Um método para o gerenciamento da confiança na identificação de recursos e detecção de ataques para a internet das coisas.

Abstract

The Internet of Things (IoT) is a concept which describes how everyday objects are connected to the Internet, transforming them into intelligent and cooperative resources. Although IoT has many similarities with wireless sensor networks, IoT presents special requirements such as net- work integration, semantic context, connection to legacy systems and heterogeneous objects, and the need to protect against specific security attacks. IoT’s resources cooperate with each other by requesting and offering services. In heterogeneous and complex environments, these features need to trust each other. The swarm concept in IoT describes the cooperation of highly independent and heterogeneous devices to perform tasks. Components of a swarm system must seamlessly discover other objects. Semantic discovery is an approach to perform this task and can run auto- matically or manually, but it does not normally address security requirements, such as trust and potential attackers. In this thesis, a method based on machine learning and a dynamic elastic window technique (JED) is presented, that automatically assesses the IoT resource trust, recog- nizes IoT semantic attributes, helping the integration of infrastructures and services, evaluating service provider attributes. The proposed method collaborates with trust management and attacks identification. In simulated and real-world data, this method was able to recognize IoT semantic attributes, On-Off attacks and fault nodes with a precision of up to 96%.