wireless communications and sensing in e-health · 2017-09-27 · 1 wireless communications and...
TRANSCRIPT
1
Wireless Communications and Sensing in e-Health
Robert J. Piechocki Communications Systems and NetworksSmart Internet Lab
02/12/15 ‘Wireless in the built environment' Radio Technology SIG event
2
Introducing SPHEREA Sensor Platform for Healthcare in a Residential EnvironmentDirector: Prof Ian Craddock• £12M over 5 years (plus £3M from industry and the Universities=£15M)• Led by University of Bristol.• Probably EPSRC’s largest current eHealth IoT research project• A team of over ~100 researchersIn collaboration with Southampton University (Health Sciences), Southampton University (Electrical Engineering), Reading University (Cybernetics), the Elizabeth Blackwell Health Research Institute, Bristol Vision Institute, Department of Experimental Psychology, School of Social and Community Medicine, School of Oral and Dental Sciences, the Centre for Medical Ethics, the Centre for Public Engagement, School of Clinical Sciences, Communications Systems & Networks Group, Intelligent Systems Group, Bristol Heart Institute, Interaction & Graphics Group, Bristol Health Partners, ALSPAC (Children of 90s), Bristol City Council, Knowle West Media Centre, Bristol NIHR Biomedical Research Unit in Nutrition, Diet & Lifestyle, Bristol NIHR Biomedical Research Unit for Cardiovascular Disease, IBM and Toshiba.
irc-sphere.ac.uk
irc-sphere.ac.uk© CSN Group 2015 3
activity monitoring via accelerometers
(&others)
water consumption, electrical consumption
temperature, light levels, humidity, air‐
quality
base‐station, +social media, +encryption
analysis, pattern extraction, feature extraction
data display
VideoEmotion, gait, activity,
interaction
irc-sphere.ac.uk© CSN Group 2015 4
Activity Recognition
© CSN Group 2015 irc-sphere.ac.uk5
Online quality assessment of humanmovement from skeleton data*
Recovered
Injured
© CSN Group 2015 irc-sphere.ac.uk6
7
Wearable IoTs (SPW-1)• Features:
• Dual ultra low power accelerometer• High energy efficiency (battery lifetime up to 1 year)• Improved wireless performance (2dB better than kit ‐ DK)• Energy‐harvesting ready
Ultra low‐power long‐term wireless connectivity
Dual polarization receiver
Energy consumption for SPW-1
irc-sphere.ac.uk© CSN Group 2015
8
What IoT radio should you choose? IEEE 802.15.4 or BLE?
Energy consumption for the transmission of a successful byte (accountingfor potential retransmissions) for various transmission power levels,considering PDUs of 39 bytes.
irc-sphere.ac.uk© CSN Group 2015
• Novel algorithms developed• No extra cost for the wearable
Application‐layer coding
• Any 2 out of 3 correct packets are sufficient to recover both A & B
CRC Error Correction
Interference, coexistence and networking
irc-sphere.ac.uk© CSN Group 2015 9
Interference mitigation: network-wide error correction
• Packet error rate (PER) is reduced by 80%• Traffic to the central hub is reduced by 57%• Computational load is distributed over all devices
irc-sphere.ac.uk© CSN Group 2015 10
Ultra Low Power Voltage Detector
11
RF Rectenna Design
RF Energy harvesting / transfer
irc-sphere.ac.uk© CSN Group 2015
Low power intermittent RF signal: infrequent transmissions with very low power Interference:multiple concurrent 2.4 GHz radios: WiFi, Bluetooth, 15.4, ZigBee Limited data: Only RSSI is readily available; Doppler, timing, angular information
requires additional hardware
RF Indoor localisation and passive sensing
Indoor propagation characteristics examples Classification example
irc-sphere.ac.uk© CSN Group 2015 12
Passive Wireless Sensing with opportunistic Energy Harvesting signal reuse
Stairs climbing activity recognition
• Complement to more traditional techniques: Video & PIR• Capability: activity recognition (walking, stairs climbing, seating
etc) vital signs (respiration)
irc-sphere.ac.uk© CSN Group 2015 13
Extracting Meaning from Data
• Multi‐resident homes• Uncertainty management for multi‐sensor fusion.• Labelled data is expensive.• Unsupervised learning of suitable features.
14 irc-sphere.ac.uk© CSN Group 2015
Conclusions
Key challenges:
• Reliable low power wireless connectivity• Battery free IoT devices• Security solutions for highly constrained devices• Making sense out of the data• User‐research, in the wild
15
[email protected], [email protected] (wireless & sensing)[email protected] (energy harvesting), [email protected] (video),[email protected] (data mining), [email protected] (networking)
Contacts:
irc-sphere.ac.uk© CSN Group 2015