Um ambiente de experimentação para redes 6LoWPAN e um módulo de autenticação baseado em chave simétrica

• Autor(a) principal: Vitali, Maycon Maia
• Data de Publicação: 2012
• Tipo de documento: Dissertação
• Idioma: por
• Título da fonte: Repositório Institucional da Universidade Federal do Espírito Santo (riUfes)
• Texto Completo: http://repositorio.ufes.br/handle/10/4262
• topic: Ciência da Computação; 004

Like traditional computer networks, Wireless Sensor Networks (WSNs) are subject to several secutity attacks, such as denial of service, spoofing addresses attack, node compromise, and other attacks specific to sensor networks. In a scenario of Smart Cities, the concerns about security takes an even greater considering that critical data will be flowing at all times by the communication paths provided by the WSN networks. In this sense, research in WSN security gains a great deal of importance, and pushes the development of research involving at the creation of new algorithms, protocols, tools and specific environments for increasing security in Wireless Sensor Networks. This work proposes a testing environment that allows to simulate mixed WSN, a way to interact with them and evaluate the WSN concerning to energy consumption, topology, and number of packets sent and received, among other features. The testing environment is capable of simulating 6LoWPAN networks, being able to access the sensor nodes using their IPv6 addresses. The proposed approach allows remote applications, using the Internet, to communicate with the 6LoWPAN network, apart from the fact that the environment is fully simulated. As proof of concept, we simulated a 6LoWPAN network, promoting an sybil attack in the network topology, and subsequently evaluated the organization of the network topology with and without the sybil node running. An interesting feature of the proposed architecture is that it enables remote testing of applications, so a machine with more processing capacity may be responsible for generating and simulating the sensor network, while applications that collect and process data are located in another machine with less processing power. As an additional contribution, we propose a new methodology for network node association based on pre-shared keys. The methodology aims to mitigate attacks where malicious nodes assume various identities in the network in order to impair the vision of the same topological. Furthermore, we present a protection mechanism by storing the pre-shared key only in volatile memory, which makes more difficult the discovery of the key in the case of a node failure, or even the destruction of the key together with the death of the sensor when your energy comes to the end.