Hardware Architectures for Low-Power In-Situ Monitoring of Wireless Embedded Systems
As wireless embedded systems transition from lab-scale research prototypes to large-scale commercial deployments, providing reliable and dependable system operation becomes absolutely crucial to ensure successful adoption. However, the untethered nature of wireless embedded systems severely limits the ability to access, debug, and control device operation after deployment— post-deployment or in-situ visibility. It is intuitive that the more information we have about a system’s operation after deployment, the better/faster we can respond upon the detection of anomalous behavior. Therefore, post-deployment visibility is a foundation upon which other runtime reliability techniques can be built. However, visibility into system operation diminishes significantly once the devices are remotely deployed, and we refer to this problem as a lack of post-deployment visibility. A fundamental factor that limits post-deployment visibility is the resource-constrained nature of these devices, in particular, the severe energy constraints typically present in them. It makes traditional reliability techniques (e.g., modular redundancy) undesirable and even infeasible. In this dissertation, we tackle the key challenge of lack of post-deployment visibility in wireless embedded systems. Specifically, we attempt to answer the following question: “Is it possible to design hardware architectures for wireless embedded systems that enable fine-grained post-deployment visibility, but impose only a minimal (or possibly even zero) power overhead?” We answer this question in the affirmative and propose three different hardware architectures named Spi-Snooper, Senergy, and TeleProbe that enable us to achieve this goal.
Raghunathan, Purdue University.
Off-Campus Purdue Users:
To access this dissertation, please log in to our