Avaliação de riscos de segurança BGP na Internet com emulação escalável

• Ano de defesa: 2021
• Autor(a) principal: Barea, Emerson Rogério Alves
• Orientador(a): Senger, Hermes
• Tipo de documento: Tese
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de São Carlos; Câmpus São Carlos
• Programa de Pós-Graduação: Programa de Pós-Graduação em Ciência da Computação - PPGCC
• Palavras-chave em Português: BGP; Ataque; Tecnologia de emulação leve e distribuída
• Palavras-chave em Inglês: Testbed; Attack; Lightweight and distributed emulation technology
• Área do conhecimento CNPq: CIENCIAS EXATAS E DA TERRA::CIENCIA DA COMPUTACAO::SISTEMAS DE COMPUTACAO
• Link de acesso: https://repositorio.ufscar.br/handle/20.500.14289/16338

Border Gateway Protocol (BGP) is the predominant routing protocol that allows different Autonomous Systems (AS) to communicate on the Internet. Attacks on BGP typically have a significant impact as they can cause unavailability and compromise sensitive data over large geographic areas. Due to the topic’s relevance, several studies were developed that address different research areas on security in BGP. An area of study that stands out is the reproduction of testbeds for detailed protocol analysis. However, a common characteristic of validating the proposals is using a limited representation in scale and fidelity of the Internet structure. This limitation makes it impossible to have a global view of the risk associated with the attacks and identify the mitigation methods. Given this scenario, it became essential to carry out an in-depth analysis of the risks to the BGP protocol at the real and global Internet infrastructure level. This analysis demands solutions representing the real Internet topology and the accurate reproduction of the attack events on real BGP software implementations. This work presents the Minimalistic Security BGP (MiniSecBGP), an emulation-based testbed, and a methodology for risk analysis of interdomain network routing infrastructures. Our solution can use either real or synthetic data about attacks, which allows a better understanding of past attacks to “what-if” scenarios to identify more robust mitigation strategies. The validation tests demonstrated that MiniSecBGP faithfully reproduces real scenarios using topology data and attack events extracted automatically from public datasets or synthetically created by the user.