparison to the traditional methods. Furthermore, a novel decomposed optimization framework for the WiMAX network planning is introduced which subdivides the overall problem into three distinct stages consisting of the network dimensioning stage which besides the primary task of evaluating the financial requirements produces a good starting network solution for the subsequent stages (Stage 1), initial sectorization and configuration of the network (Stage 2) and final network configuration (Stage 3). The proposed framework also solves two fundamental problems, which are cell planning and frequency planning, simultaneously. The feasibility of the final network solutions are then evaluated by OPNET simulator.

This thesis presents a novel methodology for generation expansion planning. The method is based on the Optimal Power Flow (OPF), a common tool for the economic operation of power systems. New generation capacity is simulated with the real power of virtual generators located at the candidate connection points, so OPF is used to plan generation expansion with respect to operating constraints of the existing network. A new method had to be developed for the direct incorporation of protection constraints in generation expansion. The modelling of new capacity with virtual generators gives access to power flow control variables. Binding constraints for generation expansion can be expressed as constrained functions of those variables. Accordingly, expected fault currents were expressed as functions of OPF variables and protection equipment specifications were converted to constraints for these functions. Thereafter, the allocation of new capacity by the OPF directly respects both system and fault constraints. The iterative approach has been proven less efficient than the later approach, but still maintains some advantages if the method is to be commercially exploited. Generator voltage control policies can also be converted to OPF constraints. The functionality of the suggested generation capacity allocation method was expanded to operate as an assessment tool of their impact on the amount of new capacity that a network can absorb. The method was expanded further, so as to consider the impact of capacity allocation on transmission losses. With a minor reformulation of the original method a new tool was designed for the optimal sitting of reactive power compensation banks for the implement of network headroom. Finally, a network planning method is presented based on the LaGrange multipliers, sensitivity by-products of the OPF solution method, which connect network constraints with generation expansion. Generation expansion is planned simultaneously with network reinforcement, so the overall optimum is achieved. The main conclusion of this work is that OPF can be used as a powerful planning, as well as operating tool. Its flexible formulation allows the incorporation of emerging constraints in generation and network expansion, such as those imposed by protection.

Increased penetration of distributed generations in distribution networks are altering the technical characteristics of the grid, pushing them to operate closer to their limits of safe and reliable operation. New renewable generators connecting to the distribution network will be constrained due to the presence of thermal and voltage constraints during times of low demand and high generation output. The traditional reinforcement planning by means of increasing the capacity of network assets can be very costly and usually ends up in overinvested network with low utilization rates of the assets. In recent years, some smart technologies have been introduced to be used to increase the utilization of network assets and provide the adequate capacity for Distributed Generations (DGs). These smart solutions can help the Distributed Network Operators (DNOs) to provide cheaper and faster network connections for DGs. This thesis presents a multi epoch Optimal Power Flow (OPF) model for capacity and voltage management of a distribution network for integrating new DGs. The model uses the smart solutions including Dynamic Line Rating (DLR), Quad-Booster (QB), Static VAR Compensator (SVC) and Automatic Network Management (ANM) for DGs as well as the traditional reinforcement options. Also the model finds the optimal connection points for new DGs to reduce the cost of network investment and DG curtailment. The multi epoch model is solved with both incremental approach where the investment is carried out incrementally and with integrated approach where the planning is done strategically anticipating the future needs of the network. It compares the application of smart solutions in short and long term planning. The proposed model is applied to a generic UK distribution network and the results are discussed.

This master thesis project analyses how an increased number of electric vehicles would affect the electric network in Stockholm. The project was started in order to help Ellevio, the DSO of Stockholm, understand how to manage the increased amount of requests of network-capacity connected to electric vehicle charging. Five different areas have been simulated in the software Tekla to investigate how many electric vehicle chargers with a rated power of 3.7 kW that can be run at the same time without exceeding network limits in terms of voltage drops in consumption points, maximum current in cables and apparent power in the secondary substation. Two different charging curves spanning over eight hours each were used during the simulations to achieve the worst plausible scenario. The first charging curve took place between 16:00-00:00 and the second between 22:00-06:00. By using two charging curves it was possible to investigate if more electric vehicle charging could occur at the same time if the charging is moved to night time. The results show that the first problem to occur in all investigated areas is overloads of cables. However, this problem can be remedied in especially areas with apartments and in the city by using equipment which can reduce the output of the chargers depending on the load magnitude of the line connection. The city network could handle highest numbers of charging without exceeding the limits, followed by the areas with apartments and at last the areas with small houses. However, it can be seen in the results that the small house areas can handle about the same number of chargers as the city network if a few cables are replaced. Finally, all areas can handle more charging without exceeding the limits if the charging is moved to night-time.
Det här examensarbetet på mastersnivå analyserar hur en utökad flotta av elbilar skulle påverka elnätet i Stockholm. Examensarbetet startades för att hjälpa Ellevio, Stockholms elnätsägare, förstå hur de ska hantera den ökade efterfrågan av elnätskapacitet för elbilsladdning. Fem områden i Stockholm har simulerats i programmet Tekla för att utreda hur många elbilar som kan laddas samtidigt med en effekt på 3.7 kW vardera utan att kablarna och nätstationen blir överbelastade samt utan att spänningsfallen i lastpunkterna blir för stora i respektive område. Under simuleringarna har två olika laddkurvor på åtta timmar vardera använts för att få fram värsta möjliga scenariot. Den ena laddkurvan var mellan klockan 16:00-00:00 och den andra mellan klockan 22:00-06:00. Genom att ha två kurvor gick det att undersöka hur mycket mer elbilsladdning som kunde vara igång samtidigt om den förflyttades till nattetid istället. Resultaten visar att det första problem som uppstår i samtliga områden är att kablarna blir överbelastade. Det här problemet går dock att avhjälpa i framförallt lägenhetsområdena och innerstaden då det finns utrustning som begränsar uteffekten på laddarna beroende på hur belastad servisen är. Innerstadsnätet var det område som klarade att ha flest laddare igång samtidigt utan att överlast eller för stora spänningsfall uppstod, därefter kom lägenhetsområdena och till sist villaområdena. Det går dock att se i resultaten att genom att byta ett fåtal kablar i villaområdena så klarar de liknande nivåer av elbilsladdning som innerstadsområdet. Slutligen går det att se hur det genom att flytta laddningen till nattekid kan vara flera laddare igång samtidigt i alla områden utan att problem med överlast eller spänningsfall uppstår.

Transportation planners need to assess users\' stress conditions on cycling infrastructures given that highly stressful situations can discourage the use of this sustainable transport mode. Not many studies have addressed the relationships between these environmental factors and users\' emotional responses in terms of objective measures of stress. The aim of this study is to explore a new approach for planning and evaluating cycling infrastructure, based on smart sensors. This new approach focuses on the perspective of monitoring parameters intrinsic to the user, such as emotions. In this perspective, the indicators of stress levels are made from directly measuring cyclists\' physiological responses throughout the journey. This approach makes use of technological resources to extract information from users through sensors and imparts this information in an integrated way to improve infrastructures for cyclists. The data were collected using stress and noise sensors, accelerometers embedded in a smartphone and a GPS. Initially, the problem is posed and we discuss how the design and evaluation of cycle paths has been addressed in the literature. The proposed approach identifies the most relevant contributions and research gaps, such as the lack of research based on objective criteria and research that involves designing and evaluating infrastructure scenarios shared with motor vehicles. In the sequence, an objective method for assessing bicycle infrastructures combining environmental assessment with vertical acceleration measurements was proposed in order to improving data collection and other procedures required for assessing the main components of cycling infrastructures. Subsequently, a tool to characterize traffic stress of cycling routes, called the Level of Traffic Stress (LTS), was validated with physiological measures in the urban context of a mediumsized Brazilian city. It was observed that there is no correlation and little agreement between the parameters. It is also emphasized that even when incorporating information about the traffic speed to the LTS tool, it was not significantly related to stress from the perspective of the cyclist. In the final analysis the influence of noise, vibration, cycle paths and period of day on stress experienced by cyclists was investigated. An analysis of the p-values and odds ratio confidence intervals shows, with a 95% confidence level, that only the period of the day influenced stress, as confirmed by the data. In this case, the chances of having stress increased by 24% in the afternoon rush hour compared to the morning rush hour. This study showed the feasibility of stress assessment in cyclists using an objective measurement method, as well as quick identification of critical levels of stress.
Os planejadores de transporte precisam avaliar as condições de estresse dos usuários em infraestruturas cicloviárias, uma vez que situações de alto estresse podem desencorajar o uso deste modo de transporte sustentável. Poucas pesquisas têm questionado se existe alguma relação entre esses fatores ambientais e a resposta emocional em termos de medidas objetivas do estresse. O objetivo deste trabalho é explorar a incorporação de sensores inteligentes no planejamento e avaliação da infraestrutura cicloviária. Essa nova abordagem é focada na perspectiva de monitorar parâmetros intrínsecos ao usuário, como as emoções. Nesta perspectiva, os indicadores dos níveis de estresse são feitos a partir da medição direta de respostas fisiológicas em ciclistas ao longo do percurso. Essa abordagem aproveita os recursos tecnológicos para extrair informações dos usuários e permitir o uso dessas informações de forma integrada para melhorar a infraestrutura dos ciclistas. Os dados foram coletados por meio de sensores de estresse, de ruído e acelerômetros incorporados a um smartphone e GPS. Inicialmente é apresentado o problema e como o projeto e avaliação das ciclovias tem sido abordada na literatura. A abordagem proposta permitiu identificar as contribuições mais relevantes e as lacunas de pesquisa, tais como, a falta de pesquisas baseadas em critérios de objetivos e a falta de pesquisas que envolvam o desenho e a avaliação de cenários de infraestrutura compartilhados com veículos automotores. Em seguida foi proposto um método objetivo de avaliação de infraestruturas cicloviárias, combinando a avaliação ambiental com medições de aceleração vertical, a fim de melhorar a coleta de dados e outros procedimentos necessários para avaliar os principais componentes das infraestruturas cicloviárias. Na sequência uma ferramenta para caracterizar o estresse de tráfego de ciclovias, denominada Level of Traffic Stress (LTS), foi validada com medidas fisiológicas no contexto urbano de uma cidade brasileira de porte médio. Observou-se que não há correlação e há pouca concordância entre esses parâmetros. Ressalta-se ainda que, mesmo ao incorporar informações de velocidade de tráfego à ferramenta LTS, não foi significativamente relacionado com estresse medido sob a perspectiva do ciclista. Por fim, foi investigada a influência do ruído, vibração, presencia ou ausência de ciclovias e período do dia no estresse experimentado pelos ciclistas. Uma análise dos valores de p e dos intervalos de confiança das razões de chances mostraram, com um nível de confiança de 95%, que apenas o período do dia influenciou o estresse. Neste caso, as chances de ter estresse aumentaram em 24% na hora pico da tarde em comparação com a hora pico da manhã. Este estudo mostrou a viabilidade da avaliação do estresse em ciclistas por meio de um método de medida objetiva além da rápida identificação dos níveis críticos de estresse.

This thesis examines the use of network governance in US airport transportation planning activities involving taxicab services for airport patrons. The research provides US airports with new insights whereby they can successfully engage with both transportation regulatory agencies and taxicab service providers in developing mutually agreeable policies that foster the development of supply-side taxicab service improvements. A mix of quantitative and qualitative research methods is used to unearth how US airports interact with these actors, and to identify attitudes held by airport staff in their engagements involving airport taxicab planning matters. The research may ultimately lead to the achievement of sustainable increases in the air passenger ground transportation modal share at US airports, resulting in both desirable long-term operational and environmental benefits for airport management, those involved with the provision of airport taxicab services, and the traveling public.

I present novel solutions to two different classes of computational problems that arise in the study of complex cellular processes. The first problem arises in the context of planning large-scale genetic cross experiments that can be used to validate predictions of multigenic perturbations made by mathematical models. (i) I present CrossPlan, a novel methodology for systematically planning genetic crosses to make a set of target mutants from a set of source mutants. CrossPlan is based on a generic experimental workflow used in performing genetic crosses in budding yeast. CrossPlan uses an integer-linear-program (ILP) to maximize the number of target mutants that we can make under certain experimental constraints. I apply it to a comprehensive mathematical model of the protein regulatory network controlling cell division in budding yeast. (ii) I formulate several natural problems related to efficient synthesis of a target mutant from source mutants. These formulations capture experimentally-useful notions of verifiability (e.g., the need to confirm that a mutant contains mutations in the desired genes) and permissibility (e.g., the requirement that no intermediate mutants in the synthesis be inviable). I present several polynomial time or fixed-parameter tractable algorithms for optimal synthesis of a target mutant for special cases of the problem that arise in practice. The second problem I address is inferring gene regulatory networks (GRNs) from single cell transcriptomic (scRNA-seq) data. These GRNs can serve as starting points to build mathematical models. (iii) I present BEELINE, a comprehensive evaluation of state-of-the-art algorithms for inferring gene regulatory networks (GRNs) from single-cell gene expression data. The evaluations from BEELINE suggest that the area under the precision-recall curve and early precision of these algorithms are moderate. Techniques that do not require pseudotime-ordered cells are generally more accurate. Based on these results, I present recommendations to end users of GRN inference methods. BEELINE will aid the development of gene regulatory network inference algorithms. (iv) Based on the insights gained from BEELINE, I propose a novel graph convolutional neural network (GCN) based supervised algorithm for GRN inference form single-cell gene expression data. This GCN-based model has a considerably better accuracy than existing supervised learning algorithms for GRN inference from scRNA-seq data and can infer cell-type specific regulatory networks.
Doctor of Philosophy
A small number of key molecules can completely change the cell's state, for example, a stem cell differentiating into distinct types of blood cells or a healthy cell turning cancerous. How can we uncover the important cellular events that govern complex biological behavior? One approach to answering the question has been to elucidate the mechanisms by which genes and proteins control each other in a cell. These mechanisms are typically represented in the form of a gene or protein regulatory network. The resulting networks can be modeled as a system of mathematical equations, also known as a mathematical model. The advantage of such a model is that we can computationally simulate the time courses of various molecules. Moreover, we can use the model simulations to predict the effect of perturbations such as deleting one or more genes. A biologist can perform experiments to test these predictions. Subsequently, the model can be iteratively refined by reconciling any differences between the prediction and the experiment. In this thesis I present two novel solutions aimed at dramatically reducing the time and effort required for this build-simulate-test cycle. The first solution I propose is in prioritizing and planning large-scale gene perturbation experiments that can be used for validating existing models. I then focus on taking advantage of the recent advances in experimental techniques that enable us to measure gene activity at a single-cell resolution, known as scRNA-seq. This scRNA-seq data can be used to infer the interactions in gene regulatory networks. I perform a systematic evaluation of existing computational methods for building gene regulatory networks from scRNA-seq data. Based on the insights gained from this comprehensive evaluation, I propose novel algorithms that can take advantage of prior knowledge in building these regulatory networks. The results underscore the promise of my approach in identifying cell-type specific interactions. These context-specific interactions play a key role in building mathematical models to study complex cellular processes such as a developmental process that drives transitions from one cell type to another

Fundamental problems in graph theory are of four types existence, construction, enumeration and optimization problems. Optimization problems lie at the interface between computer science and the field of operations research and are of primary importance in decision-making. In this thesis, two optimization problems are studied: the edge-integrity of networks and the network flow problem. An implementation of the corresponding algorithms is also realized.The edge integrity of a communication network provides a way to assess the vulnerability of the network to disruption through the destruction or failure of some of its links. While the computation of the edge-integrity of graphs in general has been proven to be NPcomplete, a recently published paper was devoted to a good algorithm using a technique of edge separation sequence for computing the edge integrity of trees. The main results of this paper will be presented and an implementation of this algorithm is achieved.The network flow problem models a distribution system in which commodities are flowing through an interconnected network. The goal is to find a maximum feasible flow and its value, given the capacity constraints for each edge. The three majors algorithms for this problem (Ford -Fulkerso n, Edmonds-Karp method, MPKM algorithm) are discussed, their complexities compared and an implementation of the Ford-Fulkerson and the MPKM algorithms is presented.
Department of Computer Science

The purpose of this research is to investigate, acquire knowledge, and better comprehend the transportation planning and modeling community. This task is accomplished through the investigation of existing social and professional networks within this community by constructing and analyzing an abstract network representation of this community. Specifically, this research explores the actors (i.e., professionals, agencies) and their relationships (i.e., ties, interactions, etc.) within the professional interpersonal (social) network where they conduct business on a regular basis. Actors and relationships are represented in terms of a nodes and links within the constructed network. The network is then analyzed in an effort to answer questions such as, who are the actors, where do interpersonal relationships exist, where are social structures found, what does the evolution of this community look like over time, and what can this evolution tell us. This study has collected information from transportation professionals directly associated with the decision-making, planning and/or modeling process within the transportation planning and modeling community. The data is collected through an in-house designed online survey disseminated to the identified target audience. The designed survey is structured to capture information required for the identification of actors and relationships (or entities and ties) within the transportation planning and modeling community. With the network constructed, analysis methods derived from mathematics, computer science and social network analysis fields are implemented to identify local and global patterns, "influential" actors, and collaborative structures as well as examine network dynamics, which transpire within the environment that these transportation professionals navigate, form bonds, and collect information on frequent basis.

The aim of this thesis was to bridge a gap in the research literature with regard to commentary on and evaluation of the relationship between British land-use planning and the management and development of the railway network in the years between 1948-94 when British railways were in public ownership. Although the research was focused on the nationalised main line system, it reviewed other rail systems where this was helpful to the analysis. The research utilised a review of the relationship between the railway network and urban form in the years to 1947 to derive analytical criteria and to serve as a point of departure for the core of the thesis. The overall relationship between the two sectors post-1948 was explored, at a broad geographical scale, with regard to institutional relationships, policy, and outcomes with regard to the spatial relationships between the railway network and patterns of urban form. The results of this research were used to derive hypotheses about the relationships which were then tested in a case study of the Manchester conurbation. The main conclusions are that there were few periods between 1948-94 when the ideological, institutional and policy frameworks necessary for a close and positive relationship between the planning and railway sectors were in place simultaneously. The contexts which were most favourable were with regard to: the location of new towns and town expansion projects in the South East in 1950s and 1960s; the improvement of railway networks in the PTE areas between 1968-79 along with the development of strategic policies for the restriction of major trip generators to CBDs; the period between 1985-94 when a surge in the property market was accompanied by BR Sectorisation, investment in other forms of fixed track transit, and the promotion of major development projects at and around stations, especially in CBDs. The research concludes by identifying opportunities for further historical research and briefly reviewing the relevance of the findings to contemporary research.

Although China has a huge cyclist population, the cycling condition in large cities is undesirable. The problem is mainly caused by the mixed-flow of cyclists and motorist on the road. Separation of cyclists and motorist is the key to solve the problem. Based on the research of the successful examples in Europe and a PEBOSCA inventory analysis of the cycling traffic in Nanjing, a set of suggestion is proposed on how to plan the bicycle network. A bicycle network separated from motor traffic and the planning on a cycling district division is introduced. The proposal also involves the suggestion on combination between cycling and public transport, and the experience route of historical culture and the natural beauty of the landscape.

Wide-area camera networks are becoming more and more common. They have widerange of commercial and military applications from video surveillance to smart home and from traffic monitoring to anti-terrorism. The design of such a camera network is a challenging problem due to the complexity of the environment, self and mutual occlusion of moving objects, diverse sensor properties and a myriad of performance metrics for different applications. In this dissertation, we consider two such challenges: camera planing and camera fusion. Camera planning is to determine the optimal number and placement of cameras for a target cost function. Camera fusion describes the task of combining images collected by heterogenous cameras in the network to extract information pertinent to a target application. I tackle the camera planning problem by developing a new unified framework based on binary integer programming (BIP) to relate the network design parameters and the performance goals of a variety of camera network tasks. Most of the BIP formulations are NP hard problems and various approximate algorithms have been proposed in the literature. In this dissertation, I develop a comprehensive framework in comparing the entire spectrum of approximation algorithms from Greedy, Markov Chain Monte Carlo (MCMC) to various relaxation techniques. The key contribution is to provide not only a generic formulation of the camera planning problem but also novel approaches to adapt the formulation to powerful approximation schemes including Simulated Annealing (SA) and Semi-Definite Program (SDP). The accuracy, efficiency and scalability of each technique are analyzed and compared in depth. Extensive experimental results are provided to illustrate the strength and weakness of each method. The second problem of heterogeneous camera fusion is a very complex problem. Information can be fused at different levels from pixel or voxel to semantic objects, with large variation in accuracy, communication and computation costs. My focus is on the geometric transformation of shapes between objects observed at different camera planes. This so-called the geometric fusion approach usually provides the most reliable fusion approach at the expense of high computation and communication costs. To tackle the complexity, a hierarchy of camera models with different levels of complexity was proposed to balance the effectiveness and efficiency of the camera network operation. Then different calibration and registration methods are proposed for each camera model. At last, I provide two specific examples to demonstrate the effectiveness of the model: 1)a fusion system to improve the segmentation of human body in a camera network consisted of thermal and regular visible light cameras and 2) a view dependent rendering system by combining the information from depth and regular cameras to collecting the scene information and generating new views in real time.

Thesis: Ph. D. in Transportation, Massachusetts Institute of Technology, Department of Civil and Environmental Engineering, 2017.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 141-147).
Public transport plays an important role in urban mobility. Public transport planners seek to improve existing public transport networks to better serve existing passengers and recruit new passengers, particularly as demand patterns change with evolving demographics and land use. One strategy for network improvement is to add new routes, which can improve service by reducing circuity in the network. In large, complex, and multi-modal public transport networks, it is challenging to determine where new routes should be added. A systematic approach for incremental network improvements, such as adding new bus services, is needed. This research proposes a new approach to network-level public transport planning by combining origin-destination (OD) level analysis with new spatial aggregation methodologies, and develops a comprehensive framework for the identification of corridors for new bus services. In the context of this framework, this dissertation contributes several new methodologies. First, it proposes a methodology for defining zones that reflect the spatial characteristics of a public transport network. This produces zonal pairs that are appropriate for OD level analysis of travel in the network. Second, the dissertation develops metrics and rules for the identification of OD pairs that can benefit from new bus services, and proposes methods for estimating the expected benefits of such services at the OD level. Finally, a new methodology for spatially clustering OD pairs into corridors is developed, based on trajectory clustering methods. This final methodology represents a new way of aggregating OD level information to accomplish the first step in bus network design: the definition of corridors for new services. The framework is demonstrated for the identification of corridors for new bus services in the London public transport network. Bus stops and rail stations are clustered into 1,000 zones. A subset of zonal OD pairs with circuitous service are identified as candidates for improvement through new bus routes. An algorithm that clusters OD pairs into corridors for bus service is developed and applied. Several promising corridors are identified, and their potential is confirmed in post-analysis.
by Cecilia A. Viggiano.
Ph. D. in Transportation

Thesis: S.M. in Transportation, Massachusetts Institute of Technology, Department of Civil and Environmental Engineering, 2014.
Cataloged from PDF version of thesis.
Includes bibliographical references (pages 145-146).
The need for optimized aviation lift planning is becoming increasingly important as the United States and her allies participate in the Global War on Terror (GWOT). As part of a comprehensive effort, our nation's fighting forces find themselves conducting operations around the globe, with this trend likely to increase, even as budget constraints limit the number of personnel and amount of equipment that is deployed. While much attention has been given to airline schedule optimization and fleet planning. the challenge of Army Aviation lift planning is unique in that it must be able to adapt to changing requirements and missions on a daily basis. In this thesis, we model Army Aviation lift planning as a service network design problem, and propose two heuristic algorithms, which compare favorably to current human planning systems. Furthermore, we apply these heuristic algorithms to long term asset planning and capacity requirement estimation for future military scenarios, and analyze how passenger flexibility affects the need for capacity.
by Matthew D. Mogensen.
S.M. in Transportation

博士
國立臺灣大學
資訊管理學研究所
94
An increasing number of new services are now running on CATV networks. The earliest CATV (Community Antenna Television) systems were constructed in small towns or semi-rural areas, where off-air television reception was poor or unavailable [. Because of their popularity and high bandwidth, CATV such networks have become one of the most popular technologies for providing a “last-mile” communication platform. The quality of CATV network systems depends to a large extent on the experience of the designers who must consider the performance constraints mandated by standards and government regulations. Consequently, the quality of CATV network design may be unreliable, and in many cases poor. In this dissertation, we study CATV networks planning problems. Mathematical formulations are used to model the planning problems, and geometric programming method, based on the proposed mathematical formulations, is adopted to solve the network planning problems. The scope and contributions of this dissertation are highlighted by the following. For the min-cost CATV networks planning problem, we propose a mathematical model to describe CATV networks planning problem. Based on some mathematical features of the model, some reformulations are necessary to solve the problem. The surrogate functions are used to reformulate the objective function and some constraints. By applying some nonlinear programming techniques, the single layer solution procedure for CATV network planning problems is developed. Some computational experiments are described and explained. From the experiment results, the solution procedure we developed is better than previous works. The comparison showed that our solution procedure is better in most of cases. The improvements on minimum costs are ranged from 51% to 92%. Based on the experiment results, we get some important finding in this problem, especially about the parameters settings in solution procedure. By the setting rules presented in this chapter, the solution quality, both the minimum cost and the scalability of the problem, can be further improved. From the analysis of the solution procedure, however, we still could not deal with problems with too many nodes. Therefore, a multilayer solution procedure is proposed in Chapter 4. By layering a large network into several smaller networks, we can divide the problem and conquer every sub problems in reasonable time. After that, we can treat each network as a macro user in upper layer, and construct the network planning problem for upper layer. By summation the costs of upper layer and every sub layers, we can get the total cost of the entire network. By the multilayer solution procedure, we can solve CATV network planning problems with more nodes. We have compared with the single-layer solution procedure and show that only 40% of time is needed in multi-layer solution procedure. On the other side, the minimum costs solved by multilayer solution procedure are ranged from 2% to 45% larger than single-layer solution procedure. By balancing the computation time and solution quality, the multilayer solution procedure still provides a way to solve a larger network in limited time. Besides the costs and computing time, we have developed algorithms for placement of drop points. In order to improve the costs of CATV networks, the placement of drop points in clusters is adjusted by proposed globally adaptive placecment algorithm. Based on experiment results, the reduced costs ranges from 9% to 13%. With tradeoff between computing time and costs, we propose partially adaptive placement algorithm, which only adjust the leave nodes on upper layer networks. Compared with globally adjustment, the computing time is reduced to 61.5% and only 4.88% cost increased. Finally, we point out three challenging issues to be tackled in the future. These issues include adjustment procedure between layers in multiplayer solution procedure, how to apply the solution procedures to other kinds of application environments, and modifications for HFC (Hybrid Fiber/Coax) networks.

博士
國立臺灣大學
電信工程學研究所
100
Integrating relay stations (RSs) into the architecture of centralized wireless access networks to support multi-hop relaying has become a major trend for 4G wireless communication systems. Though commonly believed that RSs can effectively extend the coverage and enhance throughput of a network, integrating RSs into the system can bring as many challenges as benefits. Much more sophisticated network planning is required to make full use of RS’s functionality and to optimize the performance of a relay-assisted wireless access network. In this dissertation, we study several critical issues for wireless relay networks. The first is the network deployment problem. We analyze the impacts of deployment locations on system performances for different types of network nodes. Considering different deployment scenarios and user distributions, we propose deployment strategies determining which types of stations should be deployed and their deployed locations, under a budget constraint. Due to the capacity limitation at each relay station, simply having each relay station independently relay traffics for their serving base station can only achieve limited performance gain, which motivates us to propose the concept of relay station cooperation. The second problem we studied is therefore how to coordinate relay stations to cooperatively serve the downstream users such that system capacity can be maximized without sacrificing fairness. Through analyzing the impacts of relay cooperation on its coverage and downstream rate, we have mathematically formulated this problem, and propose joint relay cooperation and resource assignment solution. The effect of relay station placement on the performance of relay station cooperation is also studied. The last problem tackled is how to select data forwarding strategies for relay stations when network coding is enabled for multi-cast data repair. We propose a Markov decision process to characterize the data forwarding process, by which a delay-optimal dynamic programming algorithm as well as two light-weight on-line heuristics are derived. To summarize, several critical issues induced when integrating relays into a wireless access network are studied in this dissertation. We believe that our analysis and observations are indispensible for the planning and performance optimization of next generation wireless relay network. Integrating relay stations (RSs) into the architecture of centralized wireless access networks to support multi-hop relaying has become a major trend for 4G wireless communication systems. Though commonly believed that RSs can effectively extend the coverage and enhance throughput of a network, integrating RSs into the system can bring as many challenges as benefits. Much more sophisticated network planning is required to make full use of RS’s functionality and to optimize the performance of a relay-assisted wireless access network. In this dissertation, we study several critical issues for wireless relay networks. The first is the network deployment problem. We analyze the impacts of deployment locations on system performances for different types of network nodes. Considering different deployment scenarios and user distributions, we propose deployment strategies determining which types of stations should be deployed and their deployed locations, under a budget constraint. Due to the capacity limitation at each relay station, simply having each relay station independently relay traffics for their serving base station can only achieve limited performance gain, which motivates us to propose the concept of relay station cooperation. The second problem we studied is therefore how to coordinate relay stations to cooperatively serve the downstream users such that system capacity can be maximized without sacrificing fairness. Through analyzing the impacts of relay cooperation on its coverage and downstream rate, we have mathematically formulated this problem, and propose joint relay cooperation and resource assignment solution. The effect of relay station placement on the performance of relay station cooperation is also studied. The last problem tackled is how to select data forwarding strategies for relay stations when network coding is enabled for multi-cast data repair. We propose a Markov decision process to characterize the data forwarding process, by which a delay-optimal dynamic programming algorithm as well as two light-weight on-line heuristics are derived. To summarize, several critical issues induced when integrating relays into a wireless access network are studied in this dissertation. We believe that our analysis and observations are indispensible for the planning and performance optimization of next generation wireless relay network.

博士
國立臺灣大學
資訊管理研究所
89
Fast evolving computer network technologies have enabled the new communication and way of life. In addition, new logistic network concept (such as supply chain network) could decrease the business transaction cost. As a result, different types of networks have widely been installed and operated. However, how to efficiently allocate the resource to satisfy the users’ quality-of-service is a difficult and important issue. In this dissertation, solution procedures based on rigorous mathematical programming and Lagrangean relaxation are proposed to solve a series of network planning and capacity management problems for computer and logistic networks. In computer networks, we consider the lightpath routing and wavelength assignment in WDM networks, access network design, backbone network design, operational support for in-service network. In logistic networks, the network planning for service network and supply chain network are addressed. By using the WDM technique, we could multiplex more than one hundred wavelengths on a single fiber, such that the bandwidth could be increased enormously. However, under the constraint of wavelength continuity, it is difficult to perform lightpath routing and wavelength assignment on each fiber efficiently. We solve this problem by Lagrangean relaxation technique in conjunction with several optimization-based algorithms. From the computational experiment, we could obtain good solution quality for network size up to 26 nodes in minutes of computational time. In access network design, we consider the network design problem with the shared tree topology. The difficulty of this work lies in the multicast traffic, which will lead to the Steiner Tree problem. By using the sophisticated problem formulation, we could solve this problem without the necessity of solving Steiner Tree problem. The optimization-based algorithms that we propose could provide near-optimal solutions to this problem for network size up to 26 nodes in minutes of computational time. In backbone network design, we consider the design problem with the system and user specified QoS requirements. The objective is to minimize the installation cost for transmission links and network switching nodes. The system specified QoS considered in this work is the average delay requirement, and the user specified QoS include the end-to-end delay requirement and the k node disjoint path survivability requirement. The difficulty of this work lies in the integer and non-convexity mathematical structure. Algorithms based on Lagrangean relaxation and add-drop heuristic could provide efficient and effective solutions to the backbone network design problem. For any in-service network, daily increasing traffic demand will cause the inferior QoS perceived by the users; two approaches -- rerouting and sizing, are proposed to address this issue. For rerouting, we consider the optimization problem to address the system optimization and user fairness issue at the same time where the objective is to minimize the system (average) delay with the user specified end-to-end delay requirement. For sizing, we consider the optimization problem to minimize the network link capacity installation cost with the system (average) delay and users’ end-to-end delay requirement. Besides integrality constraints, the nonconvexity and concavity mathematical structure makes this problem difficult. Lagrangean relaxation in conjunction with a number of getting primal heuristics are proposed to address these two problems. By accessing the solution quality to the rerouting problem, we could provide near-optimal solutions to this problem for network size up to 26 nodes in minutes of computational time. In addition, by accessing the solution quality to the sizing problem, we propose efficient and effective optimization-based algorithms to solve this problem. In service network design, we consider the location and capacity management problems of service centers. Capacity constraint and response time constraint make this problem more difficult than traditional location problem. Optimization-based algorithms based on Lagrangean relaxation could provide efficient and effective solutions to this problem for network size up to 350 nodes in minutes of computational time. In supply chain network design, besides capacitated location problem of distribution center, routing problems of transmission vehicles are considered. This problem is formulated as integer programming problem. Optimization-based algorithms based on Lagrangean relaxation could provide efficient and effective solutions to this problem.

abstract: Due to great challenges from aggressive environmental regulations, increased demand due to new technologies and the integration of renewable energy sources, the energy industry may radically change the way the power system is operated and designed. With the motivation of studying and planning the future power system under these new challenges, the development of the new tools is required. A network equivalent that can be used in such planning tools needs to be generated based on an accurate power flow model and an equivalencing procedure that preserves the key characteristics of the original system. Considering the pervasive use of the dc power flow models, their accuracy is of great concern. The industry seems to be sanguine about the performance of dc power flow models, but recent research has shown that the performance of different formulations is highly variable. In this thesis, several dc power-flow models are analyzed theoretically and evaluated numerically in IEEE 118-bus system and Eastern Interconnection 62,000-bus system. As shown in the numerical example, the alpha-matching dc power flow model performs best in matching the original ac power flow solution. Also, the possibility of applying these dc models in the various applications has been explored and demonstrated. Furthermore, a novel hot-start optimal dc power-flow model based on ac power transfer distribution factors (PTDFs) is proposed, implemented and tested. This optimal-reactance-only dc model not only matches the original ac PF solution well, but also preserves the congestion pattern obtain from the OPF results of the original ac model. Three improved strategies were proposed for applying the bus-aggregation technique to the large-scale systems, like EI and ERCOT, to improve the execution time, and memory requirements when building a reduced equivalent model. Speed improvements of up to a factor of 200 were observed.
Dissertation/Thesis
M.S. Engineering 2013