Loading...

Introducing a Supervisory Hardware Platform for Wireless Body Area Sensor Network with an Anomaly Detection Ability

Zare Dehabadi, Mohammad Saeed | 2016

654 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 49255 (05)
  4. University: Sharif University of Technology
  5. Department: Electrical Engineering
  6. Advisor(s): Jahed, Mehran
  7. Abstract:
  8. Wireless Body Area Network (WBAN) is a set of various physiological sensors for remote monitoring and controlling of the patient’s health status. Detection of anomalies, containing faults and physiological stresses, and distinguishing between them can eliminate the need for constant presence of specialists and decrease the occurrence of False Alarm Rate (FAR).A supervisory hardware platform for WBAN architecture is introduced and implemented in this thesis. The proposed architecture contains three sensors for recording of biomedical signals, heart rate, SpO2 and body temperature simultaneously and transmitting them to central node wirelessly through Wi-Fi standard. This proposed architecture benefits from a small footprint, low power, low cost, versatile with portable design. For detecting anomalies and distinguishing between them, a univariate, unsupervised, online and real time method with versatile hardware implementation has been proposed. Also a novel transient fault correction method is introduced and compared with previous methods and it is shown that the proposed method performs faster while its accuracy is completely comparable. The proposed method has been simulated on Physionet database and shown that in this method, the fastest sensor can have maximum sampling rate of 833 samples per second. For final evaluation, an experiment is designed and the proposed method has been tested using real experimental data.Finally, for modeling reliability of WBANs, a Markov model is introduced and reliability equation of WBANs is calculated based on patient’s health conditions, sensor fault occurrence rates and accuracy of anomaly detection algorithm
  9. Keywords:
  10. Anomaly Detection ; Real Time Implementation ; Reliability Model ; Wireless Body Area Network ; Fault Detection ; Error Correction ; Hardware Platform

 Digital Object List

 Bookmark

No TOC