No description available.

Submitted by Nara Lays Domingues Viana Oliveira (naradv) on 2015-07-15T14:37:25Z No. of bitstreams: 1 ROBERTO.pdf: 6262020 bytes, checksum: 76c7611d1d91674b302b17af7b02b0e4 (MD5) / Made available in DSpace on 2015-07-15T14:37:25Z (GMT). No. of bitstreams: 1 ROBERTO.pdf: 6262020 bytes, checksum: 76c7611d1d91674b302b17af7b02b0e4 (MD5) Previous issue date: 2014 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / A migração de processos é uma técnica utilizada no remapeamento de um processo para um processador mais rápido ou para aproximá-lo de outros processos com os quais se comunica frequentemente. Esta dissertação descreve o MigBSP++, um modelo de reescalonamento de processos que utiliza a técnica de migração para realizar o balanceamento de carga em sistemas paralelos. Direcionado às aplicações do tipo Bulk-Synchronous Parallel (BSP), o modelo apresentado redistribui os processos com o intuito de reduzir o tempo de cada super-passo. De modo similar ao MigBSP, o MigBSP++ combina múltiplas métricas a fim de decidir as migrações necessárias para que o sistema entre em equilíbrio sem a intervenção do usuário. As métricas utilizadas são: computação, comunicação e sobrecusto de migração. Através de sua função de decisão, chamada Potencial de Migração (PM), essas métricas são utilizadas para eleger os processos mais propícios a trazer o equilíbrio ao sistema. O MigBSP++ responde as questões necessárias para a política de migração de processos: quando realizar a migração de processos; quais processos são candidatos à migração e; para onde migrar os processos selecionados. Como contribuição científica, o MigBSP++ introduz as soluções para duas questões que estão em aberto no MigBSP: (a) a detecção de desbalanceamento de carga quando há mais processos do que processadores e; (b) a definição de quantos processos irão migrar de fato. Para a questão (a), propõe-se alteração do modo de detecção de desbalanceamento utilizada, observando o tempo total de computação de cada processador. Para a questão (b) é apresentado um algoritmo chamado de Algoritmo de Predição BSP (APBSP). Os dados de entrada do APBSP são os processos eleitos pela técnica de PM e a saída é uma lista de processos que irão, de fato, migrar proporcionando a redução do tempo do próximo super-passo. Para demonstrar os resultados da aplicação deste modelo, foram desenvolvidas duas aplicações BSP com o auxílio da biblioteca Adaptive Message Passing Interface (AMPI). Essa ferramenta oferece um arcabouço uniforme que, através da migração de processos, permite o balanceamento de carga de forma transparente ao usuário. Foram desenvolvidas as estratégias de balanceamento de carga, baseadas no MigBSP e no MigBSP++, para a realização da comparação entre elas e com as estratégias já existentes no sistema. Os resultados apontam que, nos casos onde a granularidade da tarefa é maior, os ganhos em tempo de execução são mais evidentes, podendo ser de até 46% em relação à aplicação sem balanceamento e de até 37% em relação às estratégias nativas do AMPI. Esses números sugerem que o modelo MigBSP++ tem aplicação prática e pode produzir resultados satisfatórios. / Process migration is a technique used in the remapping of a process to a faster processor or in the approaching from the processes which already have some communication among themselves. This essay describes the MigBSP++, a rescheduling process model that uses the technique of migration to perform load balancing in parallel systems. Directed to the BulkSynchronous Parallel (BSP) applications, the model redistributes the processes with the purpose of reducing the time of each super-step. Similar to MigBSP way, MigBSP++ combines multiple metrics to decide which migrations should be chosen in order to balance the entire system without the user intervention. The metrics used by the model are: computing, communication and extra costs of migration. Through its decision function, called Potential Migration (PM), these metrics are used to choose the most appropriate processes that will balance the system. MigBSP++ answers the questions about the policy process migration issues: when to perform the migration process, which processes are candidates for migration and where to migrate the selected processes. As scientific contribution, MigBSP++ introduces the solutions to two issues that were missing at MigBSP: (a) the detection of imbalance load when there are more processes than processors, and (b) the definition of how many processes will migrate indeed. On the question (a), a change of the mode of detection of imbalance is proposed, noting the total computation time for each processor. On the second question (b) an algorithm called the Prediction Algorithm BSP (PABSP) is presented. The input data of PABSP are elected process by the PM technique and the output is a list of processes that will, indeed, migrate providing a time reduction of the next super-step. To demonstrate the results of applying this model, two BSP applications have been developed with the assistance of Adaptive Message Passing Interface (AMPI) library. This tool provides a uniform framework that, through the migration process, allows a transparent load balancing to the user. Based on MigBSP and MigBSP++, load balancing strategies have been developed for the performance and comparison among new strategies and among the ones which were already in the system.The results indicate that, in cases where the granularity of the task, the gains in runtime are more evident, reaching up to 46% compared to the application without balancing, and 37% when compared to native strategies AMPI. These numbers suggest that the model MigBSP++ has practical application and can produce satisfactory results.

Les architectures de calcul haute performance de nos jours sont des architectures hiérarchiques et hétérogènes: hiérarchiques car elles sont composées d’une hiérarchie de mémoire, une mémoire distribuée entre les noeuds et une mémoire partagée entre les coeurs d’un même noeud. Hétérogènes due à l’utilisation des processeurs spécifiques appelés Accélérateurs tel que le processeur CellBE d’IBM et les CPUs de NVIDIA. La complexité de maîtrise de ces architectures est double. D’une part, le problème de programmabilité: la programmation doit rester simple, la plus proche possible de la programmation séquentielle classique et indépendante de l’architecture cible. D’autre part, le problème d’efficacité: les performances doivent êtres proches de celles qu’obtiendrait un expert en écrivant le code à la main en utilisant des outils de bas niveau. Dans cette thèse, nous avons proposé une plateforme de développement pour répondre à ces problèmes. Pour cela, nous proposons deux outils : BSP++ est une bibliothèque générique utilisant des templates C++ et BSPGen est un framework permettant la génération automatique de code hybride à plusieurs niveaux de la hiérarchie (MPI+OpenMP ou MPI + Cell BE). Basée sur un modèle hiérarchique, la bibliothèque BSP++ prend les architectures hybrides comme cibles natives. Utilisant un ensemble réduit de primitives et de concepts intuitifs, BSP++ offre une simplicité d'utilisation et un haut niveau d' abstraction de la machine cible. Utilisant le modèle de coût de BSP++, BSPGen estime et génère le code hybride hiérarchique adéquat pour une application donnée sur une architecture cible. BSPGen génère un code hybride à partir d'une liste de fonctions séquentielles et d'une description de l'algorithme parallèle. Nos outils ont été validés sur différentes applications de différents domaines allant de la vérification et du calcul scientifique au traitement d'images en passant par la bioinformatique. En utilisant une large sélection d’architecture cible allant de simple machines à mémoire partagée au machines Petascale en passant par les architectures hétérogènes équipées d’accélérateurs de type Cell BE. / Today’s high-performance computing architectures are hierarchical and heterogeneous. With a hierarchy of memory, they are composed of distributed memory between nodes and shared memory between cores of the same node. heterogeneous due to the use of specific processors called accelerators such as the CellBE IBM processor and/or NVIDIA GPUs. The programming complexity of these architectures is twofold. On the one hand, the problem of programmability: the programming should be simple, as close as possible to the conventional sequential programming and independent of the target architecture. On the other hand, the problem of efficiency: performance should be similar to those obtained by a expert in writing code by hand using low-level tools. In this thesis, we proposed a development platform to address these problems. For this, we propose two tools: BSP++ is a generic library using C++ templates and BSPGen is a framework for the automatic hybrid multi-level hierarchy (MPI + OpenMP or MPI + Cell BE) code generation.Based on a hierarchical model, the BSP++ library takes the hybrid architectures as native targets. Using a small set of primitives and intuitive concepts, BSP++ provides a simple way to use and a high level of abstraction of the target machine. Using the cost model of BSP++, BSPGen predicts and generates the appropriate hierarchical hybrid code for a given application on target architecture. BSPGen generates hybrid code from a sequential list of functions and a description of the parallel algorithm.Our tools have been validated with various applications in different fields ranging from verification to scientific computing and image processing through bioinformatics. Using a wide selection of target architecture ranging from simple shared memory machines to Petascale machines through the heterogeneous architectures equipped with Cell BE accelerators.

Submitted by Maicon Juliano Schmidt (maicons) on 2015-05-29T13:32:27Z No. of bitstreams: 1 Júlio Cezar Santos Pires_.pdf: 1421846 bytes, checksum: 20f1800d942c84a72b6cb04d163cc92f (MD5) / Made available in DSpace on 2015-05-29T13:32:27Z (GMT). No. of bitstreams: 1 Júlio Cezar Santos Pires_.pdf: 1421846 bytes, checksum: 20f1800d942c84a72b6cb04d163cc92f (MD5) Previous issue date: 2014-04-29 / CTIC/RNP - Centro de Pesquisa e Desenvolvimento em Tecnologias Digitais para informação e Comunicação / Com os constantes avanços tecnológicos, surgem novas tendências para prover uma base de serviços para a nova era da tecnologia da informação. Com isso, surgem novos paradigmas para sistemas distribuídos, como, por exemplo, a Computação em Nuvem (ou Cloud Computing), que possui como ideia base a disponibilização de recursos computacionais sob demanda por meio da Internet, permitindo, assim, a sua utilização em qualquer lugar e pelos mais diversos tipos de aplicações. Entre as principais características da computação em nuvem, tem-se a elasticidade, provisionamento de serviço e cobrança baseada na utilização efetiva dos recursos. Visando tornar estas características, na prática, possíveis, torna-se indispensável que a infraestrutura disponha de um sistema de monitoramento. Neste contexto, este trabalho apresenta o BSPMon, um sistema de monitoramento de recursos preditivo para aplicações paralelas em Cloud Computing. Com o objetivo de ter um controle fino sobre os recursos computacionais, o BSPMon coletará métricas de desempenho nos três níveis da infraestrutura: máquina física, máquina virtual e aplicação, efetuando, desta forma, um monitoramento hierárquico multinível dos recursos. De posse destas métricas de desempenho, o BSPMon efetuará predições sobre as demandas, visando melhores resultados para a tomada de decisão em situações de migração, previsão, controle sobre o SLA, provisionamento e consolidação dos recursos. O sistema proposto atuará no nível de middleware, de forma transparente para a aplicação. A partir das avaliações obtidas na predição, os resultados apontam baixa intrusividade na infraestrutura, eficiência energética e predições com taxa de acerto superior a 90%. / Due to constant technological advances, there are new trends to provide a service base for the new era of information technology. Thus, there are new paradigms for distributed systems, for example, Cloud Computing, which has as basic idea of the provision of computational resources on demand via the Internet, thus allowing their use anywhere and for many different types of applications. Among the main features of cloud computing, there is elasticity, service provisioning and billing based on the effective use of resources. In order to make these features in practice possible, it is essential that the infrastructure to have a monitoring system. In this context, this work presents the BSPMon, a monitoring system of predictive features for parallel applications in Cloud Computing. In order to have fine control over computing resources, the BSPMon collect performance metrics in three levels of infrastructure: physical machine, virtual machine and application, making thus a multilevel hierarchical resource monitoring. With such performance metrics, the BSPMon shall make predictions about the demands, to obtain better results for decision making in migration scenarios, prediction, control over the SLA, provisioning and consolidation of resources. The proposed system will operate in the middleware level, transparently to the application. From the evaluations obtained in the prediction, the results indicate low intrusiveness in infrastructure, energy efficiency and predictions accuracy rate above 90 %.

The recent growth in wireless communications presents a new challenge to multimedia communications. Digital image transmission is a very common form of multimedia communication. Due to limited bandwidth and broadcast nature of the wireless medium, it is necessary to compress and encrypt images before they are sent. On the other hand, it is important to efficiently utilize the limited energy in wireless devices. In a wireless device, two major sources of energy consumption are energy used for computation and energy used for transmission. Computation energy can be reduced by minimizing the time spent on compression and encryption. Transmission energy can be reduced by sending a smaller image file that is obtained by compressing the original highest quality image. Image quality is often sacrificed in the compression process. Therefore, users should have the flexibility to control the image quality to determine whether such a tradeoff is acceptable. It is also desirable for users to have control over image quality in different areas of the image so that less important areas can be compressed more, while retaining the details in important areas. To reduce computations for encryption, a partial encryption scheme can be employed to encrypt only the critical parts of an image file, without sacrificing security. This thesis proposes a scalable and secure image representation scheme that allows users to select different image quality and security levels. The binary space partitioning (BSP) tree presentation is selected because this representation allows convenient compression and scalable encryption. The Advanced Encryption Standard (AES) is chosen as the encryption algorithm because it is fast and secure. Our experimental result shows that our new tree construction method and our pruning formula reduces execution time, hence computation energy, by about 90%. Our image quality prediction model accurately predicts image quality to within 2-3dB of the actual image PSNR.

A series of numerical modeling studies were conducted to characterize the mass flow rates in balcony spill plumes (BSP), a type of buoyant fire plume occurring in atria. The variation of BSP mass flow rate as a function of elevation, fire size and fire compartment geometry was examined both numerically and experimentally. A new method for estimation of BSP mass flow rates, appropriate for design of smoke management systems in high-elevation atria, was developed based on simulations of BSP mass flow rate. An experimental program conducted in a 12 m high atrium measured BSP mass flow rates as well as temperatures in the fire compartment and atrium. This data was used to evaluate CFD models of the fire compartment and atrium in the experimental facility. These were implemented using the Fire Dynamics Simulator (FDS) software. The models were extended to investigate BSP behaviour at elevations up to 50 m. The removal of atrium walls in the model to allow free development of the BSP is a unique approach among published numerical modeling studies of BSP behaviour. The high-elevation CFD model was used to perform a parametric study of BSP mass flow rate as a function of elevation, fire size and fire compartment geometry. Predictions of BSP mass flow rate from this study extend to 50 m above the atrium floor, extending the range of elevations represented in the published experimental data (<= 9 m). Data from the parametric study was used to develop a new method for estimation of BSP mass flow rates at high elevations. BSP mass flow rates estimated using the new method are shown to be bounded by values estimated using existing methods based on low-elevation experimental data.

The recent growth in wireless communications presents a new challenge to multimedia communications. Digital image transmission is a very common form of multimedia communication. Due to limited bandwidth and broadcast nature of the wireless medium, it is necessary to compress and encrypt images before they are sent. On the other hand, it is important to efficiently utilize the limited energy in wireless devices. In a wireless device, two major sources of energy consumption are energy used for computation and energy used for transmission. Computation energy can be reduced by minimizing the time spent on compression and encryption. Transmission energy can be reduced by sending a smaller image file that is obtained by compressing the original highest quality image. Image quality is often sacrificed in the compression process. Therefore, users should have the flexibility to control the image quality to determine whether such a tradeoff is acceptable. It is also desirable for users to have control over image quality in different areas of the image so that less important areas can be compressed more, while retaining the details in important areas. To reduce computations for encryption, a partial encryption scheme can be employed to encrypt only the critical parts of an image file, without sacrificing security. This thesis proposes a scalable and secure image representation scheme that allows users to select different image quality and security levels. The binary space partitioning (BSP) tree presentation is selected because this representation allows convenient compression and scalable encryption. The Advanced Encryption Standard (AES) is chosen as the encryption algorithm because it is fast and secure. Our experimental result shows that our new tree construction method and our pruning formula reduces execution time, hence computation energy, by about 90%. Our image quality prediction model accurately predicts image quality to within 2-3dB of the actual image PSNR.

A series of numerical modeling studies were conducted to characterize the mass flow rates in balcony spill plumes (BSP), a type of buoyant fire plume occurring in atria. The variation of BSP mass flow rate as a function of elevation, fire size and fire compartment geometry was examined both numerically and experimentally. A new method for estimation of BSP mass flow rates, appropriate for design of smoke management systems in high-elevation atria, was developed based on simulations of BSP mass flow rate. An experimental program conducted in a 12 m high atrium measured BSP mass flow rates as well as temperatures in the fire compartment and atrium. This data was used to evaluate CFD models of the fire compartment and atrium in the experimental facility. These were implemented using the Fire Dynamics Simulator (FDS) software. The models were extended to investigate BSP behaviour at elevations up to 50 m. The removal of atrium walls in the model to allow free development of the BSP is a unique approach among published numerical modeling studies of BSP behaviour. The high-elevation CFD model was used to perform a parametric study of BSP mass flow rate as a function of elevation, fire size and fire compartment geometry. Predictions of BSP mass flow rate from this study extend to 50 m above the atrium floor, extending the range of elevations represented in the published experimental data (<= 9 m). Data from the parametric study was used to develop a new method for estimation of BSP mass flow rates at high elevations. BSP mass flow rates estimated using the new method are shown to be bounded by values estimated using existing methods based on low-elevation experimental data.

More and more embedded systems choose ARM-based micro-controllers as CPU. If no embedded OS built with the system, the application scope will be restricted. Therefore, the need of embedded OS is vital. There are many embedded OS¡¦s in the market, but the embedded Linux has many advantages and is widely accepted. Commercial embedded Linux takes less refund than other embedded OS¡¦s. The kernel and most applications are distributed in GPL open source copyright, and is highly portable to many machine platforms. Presently, the hardware key-technology is highly skilled. The margin of 3C industrial has gone down rapidly. Therefore, people focus on adapting integrated technology to practicality and innovation to make cost down. Developers choose appropriate ARM micro-controllers according to demanding functionality of their products. The microcontroller is not necessary running with Linux distribution. Two approaches can be used to resolve the embedded OS issue. The first approach is porting Linux to the platform without any refund. The second approach is to pay for commercial Linux. Embedded system peripheral devices aim at powerful functionalities and economy. For instance, UART interface is cheap and low data transfer rate. The target board communicates with host via RS-232. RS-232 acts as serial console to play dumb terminal under Linux. Industrial applications often make use of RS-xxx for UART physical transmission layer. For instance, RS-485 applies modbus protocol to build cheap monitor systems. Network transmission is a necessary function, and it generally achieves high data transfer rate application through Ethernet. The UNIX-like network socket has served network application very well. Embedded systems are usually diskless systems. In order to keep permanent data, using flash memory as block disk system is a widely adapted strategy and which operates flash memory through MTD subsystems¡][28]¡^. An MTD subsystem contains two different modules, ¡§user¡¨and ¡§driver¡¨. In the driver module, CFI¡][40]¡^ is applied to probe flash chip, partition it and provide operating function. Flash translation layer and file-system are applied in the user module. MTD BLOCK is used to emulate the flash partitions as block devices which are then mounted into Linux virtual file system¡]VFS¡^with JFFS2 type, designed according to the feature of flash devices. In this thesis, we will describe in detail the procedure of porting Linux to ARM micro-controllers. The motivation of the work is introduced in chapter 1. In chapter 2, we introduce development tools and the main flow of the porting procedure. In chapter 3, we describe the LH79525 platform and the main perepherals on the target board, then introduce the ARM programmer model. In chapter 4, we examine the required knowledge and the important issues for porting ARM Linux. In chapter 5, we describe the details of porting Linux to run with Sharp LH79525, including modifying the key source codes and adjusting kernel configuration for embedding the UART, ethernet MAC, and MTD subsystem. In chapter 6, we do step-by-step validation and apply an integrated application with the LF-314CP temperature controller¡][46]¡^ by law-chain technology for the LH79525 target board running with the ported ARM Linux. In chapter 7, we present some issues for future work and improvement, then make a conclusion for the thesis.

近年來，有關Blog的話題出現在大大小小的報導裡，由於Blog發跡於網路，有關於網路事業的活動也因此受到影響。原本只是作為個人心情日誌的Blog，如今變成商業活動中不可或缺的工具。本研究探討在不同商業經營模式下，Blog特質所能獲取的獲利情形，並訪談國內BSP經營業者及部落客（Blogger）對於Blog商業化的看法及經營情形；並藉由4C理論架構探究三種商業模式之內涵。 研究結果得知，要經營一個具有高人氣的Blog需要較早投入部落圈、有良好的寫作內容、與網友有良好互動、及定期更新。類似入口網站的BSP站台主要獲利來源可藉由收取會員加值費、與企業進行行銷專案及收取廣告費用；專職寫作的部落客（Blogger）未來可朝向收取閱讀費用為獲利來源；企業自行架設的Blog可因此累積企業知名度、宣傳產品。 本研究以商業模式為論點探討Blog的商業行為，提出這個新的網路力量之經營模式，建議後續研究者可針對Blog特性及發展探討更多元的應用議題，如Blog的分眾性、Blog專業性等。