lec20_assliving - Real-Time Ubiquitous System: A Case Study...

Info iconThis preview shows page 1. Sign up to view the full content.

View Full Document Right Arrow Icon
This is the end of the preview. Sign up to access the rest of the document.

Unformatted text preview: Real-Time Ubiquitous System: A Case Study "Combined Scheduling of Sensing and Communication for Real-Time Indoor Tracking in Assisted Living" RealLiving" M.-Y. Nam M.CS Dept. UIUC M. Z. Al-Sabbagh C.-G. Lee, J. Kim, M. Yoon AlC.ECE Dept. Ohio State University CS Dept. Seoul National University Columbus, OH Real-Time Ubiquitous System Internet Whenever, Wherever Ubiquitous System Cellular 1 Real-Time Ubiquitous Systems Integrative System: Entertainment, Medical, Sensing, and Communication Devices Integrative Systems that provide real-time interactions with humans and environment Continuous real-time monitoring of human activities realProviding real-time guidance realReal-time emergency detection and handling RealMaking our daily lives more safe and enjoyable How to integrate computing, wireless, and sensing devices providing such real-time interactions? Case Study: Technology-based Assisted Living Where is RC? Did I do this? $70,000 $190,000 in 2030 Expensive Assisted Living Facility Independent Living Where is RC? Here it is. Did I do this? Yes you did. Care Provider Age over 65 will hit 70 millions in 2030 US health-care expenditure healthwill grow to 15.9% of GDP ($2.6 trillion) Internet Independent Living with Technology Assistance 2 Real-Time tracking Human/Objects for Assisted Living Services we wish to provide: Object finding Where did I leave my eye-glasses? eyeReal-Time Tracking of Human/Object can solve Checking if actions were taken Did I eat my pills today? Behavior analysis Early symptom of diseases 3 Our Sensing Infrastructure (Ultrasonic and RFID) Wristband with "Ultrasonic Listener" and Listener" "RFID reader" reader" Tracking user: Ultrasonic Listener on a Wristband Tracking object carried by the user: RFID reader of the Wristband US listener RFID Reader Ultrasonic Sensing Crossbow Cricket Beacon Beacon Beacon Ultrasonic pulse: speed of sound RF: speed of light Listener Localization accuracy: 10cm Real-time tracking: Uses Least square and a Kalman filter with continuous sampling Tracking performance: Sampling frequency 4 RFID Based Sensing SkyeTek RFID Reader and RFID Tags M1-mini RFID Reader RFID Tags Reader Can read within 10 cm proximity Tag Entire System Operation Cricket beacon Cricket beacon Ultrasound RF Cricket beacon Schedule RS232 data Host Computer Cricket Basestation data Cricket Listener Mica2Dot M1Mini Antenna tag ID RFID Tag 5 Scheduling Issue Two types of active signals Sensing: Ultrasonic signals from multiple beacons Communication: Various RF messages from multiple beacons, listeners, listeners, and host computer Original cricket method: Carrier sensing and random arbitration Maximum possible sampling rate is limited Poor real-time tracking real- Our new scheduling method Collision-free scheduling providing much faster sampling CollisionCombined schedule of "Sensing" and "Communication" Sensing" Communication" Location-aware dynamic scheduling: further improvement of sampling rate LocationMobility-conscious scheduling: energy saving Mobility- Pipelining sensing and reporting stages Nave collision-free approach: Sequential schedule Na collisioni-th sensing Ultrasonic band RF band i-th reporting Not Used relay (i+1)th sensing US(B1) RF(B1) Not Used US(B2) RF(B2) report Pipelining sensing and reporting i-th sensing Ultrasonic band RF band (i+1)th sensing US(B1) RF(B1) US(B2) RF(B2) report relay i-th reporting 6 Concurrent Executions B1 B2 B3 B4 Simple round-robin roundB1 B1 B3 B2 B2 B4 B3 B4 Concurrent execution: Static pipeline Concurrent Executions beacon coverage = beacon interference range B1 B2 B3 B4 B1 B2 B3 B4 B5 B8 B9 B6 B10 B7 B11 B8 B5 B9 B6 B10 B7 B11 Beacon placement B1 B2 B3 B4 Conflict graph Beacon Signal Schedule B10 B1 B11 B2 B3 B5 B4 B6 B8 B7 B9 B5 B8 B9 B6 B10 B7 B11 Color Code Schedule Master Schedule Color assignment Static schedule of beacons 7 Location-aware L1 L2 B1 B2 B3 B4 Concurrent execution without location-awareness locationB1 B2 B3 B4 Concurrent execution with location-awareness locationB1 B2 Location-aware scheduling L1 B1 B2 L2 L1 B3 B4 B1 B2 L2 B3 B4 B5 B8 proximity area B6 B9 B10 B7 B11 B8 B5 B9 B6 B10 B7 B11 Beacon placement Beacon Signal Schedule Minimal subset of beacons B3 B1 B2 B5 B6 B3 B1 B1 B2 B3 B5 B6 Color Code Schedule Master Schedule Location-aware conflict graph and coloring Location-aware schedule 8 Mobility-conscious Such high-frequency sampling is not always needed highAdaptive control of sampling rate depending on user mobility slow L1 fast L2 B1 B2 B3 B4 Tracking accuracy requirement: the tracking error should be less than 20 cm in 90% of times. (Transform application-requirement to temporal requirement) Maximum tolerable inter-sampling interval for L1 for L2 B1 B2 B1 B2 B1 B2 B1 B2 B1 B2 Tracking Performance Fundamental Limit of Original Cricket Method Improvement by Our New Method 9 Location-aware Performance By using Locationaware scheduling, we can achieve a higher sampling rate resulting in less error for high mobility listeners Mobility-conscious Performance Tracking accuracy requirement: tracking error should be less than 20 cm with a probability higher than 90% Mobility-aware scheduling can meet the requirement while Mobilitysaving Listener's energy consumption. Listener' 10 Slow Train Tracking (Original Cricket Method) Slow Train Tracking (Our New Method) 11 Fast Train Tracking (Original Cricket Method) Fast Train Tracking (Our New Method) 12 Tracking human Tracking objects carried by human 13 Finding object Any Problem? Line-of-sight limitation of ultrasonic sensing 14 Success-based Self-Coloring Each beacon monitors which colors are free If at least one free color is detected, it tries to capture the color (contention based) Once it successfully capture the color, it senses the target in a contention free manner Coloring consistency 15 Location-aware scheduling (3-D Blocking Problem) Self-Coloring Method (3-D Blocking Solved!) 16 Summary Collision-free scheduling of sensing and communication for real-time tracking Combined scheduling of sensing and communication Location-aware scheduling LocationMobility-conscious scheduling Mobility- What we lean Integration of sensing, wireless, computing, and estimation theory Temporal requirement guarantee for physical interaction points Tremendous Possible Applications Sensing Bio sensors Infrared Sensors Ultrasonic sensors UWB sensors Temperature sensors Smoke sensors Laser sensors Motion sensors Visual sensors Communication WiFi Bluetooth ZigBee Ad Hoc WiBro WiMax Cellular Cable Applications Computing Desktop PC Notebook PDA Cell Phone Digital TV Home Hospital School Airport Stadium Theater Battle field 17 Scientific Challenge Integration of Control, Computing, Communication, and Sensing Physical World System theory (control & estimation) Computers Communication Networks Sensors, Actuators End-to-end timing behavior after integration? System-wide Integration Theory of Control, Computing, Communication, and Sensing For Reliable, Evolvable, Real-Time Ubiquitous Systems 18 ...
View Full Document

Ask a homework question - tutors are online