Makespan computation for GPU threads running on a single streaming multiprocessor

Graphics processors were originally developed for rendering graphics but have recently evolved towards being an architecture for general-purpose computations. They are also expected to become important parts of embedded systems hardware -- not just for graphics. However, this necessitates the development of appropriate timing analysis techniques which would be required because techniques developed for CPU scheduling are not applicable. The reason is that we are not interested in how long it takes for any given GPU thread to complete, but rather how long it takes for all of them to complete. We therefore develop a simple method for finding an upper bound on the makespan of a group of GPU threads executing the same program and competing for the resources of a single streaming multiprocessor (whose architecture is based on NVIDIA Fermi, with some simplifying assunptions). We then build upon this method to formulate the derivation of the exact worst-case makespan (and corresponding schedule) as an optimization problem. Addressing the issue of tractability, we also present a technique for efficiently computing a safe estimate of the worstcase makespan with minimal pessimism, which may be used when finding an exact value would take too long.

Efficient computation of min and max sensor values in multihop networks

Consider a wireless sensor network (WSN) where a broadcast from a sensor node does not reach all sensor nodes in the network; such networks are often called multihop networks. Sensor nodes take individual sensor readings, however, in many cases, it is relevant to compute aggregated quantities of these readings. In fact, the minimum and maximum of all sensor readings at an instant are often interesting because they indicate abnormal behavior, for example if the maximum temperature is very high then it may be that a fire has broken out. In this context, we propose an algorithm for computing the min or max of sensor readings in a multihop network. This algorithm has the particularly interesting property of having a time complexity that does not depend on the number of sensor nodes; only the network diameter and the range of the value domain of sensor readings matter.

Music and evolutionary computation

This paper presents a brief history of the western music: from its genesis to serialism and the Darmstadt school. Also some mathematical aspects of music are then presented and confronted with music as a form of art. The question is, are these two distinct aspects compatible? Can computers be of real help in automatic composition? The more appealing algorithmic approach is evolutionary computation as it offers creativity potential. Therefore, the Evolutionary Algorithms are then introduced and some results of GAs and GPs application to music generation are analysed.

On the numerical computation of the Mittag-Leffler function

Recently simple limiting functions establishing upper and lower bounds on the Mittag-Leffler function were found. This paper follows those expressions to design an efficient algorithm for the approximate calculation of expressions usual in fractional-order control systems. The numerical experiments demonstrate the superior efficiency of the proposed method.

On the numerical computation of the Mittag-Leffler function

Recently simple limiting functions establishing upper and lower bounds on the Mittag-Leffler function were found. This paper follows those expressions to design an efficient algorithm for the approximate calculation of expressions usual in fractional-order control systems. The numerical experiments demonstrate the superior efficiency of the proposed method.

A congestion management and transmission price simulator for competitive electricity markets

In a liberalized electricity market, the Transmission System Operator (TSO) plays a crucial role in power system operation. Among many other tasks, TSO detects congestion situations and allocates the payments of electricity transmission. This paper presents a software tool for congestion management and transmission price determination in electricity markets. The congestion management is based on a reformulated Optimal Power Flow (OPF), whose main goal is to obtain a feasible solution for the re-dispatch minimizing the changes in the dispatch proposed by the market operator. The transmission price computation considers the physical impact caused by the market agents in the transmission network. The final tariff includes existing system costs and also costs due to the initial congestion situation and losses costs. The paper includes a case study for the IEEE 30 bus power system.

LMP based bid formation for virtual power players operating in smart grids

Power system organization has gone through huge changes in the recent years. Significant increase in distributed generation (DG) and operation in the scope of liberalized markets are two relevant driving forces for these changes. More recently, the smart grid (SG) concept gained increased importance, and is being seen as a paradigm able to support power system requirements for the future. This paper proposes a computational architecture to support day-ahead Virtual Power Player (VPP) bid formation in the smart grid context. This architecture includes a forecasting module, a resource optimization and Locational Marginal Price (LMP) computation module, and a bid formation module. Due to the involved problems characteristics, the implementation of this architecture requires the use of Artificial Intelligence (AI) techniques. Artificial Neural Networks (ANN) are used for resource and load forecasting and Evolutionary Particle Swarm Optimization (EPSO) is used for energy resource scheduling. The paper presents a case study that considers a 33 bus distribution network that includes 67 distributed generators, 32 loads and 9 storage units.

Computational intelligence applications for future power systems

Power system planning, control and operation require an adequate use of existing resources as to increase system efficiency. The use of optimal solutions in power systems allows huge savings stressing the need of adequate optimization and control methods. These must be able to solve the envisaged optimization problems in time scales compatible with operational requirements. Power systems are complex, uncertain and changing environments that make the use of traditional optimization methodologies impracticable in most real situations. Computational intelligence methods present good characteristics to address this kind of problems and have already proved to be efficient for very diverse power system optimization problems. Evolutionary computation, fuzzy systems, swarm intelligence, artificial immune systems, neural networks, and hybrid approaches are presently seen as the most adequate methodologies to address several planning, control and operation problems in power systems. Future power systems, with intensive use of distributed generation and electricity market liberalization increase power systems complexity and bring huge challenges to the forefront of the power industry. Decentralized intelligence and decision making requires more effective optimization and control techniques techniques so that the involved players can make the most adequate use of existing resources in the new context. The application of computational intelligence methods to deal with several problems of future power systems is presented in this chapter. Four different applications are presented to illustrate the promises of computational intelligence, and illustrate their potentials.

Optimal intelligent supervisory control system in cyber-physical intelligence

This paper studies Optimal Intelligent Supervisory Control System (OISCS) model for the design of control systems which can work in the presence of cyber-physical elements with privacy protection. The development of such architecture has the possibility of providing new ways of integrated control into systems where large amounts of fast computation are not easily available, either due to limitations on power, physical size or choice of computing elements.

Efficient implementation of a dominance protocol for wireless medium access

Mestrado em Engenharia Electrotécnica e de Computadores

Algoritmos genéticos: aplicação na síntese de alguns algoritmos de controlo

Mestrado em Engenharia Electrotécnica e de Computadores. Área de Especialização de Automação e Sistemas.

eTourismGUI - Adaptative Interface

Mestrado em Engenharia Informática. Área de Especialização em Tecnologias do Conhecimento e Decisão.

Medição automática do diâmetro da aorta torácica em imagens de tomografia computorizada

O calibre da artéria aorta torácica é avaliado em situações de suspeita de patologia ou em pacientes com pre-disposição para desenvolverem doenças vasculares. A medição das suas dimensões, em duas direcções diametralmente opostas e, assim, fulcral na avaliação desta estrutura. Para tal, o exame de primeira linha definido é a Angiografia por Tomografia Computorizada (Angio-TC), injectando-se um produto de contraste na veia radial que irá opacificar os vasos, permitindo a sua distinção das estruturas adjacentes. O presente trabalho, inserido na disciplina de Dissertação/ Projecto/ Estágio Profissional do Mestrado em Engenharia de Computação e Instrumentação Médica e com a cooperação da empresa Efficientia, foi sugerido pela equipa de Angio-TC do Centro Hospitalar de Vila Nova de Gaia (CHVNG) e tem por objectivo o desenvolvimento de uma aplicação para a medição e registo automático do diâmetro da aorta torácica em nove pontos anatómicos pre-definidos em imagens de Tomografia Computorizada (TC). A aplicação foi desenvolvida no ambiente integrado de processamento e análise de imagem, Fiji, sendo a metodologia composta pelas etapas de segmentação, desenho da linha central, determinação dos planos de cortes, segmentação e medição da aorta nos planos de corte. Os resultados obtidos pela metodologia proposta são concordantes com os obtidos por especialistas para o conjunto de teste utilizado neste trabalho.

Code offloading on real-time multimedia systems: a framework for handling code mobility and code offloading in a QoS aware environment

Actualmente, os smartphones e outros dispositivos móveis têm vindo a ser dotados com cada vez maior poder computacional, sendo capazes de executar um vasto conjunto de aplicações desde simples programas de para tirar notas até sofisticados programas de navegação. Porém, mesmo com a evolução do seu hardware, os actuais dispositivos móveis ainda não possuem as mesmas capacidades que os computadores de mesa ou portáteis. Uma possível solução para este problema é distribuir a aplicação, executando partes dela no dispositivo local e o resto em outros dispositivos ligados à rede. Adicionalmente, alguns tipos de aplicações como aplicações multimédia, jogos electrónicos ou aplicações de ambiente imersivos possuem requisitos em termos de Qualidade de Serviço, particularmente de tempo real. Ao longo desta tese é proposto um sistema de execução de código remota para sistemas distribuídos com restrições de tempo-real. A arquitectura proposta adapta-se a sistemas que necessitem de executar periodicamente e em paralelo mesmo conjunto de funções com garantias de tempo real, mesmo desconhecendo os tempos de execução das referidas funções. A plataforma proposta foi desenvolvida para sistemas móveis capazes de executar o Sistema Operativo Android.

Design de cenários de aprendizagem

A cada instante surgem novas soluções de aprendizagem, resultado da evolução tecnológica constante com que nos deparamos. Estas inovações potenciam uma transmissão do conhecimento entre o educador e o educando cada vez mais simplificada, rápida e eficiente. Alguns destes avanços têm em vista a centralização no aluno, através da delegação de tarefas e da disponibilização de conteúdos, investindo na autonomia e na auto-aprendizagem, de modo a que cada aluno crie o seu próprio método de estudo, e evolua gradualmente, com o acompanhamento de um professor ou sistema autónomo de aprendizagem. Com esta investigação, é pretendido fazer um estudo dos métodos de aprendizagem ao longo do tempo até à actualidade, enumerando algumas das ferramentas utilizadas no processo de aprendizagem, indicando os vários benefícios, bem como contrapartidas do uso das mesmas. Será também analisado um caso de estudo baseado numa destas ferramentas, descrevendo o seu funcionamento e modo de interacção entre as várias entidades participantes, apresentando os resultados obtidos. O caso de estudo consistirá na criação de um cenário específico de aprendizagem, na área da saúde, analisando-o em diferentes contextos, e evidenciando as características e benefícios de cada ambiente analisado, no processo aprendizagem. Será então demonstrado como é possível optimizar os processos de aprendizagem, utilizando ferramentas de informatização e automatização desses mesmos processos, de forma tornar o processo de ensino mais célere e eficaz, num ambiente controlável, e com as funcionalidades que a tecnologia actual permite.