Exploiting Data-driven Inference Towards Low-energy Implementations in Intelligent Sensors

  • Authors:
    Naveen Verma (Princeton)
    Publication ID:
    P090944
    Publication Type:
    Presentation
    Received Date:
    19-May-2017
    Last Edit Date:
    8-Nov-2017
    Research:
    2385.001 (University of Illinois/Urbana-Champaign)
    2385.002 (Stanford University)
    2385.003 (University of California/Berkeley)

Abstract

For designers of sensor systems, faced with increasingly severe resource constraints (energy, area, bandwidth reliability), the focus on inferences from sensor data, rather than the sensor data itself, is a VERY liberating thing. While sensor data may express inferences of interest through extremely complex correlations, we now know quite broadly that these can be effectively modeled and analyzed through data-driven algorithms. What is liberating is that research in low-power systems is showing that not only can such algorithms be effectively mapped to resource-constrained implementations, but in fact such algorithms can actually relax the implementations themselves. As an example, I describe how data-driven learning enables us to select inference functions and/or parameters that are preferred from the perspective of low-energy implementation and further enables the implementations to exhibit substantially imperfect behaviors. Then, I look at how this can be exploited within systems architectures to alleviate traditional pain points (sensor acquisition, data conversion, memory operations). Measured results from several custom integrated-circuit prototypes are presented.  

4819 Emperor Blvd, Suite 300 Durham, NC 27703 Voice: (919) 941-9400 Fax: (919) 941-9450