Tracking sensitive and untrustworthy data in IoT

The Internet of Things (IoT) produces and processes large amounts of data. Among these data, some must be protected and others must be carefully handled because they come from untrusted sources. Taint analysis techniques can be used to for marking data and for monitoring their propagation at run time, so to determine how they influence the rest of the computation. Starting from the specification language IoT-LySa, we propose a Control Flow Analysis for statically predicting how tainted data spread across an IoT system and for checking whether those computations considered security critical are not affected by tainted data