Thingtegrity: A Scalable Trusted Computing Architecture for the Internet of Things

Remote attestation is used to prove the integrity of one system (prover) to another (challenger). The prover measures its configuration and transmits the result to the challenger for verification. Common attestation methods lead to complex configuration measurements (e.g., hash of all executables), which are updated every time one of the software modules changes. The updated configuration has to be distributed to all possible challengers since they need a reference to enable the verification. Recently, an idea of reducing the complexity of the configuration measurement by taking into account privileges of software modules has been presented. However, this approach has not been exhaustively analyzed since, as yet, no implementation exists. Especially in the Internet of Things (IoT) domain, where resources are constrained strictly while devices are potentially physically exposed to adversaries, attestation methodologies with reduced overhead are desireable. In this work we combine binary-, property- and privilege-based remote attestation to integrate a trusted computing architecture transparently into \iotivity, an existing IoT middleware. As a first step, we aim to enable to attestation of the integrity of complex devices with different services to constrained devices. With the help of an illustrative simulated environment, we show that our architecture reduces the effort of bootstrapping trusted relations, as well as updating single modules in the whole system, even if software and devices from different vendors are combined.