Case Studies in Low Power Motion Sensing
| Main Author: | |
|---|---|
| Language: | English |
| Published: |
The Ohio State University / OhioLINK
2009
|
| Subjects: | |
| Online Access: | http://rave.ohiolink.edu/etdc/view?acc_num=osu1250733474 |
| id |
ndltd-OhioLink-oai-etd.ohiolink.edu-osu1250733474 |
|---|---|
| record_format |
oai_dc |
| spelling |
ndltd-OhioLink-oai-etd.ohiolink.edu-osu12507334742021-08-03T05:57:06Z Case Studies in Low Power Motion Sensing DE, DEBRAJ Computer Science sensor network signal human activity radar PIR <p>Wireless Sensor Networks hold great promise as an enabling technology for a variety of applications. Considering for instance wireless networks of motion sensors, which have diverse application in defense, mobility related technologies, clinical studies etc. Performance metrices for such applications include probabilities of event detection, false alarm, classification and mis-classification, detection latency, and lifetime of sensor. In this thesis we address research issues associated with these metrics, in particular the aspects of accurate sensing and power management. Our research focuses on two case studies in motion sensing: presence detection and activity monitoring.</p><p>Presence detection is a primitive of applications of motion sensing such as room occupancy detection. Towards the goal of developing a reliable and long lived conference room occupancy sensing system, we use a Pyroelectric InfraRed (PIR) sensor enabled Trio mote. We develop an occupancy sensing algorithm that shows reliability with no observed false alarms. Power management is achieved through a number of features, including duty cycling and dynamic stabilization. Proper selection of sampling speed and duration enable fast and reliable event detection. The resulting duty cycling algorithm yields an achievable lifetime of 68 days. Based on our analysis we also propose a modification in hardware design for improved lifetime.</p><p>Activity monitoring is a primitive of applications of motion sensing such as tracking human activity level. Towards the goal of developing an energy efficient framework for reliable and accurate human activity level indexing, we use a coherent pulsed doppler radar sensor. We characterize two classes of human motion: uniform gait and milling. Reliability comes with a discrimination algorithm that distinguishes between the motions of zero, one or many people. The algorithm shows rare occurrences of false alarms. We formulated an index of human activity level that proportionally represents motion activity intensity. We propose a power management technique that adapts to activity intensity in order to save energy for sensing. Based on our research, we identify further improvements for more energy efficient, reliable and accurate activity indexing.</p> 2009-09-29 English text The Ohio State University / OhioLINK http://rave.ohiolink.edu/etdc/view?acc_num=osu1250733474 http://rave.ohiolink.edu/etdc/view?acc_num=osu1250733474 unrestricted This thesis or dissertation is protected by copyright: all rights reserved. It may not be copied or redistributed beyond the terms of applicable copyright laws. |
| collection |
NDLTD |
| language |
English |
| sources |
NDLTD |
| topic |
Computer Science sensor network signal human activity radar PIR |
| spellingShingle |
Computer Science sensor network signal human activity radar PIR DE, DEBRAJ Case Studies in Low Power Motion Sensing |
| author |
DE, DEBRAJ |
| author_facet |
DE, DEBRAJ |
| author_sort |
DE, DEBRAJ |
| title |
Case Studies in Low Power Motion Sensing |
| title_short |
Case Studies in Low Power Motion Sensing |
| title_full |
Case Studies in Low Power Motion Sensing |
| title_fullStr |
Case Studies in Low Power Motion Sensing |
| title_full_unstemmed |
Case Studies in Low Power Motion Sensing |
| title_sort |
case studies in low power motion sensing |
| publisher |
The Ohio State University / OhioLINK |
| publishDate |
2009 |
| url |
http://rave.ohiolink.edu/etdc/view?acc_num=osu1250733474 |
| work_keys_str_mv |
AT dedebraj casestudiesinlowpowermotionsensing |
| _version_ |
1719428332087410688 |