Efficient Data Tagging for Managing Privacy in the Internet of Things

The Internet of Things creates an environment where software systems are influenced and controlled by phenomena in the physical world. The goal is invisible and natural interactions with technology. However, if such systems are to provide a high-quality personalised service to individuals, they must by necessity gather information about those individuals. This leads to potential privacy invasion. Using techniques from Information Flow Control, data representing phenomena can be tagged with their privacy properties, allowing a trusted computing base to control access based on sensitivity and the system to reason about the flows of private data. For this to work well, tags must be assigned as soon as possible after phenomena are detected. Tagging within resource-constrained sensors raises worries that computing the tags may be too expensive and that useful tags are too large in relation to the data´s size and the data´s sensitivity. This paper assuages these worries, giving code templates for two small micro controllers (PIC and AVR) that effect meaningful tagging.