PhD Defense: Xiao Wu

Abstract:

The Internet of Things improves people's lives and assists research advancement by collecting and converting data about the environment. With the reduction of volume of sensor nodes, data collection and computation can be achieved in a wider range of environments, enabling applications such as health monitoring, tamper detection and industrial sensing. Small form factor sensor nodes present unique challenges in circuit design and system integration due to size and power constraints. This thesis introduces circuit techniques and system designs featuring small form factor sensor nodes: (1) a discontinuous switched-capacitor energy harvester for ultra-low power energy harvesting, (2) a fully integrated energy reservoir unit that uses a counter flow method for peak power delivery in space-constrained sensor systems, and (3) a complete wireless sensor node for accurate cellular temperature measurement, enabling implantation in a cluster of cells or large egg cells for biological studies. Just as IoT systems provide convenience to people's life and help with development in research and industry, genomics can transform precision health and enable tailored treatment plans for patients. This thesis introduces a pruning-based hardware accelerator for Pair Hidden Markov Model, which is a major time-consuming step in the secondary analysis of whole-genome sequencing.