International Journal of Information, Communication Technology and Applications
International Journal of Information, Communication Technology and Applications, ISSN 2205-0930, Volume 2 Number 1 22 September 2016
Copyright © 2016 108
Data ProcessiŶg of Physiological SeŶsor Data aŶd
Alarŵ DeterŵiŶatioŶ UtilisiŶg Activity RecogŶitioŶ
James Jin Kang
School of Information Technology
Deakin University Burwood Australia
jkang@deakin.edu.au
Tom H Luan
School of Information Technology
Deakin University Burwood Australia
tom.luan@deakin.edu.au
Henry Larkin
School of Information Technology
Deakin University Burwood Australia
henry.larkin@deakin.edu.au
Abstract: Current physiological sensors are passive and transmit sensed data to Monitoring
centre (MC) through wireless body area network (WBAN) without processing data intelligently.
We propose a solution to discern data requestors for prioritising and inferring data to reduce
transactions and conserve battery power, which is important requirements of mobile health
(mHealth). However, there is a problem for alarm determination without knowing the activity
of the user. For example, 170 beats per minute of heart rate can be normal during exercising,
however an alarm should be raised if this figure has been sensed during sleep. To solve this
problem, we suggest utilising the existing activity recognition (AR) applications. Most of health
related wearable devices include accelerometers along with physiological sensors. This paper
presents a novel approach and solution to utilise physiological data with AR so that they can
provide not only improved and efficient services such as alarm determination but also provide
richer health information which may provide content for new markets as well as additional
application services such as converged mobile health with aged care services. This has been
verified by experimented tests using vital signs such as heart pulse rate, respiration rate and
body temperature with a demonstrated outcome of AR accelerometer sensors integrated with
an Android app.
Introduction
The rapid prevalence of wearables and body sensors allows additional services to applications
of human activity recognition technologies as seen in Figure 1. AR has become an emerging