Configuring Data Flows in the Internet of Things for Security and Privacy Requirements

Abstract

The Internet of Things is a highly distributed, highly dynamic environment where data can flow among entities (the ‘things’) in complex data flow configurations. For data secrecy, it is important that only certain data flows be allowed. Research in this area is often based on the use of the well-known lattice model. However, as shown in previous papers, by using a basic result of directed graph theory (or of order theory) it is possible to use a less constrained model based on partial orders, for which a formal notion of secrecy can be defined. We define a notion of ‘allowed contents’ for each ‘thing’ and then the data flows follow by inclusion relationships. By taking advantage of transitivity of data flows and of strongly connected component algorithms, these data flow relationships can then be simplified. It is shown that several data flow relationships can coexist in a network. Two small examples are presented, one on hospital applications and another on e-commerce. Implementation issues are discussed.