Synthesis of FPGA-based accelerators implementing recursive algorithms

O desenvolvimento de sistemas computacionais é um processo complexo, com múltiplas etapas, que requer uma análise profunda do problema, levando em consideração as limitações e os requisitos aplicáveis. Tal tarefa envolve a exploração de técnicas alternativas e de algoritmos computacionais para optimizar o sistema e satisfazer os requisitos estabelecidos. Neste contexto, uma das mais importantes etapas é a análise e implementação de algoritmos computacionais. Enormes avanços tecnológicos no âmbito das FPGAs (Field-Programmable Gate Arrays) tornaram possível o desenvolvimento de sistemas de engenharia extremamente complexos. Contudo, o número de transístores disponíveis por chip está a crescer mais rapidamente do que a capacidade que temos para desenvolver sistemas que tirem proveito desse crescimento. Esta limitação já bem conhecida, antes de se revelar com FPGAs, já se verificava com ASICs (Application-Specific Integrated Circuits) e tem vindo a aumentar continuamente. O desenvolvimento de sistemas com base em FPGAs de alta capacidade envolve uma grande variedade de ferramentas, incluindo métodos para a implementação eficiente de algoritmos computacionais. Esta tese pretende proporcionar uma contribuição nesta área, tirando partido da reutilização, do aumento do nível de abstracção e de especificações algorítmicas mais automatizadas e claras. Mais especificamente, é apresentado um estudo que foi levado a cabo no sentido de obter critérios relativos à implementação em hardware de algoritmos recursivos versus iterativos. Depois de serem apresentadas algumas das estratégias para implementar recursividade em hardware mais significativas, descreve-se, em pormenor, um conjunto de algoritmos para resolver problemas de pesquisa combinatória (considerados enquanto exemplos de aplicação). Versões recursivas e iterativas destes algoritmos foram implementados e testados em FPGA. Com base nos resultados obtidos, é feita uma cuidada análise comparativa. Novas ferramentas e técnicas de investigação que foram desenvolvidas no âmbito desta tese são também discutidas e demonstradas.

A generalization of the Hoffman-Lovász upper bound on the independence number of a regular graph

A family of quadratic programming problems whose optimal values are upper bounds on the independence number of a graph is introduced. Among this family, the quadratic programming problem which gives the best upper bound is identified. Also the proof that the upper bound introduced by Hoffman and Lovász for regular graphs is a particular case of this family is given. In addition, some new results characterizing the class of graphs for which the independence number attains the optimal value of the above best upper bound are given. Finally a polynomial-time algorithm for approximating the size of the maximum independent set of an arbitrary graph is described and the computational experiments carried out on 36 DIMACS clique benchmark instances are reported.

Politopo de Birkhoff acíclico

Neste trabalho estabelece-se uma interpreta c~ao geom etrica, em termos da teoria dos grafos, para v ertices, arestas e faces de uma qualquer dimens~ao do politopo de Birkho ac clico, Tn = n(T), onde T e uma arvore com n v ertices. Generaliza-se o resultado obtido por G. Dahl, [18], para o c alculo do di^ametro do grafo G( t n), onde t n e o politopo das matrizes tridiagonais duplamente estoc asticas. Adicionalmente, para q = 0; 1; 2; 3 s~ao obtidas f ormulas expl citas para a contagem do n umero de q faces do politopo de Birkho tridiagonal, t n, e e feito o estudo da natureza geom etrica dessas mesmas faces. S~ao, tamb em, apresentados algoritmos para efectuar contagens do n umero de faces de dimens~ao inferior a de uma dada face do politopo de Birkho ac clico.

Optimisation of viscoelastic treatments using genetic algorithms

Viscoelastic treatments are one of the most efficient treatments, as far as passive damping is concerned, particularly in the case of thin and light structures. In this type of treatment, part of the strain energy generated in the viscoelastic material is dissipated to the surroundings, in the form of heat. A layer of viscoelastic material is applied to a structure in an unconstrained or constrained configuration, the latter proving to be the most efficient arrangement. This is due to the fact that the relative movement of both the host and constraining layers cause the viscoelastic material to be subjected to a relatively high strain energy. There are studies, however, that claim that the partial application of the viscoelastic material is just as efficient, in terms of economic costs or any other form of treatment application costs. The application of patches of material in specific and selected areas of the structure, thus minimising the extension of damping material, results in an equally efficient treatment. Since the damping mechanism of a viscoelastic material is based on the dissipation of part of the strain energy, the efficiency of the partial treatment can be correlated to the modal strain energy of the structure. Even though the results obtained with this approach in various studies are considered very satisfactory, an optimisation procedure is deemed necessary. In order to obtain optimum solutions, however, time consuming numerical simulations are required. The optimisation process to use the minimum amount of viscoelastic material is based on an evolutionary geometry re-design and calculation of the modal damping, making this procedure computationally costly. To avert this disadvantage, this study uses adaptive layerwise finite elements and applies Genetic Algorithms in the optimisation process.

Necessary and sufficient conditions for a Hamiltonian graph

A graph is singular if the zero eigenvalue is in the spectrum of its 0-1 adjacency matrix A. If an eigenvector belonging to the zero eigenspace of A has no zero entries, then the singular graph is said to be a core graph. A ( k,t)-regular set is a subset of the vertices inducing a k -regular subgraph such that every vertex not in the subset has t neighbours in it. We consider the case when k=t which relates to the eigenvalue zero under certain conditions. We show that if a regular graph has a ( k,k )-regular set, then it is a core graph. By considering the walk matrix we develop an algorithm to extract ( k,k )-regular sets and formulate a necessary and sufficient condition for a graph to be Hamiltonian.

Spectra of graphs obtained by a generalization of the join graph operation

Taking a Fiedler’s result on the spectrum of a matrix formed from two symmetric matrices as a motivation, a more general result is deduced and applied to the determination of adjacency and Laplacian spectra of graphs obtained by a generalized join graph operation on families of graphs (regular in the case of adjacency spectra and arbitrary in the case of Laplacian spectra). Some additional consequences are explored, namely regarding the largest eigenvalue and algebraic connectivity.