18 May 2017 Context-aware system design
Author Affiliations +
Abstract
The Internet of Things envisions a web-connected infrastructure of billions of sensors and actuation devices. However, the current state-of-the-art presents another reality: monolithic end-to-end applications tightly coupled to a limited set of sensors and actuators. Growing such applications with new devices or behaviors, or extending the existing infrastructure with new applications, involves redesign and redeployment. We instead propose a modular approach to these applications, breaking them into an equivalent set of functional units (context engines) whose input/output transformations are driven by general-purpose machine learning, demonstrating an improvement in compute redundancy and computational complexity with minimal impact on accuracy. In conjunction with formal data specifications, or ontologies, we can replace application-specific implementations with a composition of context engines that use common statistical learning to generate output, thus improving context reuse. We implement interconnected context-aware applications using our approach, extracting user context from sensors in both healthcare and grid applications. We compare our infrastructure to single-stage monolithic implementations with single-point communications between sensor nodes and the cloud servers, demonstrating a reduction in combined system energy by 22-45%, and multiplying the battery lifetime of power-constrained devices by at least 22x, with easy deployment across different architectures and devices.
© (2017) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Christine S. Chan, Christine S. Chan, Michael H. Ostertag, Michael H. Ostertag, Alper Sinan Akyürek, Alper Sinan Akyürek, Tajana Šimunić Rosing, Tajana Šimunić Rosing, } "Context-aware system design", Proc. SPIE 10194, Micro- and Nanotechnology Sensors, Systems, and Applications IX, 101940B (18 May 2017); doi: 10.1117/12.2263232; https://doi.org/10.1117/12.2263232
PROCEEDINGS
17 PAGES


SHARE
Back to Top