4th International ICST Conference on Body Area Networks

Research Article

Optimal Time-Resource Allocation for Activity-Detection via Multimodal Sensing

Download338 downloads
  • @INPROCEEDINGS{10.4108/ICST.BODYNETS2009.6030,
        author={Gautam Thatte and Viktor Rozgic and Ming Li and Sabyasachi Ghosh and Urbashi Mitra and Shri Narayanan and Murali Annavaram and Donna Spruijt-Metz},
        title={Optimal Time-Resource Allocation for Activity-Detection via Multimodal Sensing},
        proceedings={4th International ICST Conference on Body Area Networks},
        publisher={ICST},
        proceedings_a={BODYNETS},
        year={2010},
        month={5},
        keywords={},
        doi={10.4108/ICST.BODYNETS2009.6030}
    }
    
  • Gautam Thatte
    Viktor Rozgic
    Ming Li
    Sabyasachi Ghosh
    Urbashi Mitra
    Shri Narayanan
    Murali Annavaram
    Donna Spruijt-Metz
    Year: 2010
    Optimal Time-Resource Allocation for Activity-Detection via Multimodal Sensing
    BODYNETS
    ICST
    DOI: 10.4108/ICST.BODYNETS2009.6030
Gautam Thatte1,*, Viktor Rozgic1,*, Ming Li1,*, Sabyasachi Ghosh1,*, Urbashi Mitra1,*, Shri Narayanan1,*, Murali Annavaram1,*, Donna Spruijt-Metz2,*
  • 1: Ming Hsieh Department of Electrical Engineering, University of Southern California, USA
  • 2: Keck School of Medicine, University of Southern California, USA
*Contact email: thatte@usc.edu, rozgic@usc.edu, mingli@usc.edu, sabyasag@usc.edu, ubli@usc.edu, shri@sipi.usc.edu, annavara@usc.edu, dmetz@usc.edu

Abstract

The optimal allocation of measurements for activity-level de- tection in a wireless body area network (WBAN) for health- monitoring applications is considered. The WBAN with heterogeneous sensors is deployed in a simple star topol- ogy with the fusion center receiving a fixed number of mea- surements from the sensors; the number of measurements allocated to each sensor is optimized to minimize the prob- ability of detection error at the fusion center. An analysis of the two-sensor case with binary hypotheses is presented. Since the number of measurements is an integer, an exhaus- tive search (grid search) is traditionally employed to de- termine the optimal allocation of measurements. However, such a search is computationally expensive. To this end, an alternate continuous-valued vector optimization is derived which yields approximately optimal allocations which can be found with lower complexity. Numerical case studies based on experimental data for different key activity-states are presented. It is observed that the Kullback-Leibler (KL) distances between the distributions associated with the hy- potheses dominate the optimal allocation of measurements.