This research advances the knowledge and capability for securely performing general-purpose computation and communication on zero-power devices that rely on harvested energy stored in extremely small reservoirs such as capacitors. A type of zero-power device, computational RFIDs (CRFIDs) harvest RF energy and endure continual interruptions to power. Complete loss of RAM on the order of every second makes the notion of a computational checkpoint fundamental to this model of computing. A significant problem is how to perform effective computation and checkpoints in a secure and privacy-preserving manner. The most fundamental question is: how to securely make forward progress of computation on zero-power devices? The exploration of checkpointing strategies that exploit the energy-centric properties of computational RFIDs will lay a solid foundation for designing secure software and systems for pervasive devices. Beyond security and privacy, the research seeks to discover what kinds of computational problems can be solved in a practical sense on extremely resource-constrained devices. While the experiments focus on computational RFIDs, many of the techniques will apply broadly to devices with batteries as well.
The impact of this work will lead to improved security for computation in low-power, untrusted infrastructure. The expected results will be published in venues at the intersection of security & privacy and low-power computation. Potential applications include improved security and privacy for (1) implantable medical devices and (2) sensors embedded in concrete components of bridges and roads.
