Submitted by Irene Nascimento (irene.kessia@ufpe.br) on 2016-07-12T18:32:53Z No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Final_bib.pdf: 4325542 bytes, checksum: 5cafa644d256b743ce0f06490e4d5920 (MD5) / Made available in DSpace on 2016-07-12T18:32:53Z (GMT). No. of bitstreams: 2 license_rdf: 1232 bytes, checksum: 66e71c371cc565284e70f40736c94386 (MD5) Tese_Final_bib.pdf: 4325542 bytes, checksum: 5cafa644d256b743ce0f06490e4d5920 (MD5) Previous issue date: 2015-08-20 / Atualmente, a evolução na arquitetura dos FPGAs (Field programable gate arrays) permite que os mesmos sejam empregados em aplicações que vão desde a prototipação rápida de circuitos digitais simples a coprocessadores para computação de alto desempenho. Entretanto, a utilização eficiente dessas arquiteturas é fortemente dependente, entre outros fatores, da ferramenta de síntese empregada. O desafio das ferramentas de síntese está em converter a lógica do projetista em circuitos que utilizem de maneira efetiva a área do chip, não degradem a frequência de operação e que, sobretudo, sejam eficientes em reduzir o consumo de energia. Nesse sentido, pesquisadores e grandes fabricantes de FPGA estão, frequentemente, desenvolvendo novas ferramentas com vistas a esses objetivos, que se caracterizam por serem conflitantes. O fluxo de síntese de projetos baseados em FPGAs engloba as etapas de otimização lógica, mapeamento, agrupamento, posicionamento e roteamento. Essas fases são dependentes, de forma que, otimizações nas etapas iniciais produzem impactos positivos nas etapas posteriores. No âmbito deste trabalho de doutorado, estamos propondo uma metodologia para otimização do fluxo de síntese, especificamente, nas etapas de mapeamento e agrupamento. Classicamente, a etapa de mapeamento é realizada mediante heurísticas que determinam uma solução para o problema, mas que, não permitem a busca por soluções ótimas, ou que beneficiam um objetivo em detrimento de outros. Desta forma, estamos propondo a utilização de uma abordagem multiobjetivo baseada em algoritmo genético e de uma abordagem multiobjetivo baseada em colônia artificial de abelhas que, associadas a heurísticas específicas do problema, permitem que sejam obtidas soluções de melhor qualidade e que resultam em circuitos finais com área reduzida, ganhos na frequência de operação e com menor consumo de potência dinâmica. Além disso, propomos uma nova abordagem de agrupamento multiobjetivo que se diferencia do estado da arte, por utilizar uma técnica de predição e por considerar características dinâmicas do problema, produzindo circuitos mais eficientes e que facilitam a tarefa das etapas de posicionamento e roteamento. Toda a metodologia proposta foi integrada ao fluxo acadêmico do VTR (Verilog to routing), um projeto código aberto e colaborativo que conta com múltiplos grupos de pesquisa, conduzindo trabalhos nas áreas de desenvolvimento de arquitetura de FPGAs e de novas ferramentas de síntese. Além disso, utilizamos como benchmark, um conjunto dos 20 maiores circuitos do MCNC (Microelectronics Center of North Carolina) que são frequentemente utilizados em pesquisas da área. O resultado do emprego integrado das ferramentas frutos da metodologia proposta permite a redução de importantes aspectos pós-roteamento avaliados. Em comparação ao estado da arte, são obtidas, em média, redução na área dos circuitos de até 19%, além da redução do caminho crítico em até 10%, associada à diminuição na potência dinâmica total estimada de até 18%. Os experimentos também mostram que as metodologias de mapeamento propostas são computacionalmente mais custosas em comparação aos métodos presentes no estado da arte, podendo ser até 4,7x mais lento. Já a metodologia de agrupamento apresentou pouco ou nenhum overhead em comparação ao metodo presente no VTR. Apesar do overhead presente no mapeamento, os métodos propostos, quando integrados ao fluxo completo, podem reduzir o tempo de execução da síntese em cerca de 40%, isto é o resultado da produção de circuitos mais simples e que, consequentemente, favorecem as etapas de posicionamento e roteamento. / Nowadays, the evolution of FPGAs (Field Programmable Gate Arrays) allows them to be employed in applications from rapid prototyping of digital circuits to coprocessor of high performance computing. However, the efficient use of these architectures is heavily dependent, among other factors, on the employed synthesis tool. The synthesis tools challenge is in converting the designer logic into circuits using effectively the chip area, while, do not degrade the operating frequency and, especially, are efficient in reducing power consumption. In this sense, researchers and major FPGA manufacturers are often developing new tools to achieve those goals, which are characterized by being conflicting. The synthesis flow of projects based on FPGAs comprises the steps of logic optimization, mapping, packing, placement and routing. These steps are dependent, such that, optimizations in the early stages bring positive results in later steps. As part of this doctoral work, we propose a methodology for optimizing the synthesis flow, specifically, on the steps of mapping and grouping. Classically, the mapping step is performed by heuristics which determine a solution to the problem, but do not allow the search for optimal solutions, or that benefit a goal at the expense of others. Thus, we propose the use of a multi-objective approach based on genetic algorithm and a multi-objective approach based on artificial bee colony that, combined with problem specific heuristics, allows a better quality of solutions are obtained, yielding circuits with reduced area, operating frequency gains and lower dynamic power consumption. In addition, we propose a new multi-objective clustering approach that differs from the state-of-the-art, by using a prediction technique and by considering dynamic characteristics of the problem, producing more efficient circuits and that facilitate the tasks of placement and routing steps . The proposal methodology was integrated into the VTR (Verilog to routing) academic flow, an open source and collaborative project that has multiple research groups, conducting work in the areas of FPGA architecture development and new synthesis tools. Furthermore, we used a set of the 20 largest MCNC (Microelectronics Center of North Carolina) benchmark circuits that are often used in research area. The results of the integrated use of tools based on the proposed methodology allow the reduction of important post-routing aspects evaluated. Compared to the stateof- the-art, are achieved, on average, 19% reduction in circuit area, besides 10% reduction in critical path, associated with 18% decrease in the total dynamic estimated power. The experiments also reveal that proposed mapping methods are computationally more expensive in comparison to methods in the state-of-the-art, and may even be 4.7x slower. However, the packing methodology presented little or no overhead compared to the method in VTR. Although the present overhead mapping, the proposed methods, when integrated into the complete flow, can reduce the running time of the synthesis by approximately 40%, which is the result of more simple circuits and which, consequently, favor the steps of placement and routing.

Submitted by Renata Lopes (renatasil82@gmail.com) on 2017-03-06T11:57:32Z No. of bitstreams: 1 grasielereginaduarte.pdf: 2553018 bytes, checksum: e0b9afbcc0b18965321f8db8ea7d38b8 (MD5) / Approved for entry into archive by Adriana Oliveira (adriana.oliveira@ufjf.edu.br) on 2017-03-06T20:19:40Z (GMT) No. of bitstreams: 1 grasielereginaduarte.pdf: 2553018 bytes, checksum: e0b9afbcc0b18965321f8db8ea7d38b8 (MD5) / Made available in DSpace on 2017-03-06T20:19:40Z (GMT). No. of bitstreams: 1 grasielereginaduarte.pdf: 2553018 bytes, checksum: e0b9afbcc0b18965321f8db8ea7d38b8 (MD5) Previous issue date: 2015-03-06 / CAPES - Coordenação de Aperfeiçoamento de Pessoal de Nível Superior / Os problemas de otimização estão presentes em diversas áreas de atuação da sociedade e o uso de algoritmos bio-inspirados para a resolução de problemas complexos deste tipo vem crescendo constantemente. O Algoritmo Colônia de Abelhas Artificiais (ABC – do inglês Artificial Bee Colony) é um algoritmo bio-inspirado proposto em 2005 para a resolução de problemas de otimização multimodais e multidimensionais. O fenômeno natural que inspirou o desenvolvimento do ABC foi o comportamento inteligente observado em colônias de abelhas, mais especificamente no forrageamento. O ABC foi proposto inicialmente para ser aplicado na resolução de problemas sem restrições. Este trabalho avalia o desempenho do ABC quando aplicado na resolução de problemas de otimização com restrições. Para o tratamento das restrições, métodos de penalização serão incorporados ao ABC. São analisados diversos métodos de penalização, de diferentes tipos, com o objetivo de identificar com qual deles o algoritmo apresenta melhor desempenho. Além disto, são avaliadas possíveis limitações e cuidados que devem ser tomados ao combinar métodos de penalização ao ABC. O algoritmo proposto é avaliado através da resolução de problemas de otimização encontrados na literatura. Vários experimentos computacionais são realizados e gráficos e tabelas são gerados para demonstração dos resultados obtidos que também são discutidos. / Optimization problems are present in several areas of society and the use of bio-inspired algorithms to solve complex problems of this type has been growing constantly. The Artificial Bee Colony Algorithm (ABC) is a bio-inspired algorithm proposed in 2005 for solving multimodal and multidimensional optimization problems. The natural phenomenon that inspired the development of the ABC was intelligent behavior observed in bee colonies, more specifically in foraging. The ABC was initially proposed to be applied to solve unconstrained problems. This study evaluates the performance of ABC when applied in solving constrained optimization problems. For the treatment of constraints, penalty methods will be incorporated into the ABC. Several penalty methods, of different types, are analyzed with the goal of identifying which of these penalty methods offers better performance. Furthermore, possible limitations and care that should be taken when combining penalty methods to ABC are evaluated. The proposed algorithm is evaluated by solving optimization problems found in the literature. Several computational experiments are performed and graphs and tables are generated for demonstration of the obtained results which are also discussed.

Recently, continuous-valued optimization problems have received a great amount of focus, resulting in optimization algorithms which are very efficient within the continuous-valued space. Many optimization problems are, however, defined within the binary-valued problem space. These continuous-valued optimization algorithms can not operate directly on a binary-valued problem representation, without algorithm adaptations because the mathematics used within these algorithms generally fails within a binary problem space. Unfortunately, such adaptations may alter the behavior of the algorithm, potentially degrading the performance of the original continuous-valued optimization algorithm. Additionally, binary representations present complications with respect to increasing problem dimensionality, interdependencies between dimensions, and a loss of precision. This research investigates the possibility of applying continuous-valued optimization algorithms to solve binary-valued problems, without requiring algorithm adaptation. This is achieved through the application of a mapping technique, known as angle modulation. Angle modulation effectively addresses most of the problems associated with the use of a binary representation by abstracting a binary problem into a four-dimensional continuous-valued space, from which a binary solution is then obtained. The abstraction is obtained as a bit-generating function produced by a continuous-valued algorithm. A binary solution is then obtained by sampling the bit-generating function. This thesis proposes a number of population-based angle-modulated continuous-valued algorithms to solve binary-valued problems. These algorithms are then compared to binary algorithm counterparts, using a suite of benchmark functions. Empirical analysis will show that the angle-modulated continuous-valued algorithms are viable alternatives to binary optimization algorithms. Copyright 2012, University of Pretoria. All rights reserved. The copyright in this work vests in the University of Pretoria. No part of this work may be reproduced or transmitted in any form or by any means, without the prior written permission of the University of Pretoria. Please cite as follows: Pamparà, G 2012, Angle modulated population based algorithms to solve binary problems, MSc dissertation, University of Pretoria, Pretoria, viewed yymmdd < http://upetd.up.ac.za/thesis/available/etd-02242012-090312 / > C12/4/188/gm / Dissertation (MSc)--University of Pretoria, 2012. / Computer Science / unrestricted

L'optimisation difficile représente une classe de problèmes dont la résolution ne peut être obtenue par une méthode exacte en un temps polynomial.Trouver une solution en un temps raisonnable oblige à trouver un compromis quant à son exactitude.Les métaheuristiques sont une classe d'algorithmes permettant de résoudre de tels problèmes, de manière générique et efficiente (i.e. trouver une solution satisfaisante selon des critères définis: temps, erreur, etc.).Le premier chapitre de cette thèse est notamment consacré à la description de cette problématique et à l'étude détaillée de deux familles de métaheuristiques à population, les algorithmes évolutionnaires et les algorithmes d'intelligence en essaim.Afin de proposer une approche innovante dans le domaine des métaheuristiques, ce premier chapitre présente également la notion d'analyse de sensibilité.L'analyse de sensibilité permet d'évaluer l'influence des paramètres d'une fonction sur son résultat.Son étude caractérise globalement le comportement de la fonction à optimiser (linéarité, influence, corrélation, etc.) sur son espace de recherche.L'incorporation d'une méthode d'analyse de sensibilité au sein d'une métaheuristique permet d'orienter sa recherche le long des dimensions les plus prometteuses.Deux algorithmes réunissant ces notions sont proposés aux deuxième et troisième chapitres.Pour le premier algorithme, ABC-Morris, la méthode de Morris est introduite dans la métaheuristique de colonie d'abeilles artificielles (ABC).Cette inclusion est dédiée, les méthodes reposant sur deux équations similaires.Afin de généraliser l'approche, une nouvelle méthode, NN-LCC, est ensuite développée et son intégration générique est illustrée sur deux métaheuristiques, ABC avec taux de modification et évolution différentielle.L'efficacité des approches proposées est testée sur le jeu de données de la conférence CEC 2013. L'étude se réalise en deux parties: une analyse classique de la méthode vis-à-vis de plusieurs algorithmes de la littérature, puis vis-à-vis de l'algorithme d'origine en désactivant un ensemble de dimensions, provoquant une forte disparité des influences / Hard optimization stands for a class of problems which solutions cannot be found by an exact method, with a polynomial complexity.Finding the solution in an acceptable time requires compromises about its accuracy.Metaheuristics are high-level algorithms that solve these kind of problems. They are generic and efficient (i.e. they find an acceptable solution according to defined criteria such as time, error, etc.).The first chapter of this thesis is partially dedicated to the state-of-the-art of these issues, especially the study of two families of population based metaheuristics: evolutionnary algorithms and swarm intelligence based algorithms.In order to propose an innovative approach in metaheuristics research field, sensitivity analysis is presented in a second part of this chapter.Sensitivity analysis aims at evaluating arameters influence on a function response. Its study characterises globally a objective function behavior (linearity, non linearity, influence, etc.), over its search space.Including a sensitivity analysis method in a metaheuristic enhances its seach capabilities along most promising dimensions.Two algorithms, binding these two concepts, are proposed in second and third parts.In the first one, ABC-Morris, Morris method is included in artificial bee colony algorithm.This encapsulation is dedicated because of the similarity of their bare bone equations, With the aim of generalizing the approach, a new method is developped and its generic integration is illustrated on two metaheuristics.The efficiency of the two methods is tested on the CEC 2013 conference benchmark. The study contains two steps: an usual performance analysis of the method, on this benchmark, regarding several state-of-the-art algorithms and the comparison with its original version when influences are uneven deactivating a subset of dimensions

Yes / In dealing with big data, we need effective algorithms; effectiveness that depends, among others, on the ability to remove outliers from the data set, especially when dealing with classification problems. To this aim, support vector finder algorithms have been created to save just the most important data in the data pool. Nevertheless, existing classification algorithms, such as Fuzzy C-Means (FCM), suffer from the drawback of setting the initial cluster centers imprecisely. In this paper, we avoid existing shortcomings and aim to find and remove unnecessary data in order to speed up the final classification task without losing vital samples and without harming final accuracy; in this sense, we present a unique approach for finding support vectors, named evolutionary Pentagon Support Vector (PSV) finder method. The originality of the current research lies in using geometrical computations and evolutionary algorithms to make a more effective system, which has the advantage of higher accuracy on some data sets. The proposed method is subsequently tested with seven benchmark data sets and the results are compared to those obtained from performing classification on the original data (classification before and after PSV) under the same conditions. The testing returned promising results.

O ambiente de produção abordado neste trabalho é o Job Shop Flexível (JSF), uma generalização do Job Shop (JS). O problema de programação de tarefas, ou jobs, no ambiente JS é classificado por Garey; Johnson e Sethi (1976) como NP-Difícil e o JSF é, no mínimo, tão difícil quanto o JS. O JSF é composto por um conjunto de jobs, cada qual constituído por operações. Cada operação deve ser processada individualmente, sem interrupção, em uma única máquina de um subconjunto de máquinas habilitadas. O principal critério de desempenho considerado é a minimização dos atrasos dos jobs. São apresentados modelos de Programação Linear Inteira Mista (PLIM) para minimizar o atraso total e o instante de término da última operação, o makespan. São propostas novas regras de prioridade dos jobs, além de adaptações de regras da literatura. Tais regras são utilizadas por heurísticas construtivas e são aliadas a estratégias cujo objetivo é explorar características específicas do JSF. Visando aprimorar as soluções inicialmente obtidas, são propostas buscas locais e outros mecanismos de melhoria utilizados no desenvolvimento de três meta-heurísticas de diferentes categorias. Essas meta-heurísticas são: Iterated Local Search (ILS), classificada como meta-heurística de trajetória; Greedy Randomized Adaptive Search (GRASP), meta-heurística construtiva; e Artificial Bee Colony (ABC), meta-heurística populacional recentemente proposta. Esses métodos foram selecionados por alcançarem bons resultados para diversos problemas de otimização da literatura. São realizados experimentos computacionais com 600 instâncias do JSF, permitindo comparações entre os métodos de resolução. Os resultados mostram que explorar as características do problema permite que uma das regras de prioridade propostas supere a melhor regra da literatura em 81% das instâncias. As meta-heurísticas ILS, GRASP e ABC chegam a conseguir mais de 31% de melhoria sobre as soluções iniciais e a obter atrasos, em média, somente 2,24% superiores aos das soluções ótimas. Também são propostas modificações nas meta-heurísticas que permitem obter melhorias ainda mais expressivas sem aumento do tempo de execução. Adicionalmente é estudada uma versão do JSF com operações de Montagem e Desmontagem (JSFMD) e os experimentos realizados com um conjunto de 150 instâncias também indicam o bom desempenho dos métodos desenvolvidos. / The production environment addressed herein is the Flexible Job Shop (FJS), a generalization of the Job Shop (JS). In the JS environment, the jobs scheduling problem is classified by Garey; Johnson and Sethi (1976) as NP-Hard and the FJS is at least as difficult as the JS. FJS is composed of a set of jobs, each consisting of operations. Each operation must be processed individually, without interruption, in a single machine of a subset of enabled machines. The main performance criterion is minimizing the jobs tardiness. Mixed Integer Linear Programming (MILP) models are presented. These models minimize the total tardiness and the completion time of the last operation, makespan. New priority rules of jobs are proposed, as well as adaptations of rules from the literature. These rules are used by constructive heuristics and are combined with strategies aimed at exploiting specific characteristics of FSJ. In order to improve the solutions initially obtained, local searches and other improvement mechanisms are proposed and used in the development of metaheuristics of three different categories. These metaheuristics are: Iterated Local Search (ILS), classified as trajectory metaheuristic; Greedy Randomized Adaptive Search (GRASP), constructive metaheuristic, and Artificial Bee Colony (ABC), recently proposed population metaheuristic. These methods were selected owing to their good results for various optimization problems in the literature. Computational experiments using 600 FJS instances are carried out to allow comparisons between the resolution methods. The results show that exploiting the characteristics of the problem allows one of the proposed priority rules to exceed the best literature rule in about 81% of instances. Metaheuristics ILS, GRASP and ABC achieve more than 31% improvement over the initial solutions and obtain an average tardiness only 2.24% higher than the optimal solutions. Modifications in metaheuristics are proposed to obtain even more significant improvements without increased execution time. Additionally, a version called Disassembly and Assembly FSJ (DAFJS) is studied and the experiments performed with a set of 150 instances also indicate good performance of the methods developed.

O ambiente de produção abordado neste trabalho é o Job Shop Flexível (JSF), uma generalização do Job Shop (JS). O problema de programação de tarefas, ou jobs, no ambiente JS é classificado por Garey; Johnson e Sethi (1976) como NP-Difícil e o JSF é, no mínimo, tão difícil quanto o JS. O JSF é composto por um conjunto de jobs, cada qual constituído por operações. Cada operação deve ser processada individualmente, sem interrupção, em uma única máquina de um subconjunto de máquinas habilitadas. O principal critério de desempenho considerado é a minimização dos atrasos dos jobs. São apresentados modelos de Programação Linear Inteira Mista (PLIM) para minimizar o atraso total e o instante de término da última operação, o makespan. São propostas novas regras de prioridade dos jobs, além de adaptações de regras da literatura. Tais regras são utilizadas por heurísticas construtivas e são aliadas a estratégias cujo objetivo é explorar características específicas do JSF. Visando aprimorar as soluções inicialmente obtidas, são propostas buscas locais e outros mecanismos de melhoria utilizados no desenvolvimento de três meta-heurísticas de diferentes categorias. Essas meta-heurísticas são: Iterated Local Search (ILS), classificada como meta-heurística de trajetória; Greedy Randomized Adaptive Search (GRASP), meta-heurística construtiva; e Artificial Bee Colony (ABC), meta-heurística populacional recentemente proposta. Esses métodos foram selecionados por alcançarem bons resultados para diversos problemas de otimização da literatura. São realizados experimentos computacionais com 600 instâncias do JSF, permitindo comparações entre os métodos de resolução. Os resultados mostram que explorar as características do problema permite que uma das regras de prioridade propostas supere a melhor regra da literatura em 81% das instâncias. As meta-heurísticas ILS, GRASP e ABC chegam a conseguir mais de 31% de melhoria sobre as soluções iniciais e a obter atrasos, em média, somente 2,24% superiores aos das soluções ótimas. Também são propostas modificações nas meta-heurísticas que permitem obter melhorias ainda mais expressivas sem aumento do tempo de execução. Adicionalmente é estudada uma versão do JSF com operações de Montagem e Desmontagem (JSFMD) e os experimentos realizados com um conjunto de 150 instâncias também indicam o bom desempenho dos métodos desenvolvidos. / The production environment addressed herein is the Flexible Job Shop (FJS), a generalization of the Job Shop (JS). In the JS environment, the jobs scheduling problem is classified by Garey; Johnson and Sethi (1976) as NP-Hard and the FJS is at least as difficult as the JS. FJS is composed of a set of jobs, each consisting of operations. Each operation must be processed individually, without interruption, in a single machine of a subset of enabled machines. The main performance criterion is minimizing the jobs tardiness. Mixed Integer Linear Programming (MILP) models are presented. These models minimize the total tardiness and the completion time of the last operation, makespan. New priority rules of jobs are proposed, as well as adaptations of rules from the literature. These rules are used by constructive heuristics and are combined with strategies aimed at exploiting specific characteristics of FSJ. In order to improve the solutions initially obtained, local searches and other improvement mechanisms are proposed and used in the development of metaheuristics of three different categories. These metaheuristics are: Iterated Local Search (ILS), classified as trajectory metaheuristic; Greedy Randomized Adaptive Search (GRASP), constructive metaheuristic, and Artificial Bee Colony (ABC), recently proposed population metaheuristic. These methods were selected owing to their good results for various optimization problems in the literature. Computational experiments using 600 FJS instances are carried out to allow comparisons between the resolution methods. The results show that exploiting the characteristics of the problem allows one of the proposed priority rules to exceed the best literature rule in about 81% of instances. Metaheuristics ILS, GRASP and ABC achieve more than 31% improvement over the initial solutions and obtain an average tardiness only 2.24% higher than the optimal solutions. Modifications in metaheuristics are proposed to obtain even more significant improvements without increased execution time. Additionally, a version called Disassembly and Assembly FSJ (DAFJS) is studied and the experiments performed with a set of 150 instances also indicate good performance of the methods developed.

<p>U disertaciji je dato jedno rešenje automatskog operativnog<br />upravljanja dvosmernom brodskom prevodnicom sa jednom komorom.<br />Kreiran je ekspertski sistem zasnovan na rasplinutoj (fuzzy) logici.<br />Upravljački sistem je testiran na modelu brodske prevodnice koji je<br />kreiran na osnovu statističkih podataka o gustini saobraćaja na<br />hidrosistemu DTD (Dunav-Tisa-Dunav), na osnovu tehničke<br />dokumentacije brodske prevodnice i na osnovu razgovora sa<br />operaterima. Sistem je zatim optimizovan globalnim algoritmima<br />optimizacije. Dobijeno rešenje se pokazalo značajno bolje u poređenju<br />sa standardnim algoritmima odluke.</p> / <p>This thesis presents a solution to automatic control of a two-way one-channel<br />ship lock. Expert system based on fuzzy logic is designed. This control<br />system is tested on model of ship lock created using statistical data of<br />transportation density on DTD (Danube-Tisa-Danube) channel, using<br />technical documentation of ship lock and interview with operators. The<br />system is further optimized with global optimization techniques. Given<br />solution proved to be significantly better than standard decision algorithms.<br />&nbsp;</p>

La dissémination d’information dans les réseaux VANET est une opération fondamentale pour la sécurité routière. Il est dès lors nécessaire de concevoir et mettre en oeuvre des algorithmes efficaces et adaptatifs pour la dissémination d’informations sélectives et pertinentes.Dans ce travail, des approches Bio-inspirées sont proposées, à partir des comportements auto-organisés des essaims comme les colonies de fourmis et d’abeilles. Ces approches visent à fournir à chaque véhicule des informations en provenance de son environnement et alerter les conducteurs. Dans la première approche, le système de communication direct et indirect des fourmis est utilisé. Les fourmis partagent les informations sur les sources de nourriture avec des membres de la colonie en sécrétant la phéromone sur leurs chemins. La deuxième approche est inspirée par le système de communication des abeilles. Les abeilles partagent les informations à propos des sources de nourriture avec les autres membres de la ruche par des messages spécifiques, selon l’importance de ces sources.Une nouvelle mesure de "pertinence" associée aux messages est définie, par analogie à la sécrétion des phéromones des fourmis et au niveau de l’intensité des messages pour les abeilles, pour disséminer des informations de sécurité dans une zone géographique. Les simulations sont effectuées en utilisant le simulateur NS2 pour mesurer l’efficacité des approches proposées sous différentes conditions, en particulier en termes de densités et vitesses des véhicules. / Information dissemination in Vehicular Ad hoc Networks (VANETs) is a fundamental operation to increase the safety awareness among vehicles on roads. Thus, the design and implementation of efficient and scalable algorithms for relevant information dissemination constitutes a major issue that should be tackled.In this work, bio-inspired information dissemination approaches are proposed, that use self-organization principles of swarms such as Ant and Honey Bee colonies. These approaches are targeted to provide each vehicle with the required information about its surrounding and assist drivers to be aware of undesirable road conditions. In the first approach, Ant’s direct and indirect communication systems are used. Ants share information about food findings with colony members by throwing pheromone on the returning to the nest. The second, an RSU-based approach is inspired by the Bee communication system. Bees share profitable food sources with hive-mates in their hive by specific messages.A “relevance” value associated to the emergency messages is defined as an analogue to pheromone throwing in Ant colony, and as an analogue to profitability level in Bee colony, to disseminate safety information within a geographical area. Simulations are conducted using NS2 network simulator and relevant metrics are evaluated under different node speeds and network densities to show the effectiveness of the proposed approaches.