As complementary metal-oxide semiconductor (CMOS) technology faces more and more severe physical barriers down the path of continuously feature size scaling, innovative nano-scale devices and other post-CMOS technologies have been developed to enhance future circuit design and computation. These nanotechnologies have shown promising potentials to achieve magnitude improvement in performance and integration density. The substitution of CMOS transistors with nano-devices is expected to not only continue along the exponential projection of Moore's Law, but also raise significant challenges and opportunities, especially in the field of electronic design automation. The major obstacles that the designers are experiencing with emerging nanotechnology design include: i) the existing computer-aided design (CAD) approaches in the context of conventional CMOS Boolean design cannot be directly employed in the nanoelectronic design process, because the intrinsic electrical characteristics of many nano-devices are not best suited for Boolean implementations but demonstrate strong capability for implementing non-conventional logic such as threshold logic and reversible logic; ii) due to the density and size factors of nano-devices, the defect rate of nanoelectronic system is much higher than conventional CMOS systems, therefore existing design paradigms cannot guarantee design quality and lead to even worse result in high failure ratio. Motivated by the compelling potentials and design challenges of emerging post-CMOS technologies, this dissertation work focuses on fundamental design methodologies to effectively and efficiently achieve high quality nanoscale design. A novel programmable logic element (PLE) is first proposed to explore the versatile functionalities of threshold gates (TGs) and multi-threshold threshold gates (MTTGs). This PLE structure can realize all three- or four-variable logic functions through configuring binary control bits. This is the first single threshold logic structure that provides complete Boolean logic implementation. Based on the PLEs, a reconfigurable architecture is constructed to offer dynamic reconfigurability with little or no reconfiguration overhead, due to the intrinsic self-latching property of nanopipelining. Our reconfiguration data generation algorithm can further reduce the reconfiguration cost. To fully take advantage of such threshold logic design using emerging nanotechnologies, we also developed a combinational equivalence checking (CEC) framework for threshold logic design. Based on the features of threshold logic gates and circuits, different techniques of formulating a given threshold logic in conjunctive normal form (CNF) are introduced to facilitate efficient SAT-based verification. Evaluated with mainstream benchmarks, our hybrid algorithm, which takes into account both input symmetry and input weight order of threshold gates, can efficiently generate CNF formulas in terms of both SAT solving time and CNF generating time. Then the reversible logic synthesis problem is considered as we focus on efficient synthesis heuristics which can provide high quality synthesis results within a reasonable computation time. We have developed a weighted directed graph model for function representation and complexity measurement. An atomic transformation is constructed to associate the function complexity variation with reversible gates. The efficiency of our heuristic lies in maximally decreasing the function complexity during synthesis steps as well as the capability to climb out of local optimums. Thereafter, swarm intelligence, one of the machine learning techniques is employed in the space searching for reversible logic synthesis, which achieves further performance improvement. To tackle the high defect-rate during the emerging nanotechnology manufacturing process, we have developed a novel defect-aware logic mapping framework for nanowire-based PLA architecture via Boolean satisfiability (SAT). The PLA defects of various types are formulated as covering and closure constraints. The defect-aware logic mapping is then solved efficiently by using available SAT solvers. This approach can generate valid logic mapping with a defect rate as high as 20%. The proposed method is universally suitable for various nanoscale PLAs, including AND/OR, NOR/NOR structures, etc. In summary, this work provides some initial attempts to address two major problems confronting future nanoelectronic system designs: the development of electronic design automation tools and the reliability issues. However, there are still a lot of challenging open questions remain in this emerging and promising area. We hope our work can lay down stepstones on nano-scale circuit design optimization through exploiting the distinctive characteristics of emerging nanotechnologies. / Ph. D.

This thesis presents a new ant-optimized biclustering technique known as SNAP biclustering, which runs faster and produces results of superior quality to previous techniques. Biclustering techniques have been designed to compensate for the weaknesses of classical clustering algorithms by allowing cluster overlap, and allowing vectors to be grouped for a subset of their defined features. These techniques have performed well in many problem domains, particularly DNA microarray analysis and collaborative filtering. A motivation for this work has been the biclustering technique known as bicACO, which was the first to use ant colony optimization. As bicACO is time intensive, much emphasis was placed on decreasing SNAP's runtime. The superior speed and biclustering results of SNAP are due to its improved initialization and solution construction procedures. In experimental studies involving the Yeast Cell Cycle DNA microarray dataset and the MovieLens collaborative filtering dataset, SNAP has run at least 22 times faster than bicACO while generating superior results. Thus, SNAP is an effective choice of technique for microarray analysis and collaborative filtering applications. / Master of Science

Roteamento por colônia de formigas é um método de roteamento em redes de comunicação, e diversos algoritmos foram propostos nos últimos anos baseado nessa estrutura. Todos esses algoritmos produzem excelentes resultados, provando a sua eficiência e eficácia. Este trabalho apresenta os resultados de desempenho dos principais algoritmos encontrados na literatura, e com base nesses resultados, propõe um novo algoritmo com desempenho equivalente e com uma complexidade computacional menor. O trabalho é focalizado em redes tipo datagrama com topologia irregular, descrevendo suas propriedades e características e realizando uma análise e comparação de seus desempenhos em um ambiente de simulação. / Ant Colony Routing is an adaptive method for routing in communication networks, and several algorithms have been proposed in the last years based on this framework. All these algorithms show excellent results, proving their efficiency and efficacy. This work presents the results of the performance of the main algorithms found in the literature, and based on these results, it proposes a novel algorithm that has a similar performance but with a lower computational complexity. The work is focused in datagram like networks with irregular topology, describing its characteristics and properties. The performances in an simulation environment are analysed and compared.

Um desafio recorrente em processos produtivos é o desenvolvimento de sistemas de monitoração e diagnóstico. Esses sistemas ajudam na detecção de mudanças inesperadas e interrupções, prevenindo perdas e mitigando riscos. Redes Neurais Artificiais (RNA) têm sido largamente utilizadas na criação de sistemas de monitoração. Normalmente as RNA utilizadas para resolver este tipo de problema são criadas levando-se em conta apenas parâmetros como o número de entradas, saídas e quantidade de neurônios nas camadas escondidas. Assim, as redes resultantes geralmente possuem uma configuração onde há uma total conexão entre os neurônios de uma camada e os da camada seguinte, sem que haja melhorias em sua topologia. Este trabalho utiliza o algoritmo de Otimização por Colônia de Formigas (OCF) para criar redes neurais otimizadas. O algoritmo de busca OCF utiliza a técnica de retropropagação de erros para otimizar a topologia da rede neural sugerindo as melhores conexões entre os neurônios. A RNA resultante foi aplicada para monitorar variáveis do reator de pesquisas IEA-R1 do IPEN. Os resultados obtidos mostram que o algoritmo desenvolvido é capaz de melhorar o desempenho do modelo que estima o valor de variáveis do reator. Em testes com diferentes números de neurônios na camada escondida, utilizando como comparativos o erro quadrático médio, o erro absoluto médio e o coeficiente de correlação, o desempenho da RNA otimizada foi igual ou superior ao da tradicional. / A recurring challenge in production processes is the development of monitoring and diagnosis systems. Those systems help on detecting unexpected changes and interruptions, preventing losses and mitigating risks. Artificial Neural Networks (ANN) have been extensively used in creating monitoring systems. Usually the ANN used to solve this kind of problem are created by taking into account only parameters as the number of inputs, outputs, and number of neurons in the hidden layers. This way, the result networks are generally fully connected and have no improvements in its topology. This work uses an Ant Colony Optimization (ACO) algorithm to create a tuned neural networks. The ACO search algorithm uses Back Error Propagation (BP) to optimize the network topology by suggesting the best neuron connections. The outcome ANN was applied to monitoring the IEA-R1 research reactor at IPEN. The results show that the algorithm is able to improve the performance of the model which estimates the values of the reactor variables. In tests with different numbers of neurons in the hidden layer, using as comparison the mean squared error, the mean absolute error, and the correlation coefficient, the performance of the optimized ANN proved equal or better than the equivalent traditional neural networks.

We present two Swarm Intelligence based approaches for data clustering. The first algorithm, Fuzzy Ants, presented in this thesis clusters data without the initial knowledge of the number of clusters. It is a two stage algorithm. In the first stage the ants cluster data to initially create raw clusters which are refined using the Fuzzy C Means algorithm. Initially, the ants move the individual objects to form heaps. The centroids of these heaps are redefined by the Fuzzy C Means algorithm. In the second stage the objects obtained from the Fuzzy C Means algorithm are hardened according to the maximum membership criteria to form new heaps. These new heaps are then moved by the ants. The final clusters formed are refined by using the Fuzzy C Means algorithm. Results from experiments with 13 datasets show that the partitions produced are competitive with those from FCM. The second algorithm, Fuzzy ant clustering with centroids, is also a two stage algorithm, it requires an initial knowledge of the number of clusters in the data. In the first stage of the algorithm ants move the cluster centers in feature space. The cluster centers found by the ants are evaluated using a reformulated Fuzzy C Means criterion. In the second stage the best cluster centers found are used as the initial cluster centers for the Fuzzy C Means algorithm. Results on 18 datasets show that the partitions found by FCM using the ant initialization are better than those from randomly initialized FCM. Hard C Means was also used in the second stage and the partitions from the ant algorithm are better than from randomly initialized Hard C Means. The Fuzzy Ants algorithm is a novel method to find the number of clusters in the data and also provides good initializations for the FCM and HCM algorithms. We performed sensitivity analysis on the controlling parameters and found the Fuzzy Ants algorithm to be very sensitive to the Tcreateforheap parameter. The FCM and HCM algorithms, with random initializations can get stuck in a bad extrema, the Fuzzy ant clustering with centroids algorithm successfully avoids these bad extremas.

Opposition-based learning (OBL) was recently proposed to extend di erent machine learning algorithms. The main idea of OBL is to consider opposite estimates, actions or states as an attempt to increase the coverage of the solution space and to reduce exploration time. OBL has already been applied to reinforcement learning, neural networks and genetic algorithms. This thesis explores the application of OBL to ant algorithms. Ant algorithms are based on the trail laying and following behaviour of ants. They have been successfully applied to many complex optimization problems. However, like any other technique, they can benefit from performance improvements. Thus, this work was motivated by the idea of developing more complex pheromone and path selection behaviour for the algorithm using the concept of opposition. This work proposes opposition-based extensions to the construction and update phases of the ant algorithm. The modifications that focus on the solution construction include three direct and two indirect methods. The three direct methods work by pairing the ants and synchronizing their path selection. The two other approaches modify the decisions of the ants by using opposite-pheromone content. The extension of the update phase lead to an approach that performs additional pheromone updates using opposite decisions. Experimental validation was done using two versions of the ant algorithm: the Ant System and the Ant Colony System. The di erent OBL extensions were applied to the Travelling Salesman Problem (TSP) and to the Grid World Problem (GWP). Results demonstrate that the concept of opposition is not easily applied to the ant algorithm. One pheromone-based method showed performance improvements that were statistically significant for the TSP. The quality of the solutions increased and more optimal solutions were found. The extension to the update phase showed some improvements for the TSP and led to accuracy improvements and a significant speed-up for the GWP. The other extensions showed no clear improvement. The proposed methods for applying opposition to the ant algorithm have potential, but more investigations are required before ant colony optimization can fully benefit from opposition. Most importantly, fundamental theoretical work with graphs, specifically, clearly defining opposite paths or opposite path components, is needed. Overall, the results indicate that OBL ideas can be beneficial for ant algorithms.

Opposition-based learning (OBL) was recently proposed to extend di erent machine learning algorithms. The main idea of OBL is to consider opposite estimates, actions or states as an attempt to increase the coverage of the solution space and to reduce exploration time. OBL has already been applied to reinforcement learning, neural networks and genetic algorithms. This thesis explores the application of OBL to ant algorithms. Ant algorithms are based on the trail laying and following behaviour of ants. They have been successfully applied to many complex optimization problems. However, like any other technique, they can benefit from performance improvements. Thus, this work was motivated by the idea of developing more complex pheromone and path selection behaviour for the algorithm using the concept of opposition. This work proposes opposition-based extensions to the construction and update phases of the ant algorithm. The modifications that focus on the solution construction include three direct and two indirect methods. The three direct methods work by pairing the ants and synchronizing their path selection. The two other approaches modify the decisions of the ants by using opposite-pheromone content. The extension of the update phase lead to an approach that performs additional pheromone updates using opposite decisions. Experimental validation was done using two versions of the ant algorithm: the Ant System and the Ant Colony System. The di erent OBL extensions were applied to the Travelling Salesman Problem (TSP) and to the Grid World Problem (GWP). Results demonstrate that the concept of opposition is not easily applied to the ant algorithm. One pheromone-based method showed performance improvements that were statistically significant for the TSP. The quality of the solutions increased and more optimal solutions were found. The extension to the update phase showed some improvements for the TSP and led to accuracy improvements and a significant speed-up for the GWP. The other extensions showed no clear improvement. The proposed methods for applying opposition to the ant algorithm have potential, but more investigations are required before ant colony optimization can fully benefit from opposition. Most importantly, fundamental theoretical work with graphs, specifically, clearly defining opposite paths or opposite path components, is needed. Overall, the results indicate that OBL ideas can be beneficial for ant algorithms.

In this dissertation, we consider the clustering problem in data sets with unknown number of clusters having arbitrary shapes, intracluster and intercluster density variations. We introduce a clustering methodology which is composed of three methods that ensures extraction of local density and connectivity properties, data set reduction, and clustering. The first method constructs a unique neighborhood for each data point using the connectivity and density relations among the points based upon the graph theoretical concepts, mainly Gabriel Graphs. Neighborhoods subsequently connected form subclusters (closures) which constitute the skeleton of the clusters. In the second method, the external shape concept in computational geometry is adapted for data set reduction and cluster visualization. This method extracts the external shape of a non-convex n-dimensional data set using Delaunay triangulation. In the third method, we inquire the applicability of Swarm Intelligence to clustering using Ant Colony Optimization (ACO). Ants explore the data set so that the clusters are detected using density break-offs, connectivity and distance information. The proposed ACO-based algorithm uses the outputs of the neighborhood construction (NC) and the external shape formation. In addition, we propose a three-phase clustering algorithm that consists of NC, outlier detection and merging phases. We test the strengths and the weaknesses of the proposed approaches by extensive experimentation with data sets borrowed from literature and generated in a controlled manner. NC is found to be effective for arbitrary shaped clusters, intracluster and intercluster density variations. The external shape formation algorithm achieves significant reductions for convex clusters. The ACO-based and the three-phase clustering algorithms have promising results for the data sets having well-separated clusters.

In einer von Vernetzung und Globalisierung geprägten Umwelt wächst der Wettbewerbsdruck auf die Unternehmen am Markt stetig. Die effektive Nutzung der Ressourcen einerseits und die enge Zusammenarbeit mit Lieferanten und Kunden andererseits führen für nicht wenige Unternehmen des industriellen Sektors zu entscheidenden Wettbewerbsvorteilen, die das Fortbestehen jener Unternehmen am Markt sichern. Viele Unternehmen verstehen sich aus diesem Grund als Bestandteil so genannter Supply Chains. Die unternehmensübergreifende Steuerung und Optimierung des Wertschöpfungsprozesses stellt ein charakteristisches Problem des Supply Chain Managements dar und besitzt zur Erzielung von Wettbewerbsvorteilen hohes Potential. Produktionsnetzwerke sind ein wesentlicher Forschungsschwerpunkt der Professur für Produktionswirtschaft und Industriebetriebslehre an der TU Chemnitz. Das Extended Value Chain Management (EVCM) stellt ein kompetenzorientiertes Konzept für die Bildung und zum Betrieb hierarchieloser temporärer regionaler Produktionsnetzwerke im Sinne virtueller Unternehmen dar. Gegenstand dieser Arbeit ist ein diskretes Optimierungsproblem, dass einen mehrstufigen Entscheidungsprozesses unter Berücksichtigung mehrerer Ziele abbildet, der sich bei der Auswahl möglicher Partner in einem Produktionsnetzwerk nach dem Betreiberkonzept des EVCM ergibt. Da mehrere Zielstellungen bestehen, werden grundlegende Methoden der mehrkriteriellen Optimierung und Entscheidung erörtert. Neben der Vorstellung des Problems sollen mehrzielorientierte Ansätze im Sinne einer Pareto-Optimierung auf Basis des Dynamic Programmings als Verfahren zur Bestimmung von Optimallösungen sowie Ant Colony Optimization zur näherungsweisen Lösung vorgestellt werden. Darauf aufbauend werden verschiedene Möglichkeiten der Hybridisierung beider Methoden diskutiert. Die entwickelten Ansätze werden auf ihre Eignung im Rahmen der informationstechnischen Umsetzung des EVCM-Konzepts untersucht und einer Evaluierung unterzogen. Hierzu werden verschiedene Kennzahlen zur Beurteilung der Verfahren entwickelt. Die modellierten Algorithmen und entwickelten Konzepte beschränken sich nicht ausschließlich auf das betrachtete Problem, sondern können leicht auf Probleme mit ähnlichen Eigenschaften übertragen werden. Insbesondere das NP-vollständige mehrkriterielle Kürzeste-Wege-Problem stellt einen Spezialfall des behandelten Optimierungsproblems dar.

In the last decades there has been a lot of interest in computational models and metaheuristics algorithms capable to solve combinatorial optimization problems. The recent trend is to define these algorithms taking inspiration by the observation of natural systems. In this thesis the Ant Colony System (ACS) is presented which has been inspired by the observation of real ant colonies. ACS is initially proposed to solve the symmetric and asymmetric travelling salesman problems where it is shown to be competitive with other metaheuristics. Although this is an interesting and promising result, it was immediately clear that ACS, as well as other metaheuristics, in many cases cannot compete with specialized local search methods. An interesting trend is therefore to couple metaheuristics with a local optimizer, giving birth to so-called hybrid methods. Along this line, the thesis investigates MACS-VRPTW (Multiple ACS for the Vehicle Routing Problem with Time Windows) and HAS-SOP: Hybrid Ant System for the Sequential Ordering Problem (SOP). In the second part the thesis introduces some modifications of the original ACS algorithm. These modifications are able to speed up the method and to make it more competitive in case of large problem instances. The resulting framework, called Enhanced Ant Colony System is tested for the SOP. Finally the thesis presents the application of ACS to solve real-life vehicle routing problems where additional constraints and stochastic information are included. / Doctorat en Sciences de l'ingénieur / info:eu-repo/semantics/nonPublished