Data Provenance for IoT With Light Weight Authentication and Privacy Preservation

The Internet of Things (IoT) engulfs a large number of interconnected heterogeneous devices from a wide range of pervasive application areas including health-care systems, energy management, environmental monitoring, and home and commercial automation. Although IoT is considered an enabling technology for a variety of services, it also raises many security and privacy concerns. This article focuses on developing secure protocols for data provenance with authentication and privacy preservation in IoT systems. Protocols for two scenarios are presented, one when an IoT device is directly connected to a wireless gateway and the other when an IoT device is indirectly connected to the wireless gateway through multiple hops of other IoT devices. The proposed protocols use physically unclonable functions along with wireless link fingerprints derived from the wireless channel characteristics between two communicating entities. This results in protocols which are not only efficient in terms of computational complexity and energy requirements but are also safe against various types of attacks including physical and cloning attacks. Experimental results show that in comparison to existing protocols, the proposed protocols are up to 100% more accurate in detecting attacks on data provenance and can save up to 83.8% and 73.5% energy consumption for the IoT devices in terms of CPU and radio energy, respectively.