• JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  • JoomlaWorks Simple Image Rotator
  Bookmark and Share
Doctoral Thesis
Full name
Luiz Arthur Feitosa dos Santos
Knowledge Area
Date of Defense
São Paulo, 2016
Title in Portuguese
Uma abordagem autonômica para mitigar ciberataques em redes de computadores
Keywords in Portuguese
Crime Por Computador
Redes De Computadores
Segurança De Computadores
Abstract in Portuguese
Nos últimos anos observa-se o crescimento dos problemas relacionados com segurança em redes de computadores locais, que frequentemente são alvos ou fontes de diversos tipos de ciberataques. Em parte, isso ocorre porque as redes locais estão se tornando extremamente dinâmicas e heterogêneas. Como agravante, os ataques à segurança estão mais sosticados, pois muitos são compostos por várias etapas e utilizam diferentes métodos para concretizar a investida, o que diculta a iden- ticação e reação contra essas ameaças. Desta maneira, manter a segurança em ambientes de rede tão heterogêneos e dinâmicos que são frequentemente expostos a pragas digitais, torna-se uma tarefa complexa para o administrador de rede. Nesse contexto, o presente trabalho tem por objetivo desenvolver uma arquitetura autonômica que mantenha a segurança de redes de computadores, exigindo o mínimo de intervenção humana. Para alcançar esse objetivo, propõem-se uma abordagem que emprega aprendizagem de máquina para processar, similarmente à memória humana, históricos de uso da rede e alertas de segurança, para extrair regras de segurança que são aplicadas autonomicamente na rede, por intermédio da tecnologia OpenFlow. Tal arquitetura, ainda propõem utilizar mensagens postadas em redes sociais, para extrair alertas de cibersegurança que auxiliem o administrador da rede a elucidar problemas na rede local. Nos experimentos executados, a arquitetura autonômica proposta conseguiu mitigar até 97,5 por cento dos pacotes maliciosos referentes a ataques DDoS.
Title in English
An autonomic approach to mitigate cyberattacks in computer networks
Abstract in English
Nowadays it is possible to observe the growth of computer security problems, mainly the threats related with local computer networks, which are often targets or sources of many types of security attacks. One reason for this is that local networks are becoming extremely dynamic and heterogeneous. To aggravate this situation, the security threats are becoming more sophisticated too. For instance, many attacks use several steps and dierent methods to achieve their objective, which usually complicates the identication and reaction against these threats. In this scenario, it is a hard task for a human to deal with cyber attacks, mainly due the increasing number of users and heterogeneous devices in LAN encouraged by practices such as BYOD, that constantly brings new threats to these environments. Therefore, to reduce the necessity of human interaction to maintain the network security, we propose an autonomic approach that uses machine learning to process, in a similar way of human memory, the historical of network usage and security alerts, to generate security rules, that are imposed to network using SDN resources, to mitigate cyber attacks. In addition, we propose a method of extracting cyber alerts based on messages posted on social networks, which can be used to prevent security problems in computing networks. During the experiments, the au- tonomic architecture proposed was able to mitigate 97.5 percent of malicious packets generated by DDoS attacks.
WARNING - Viewing this document is conditioned on your acceptance of the following terms of use:
This document is only for private use for research and teaching activities. Reproduction for commercial use is forbidden. This rights cover the whole data about this document as well as its contents. Any uses or copies of this document in whole or in part must include the author's name.
Publishing Date
WARNING: Learn what derived works are clicking here.
All rights of the thesis/dissertation are from the authors
Digital Library of Theses and Dissertations of USP. Copyright © 2001-2023. All rights reserved.