A high-level authorization framework for software-defined networks

• Ano de defesa: 2017
• Autor(a) principal: ROSENDO, Daniel
• Orientador(a): KELNER, Judith
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: eng
• Instituição de defesa: Universidade Federal de Pernambuco
• Programa de Pós-Graduação: Programa de Pos Graduacao em Ciencia da Computacao
• País: Brasil
• Palavras-chave em Português: Redes de computadores; Internet das coisas
• Link de acesso: https://repositorio.ufpe.br/handle/123456789/25356

Network Access Control (NAC) management is a critical task. Misconfigurations may result in vulnerabilities that may compromise the overall network security. Traditional access control setups rely on firewalls, IEEE 802.1x, VLAN, ACL, and LDAP. These approaches work well for stable and small networks and are hard to integrate and configure. Besides, they are inflexible and require per-device and vendor-specific configurations, being error-prone. The Software-Defined Networking (SDN) paradigm overcomes architectural problems of traditional networks, simplifies the network design and operation, and offers new opportunities (programmability, flexibility, dynamicity, and standardization) to manage these issues. Furthermore, SDN reduces the human intervention, which in turn also reduce operational costs and misconfigurations. Despite this, access control management remains a challenge, once managing security policies involves dealing with a large set of access control rules; detection of conflicting policies; defining priorities; delegating rights; reacting to dynamic network states and events. This dissertation explores the use of SDN to mitigate these problems. We present HACFlow, a novel SDN framework for network access control management based on the OrBAC model. HACFlow aims to simplify and automate the NAC management. It allows network operators to govern rights of network entities by defining dynamic, fine-grained, and high-level access control policies. To illustrate the operation of HACFlow we present through a step by step how the main management tasks are executed. Our study case is a Smart City network environment. We conducted many experiments to analyze the scalability and performance of HACFlow, and the results show that it requires a time in the order of milliseconds to execute all the management tasks, even managing many policies. Besides, we compare HACFlow against related approaches.