University of Virginia Wireless Sensor Networks Sept. 12, 2007 University of Virginia Jack Stankovic
University of Virginia Ad Hoc Wireless Sensor Networks Sensors Actuators CPUs/Memory Radio Minimal capacity 1000s Self-organize
University of Virginia Mica2 and Mica2Dot ATMega 128L 8-bit, 8MHz, 4KB EEPROM, 4KB RAM, 128KB flash Chipcon CC100 multichannel radio (Manchester encoding, FSK). Up to ft.
University of Virginia Sensor Board
University of Virginia Exciting Potential The Internet Gets Physical “Sensing technologies will be one of the hallmarks of this century” 1980 => decade of microcomputers 1990 => decade of the Internet 2000 => decades of WSN
University of Virginia Application Spectrum
University of Virginia Research Overview Application Driven – real world –Building Systems Technology Driven – focused research problems Large Group/Team/Partnering –Recent Past: CMU, Berkeley, UIUC, Washington Univ., Johns Hopkins –Current: UIUC, Minnesota, Harvard, Microsoft, UVA Medical School
University of Virginia Applications/Testbeds AlarmNet – Assisted Living and Residential Monitoring Network Luster - Environmental Science VigilNet – Military surveillance, tracking and classification
University of Virginia AlarmNet - Medical System Architecture InternetInternet Internet PDAs Nurses Stations
University of Virginia
Smart Living Health Spaces
University of Virginia Research Questions in AlarmNet Flexible and Dynamic Privacy Security Form factors for sensor nodes –Unobtrusiveness Mobility –Routing for 2 mobile end points Localization In-network preliminary diagnoses Define and meet real-time requirements including alarms Power Management Data Association
University of Virginia System: Luster Measuring light under shrub thickets Deployment on Hog Island, Eastern Shore of VA
University of Virginia Research Questions in Luster Reliable collection of and dissemination of data even when wireless links are down Deployment time validation Run-time self-healing Scaling
University of Virginia 1. An unmanned plane (UAV) deploys motes 2. Motes establish an sensor network with power management 3.Sensor network detects vehicles and wakes up the sensor nodes Zzz... Energy Efficient Surveillance System Sentry
University of Virginia VigilNet Architecture
University of Virginia Demo System Layout Tent 200 XSM Motes 3 Bases (Tripwires) 300 by 200 Meters in T-shape Inter-tripwire communication Via wireless LAN 300 meters, 30 motes each line, 4 non-uniform lines 200M200M
University of Virginia Results of Actual Test
University of Virginia Overview of Demo Scenarios Tracking multiple targets (people, vehicles, and then people and vehicles) –3 crossing people –Vehicle followed by person –2 vehicles following each other about 50 meters apart –Large versus small vehicles –People and people with weapons Fault Tolerance/Robustness –Kill 20% of the nodes –Kill base stations
University of Virginia For related other publications: Florida
University of Virginia C&C Mote Field N 300M by 200 M T shape Berkeley
University of Virginia Spotlight - Localization μSpotlight (projector, Mica2 motes, laptop) – DEMO at ACM/IEEE IPSN 05 Spotlight (telescope mount, diode laser, XSM motes, laptop) (Sent to Berkeley) Demo at upcoming SenSys 2005
University of Virginia Sentry-Based Power Management (SBPM) Two classes of nodes: sentries and non-sentries – Sentries are awake – Non-sentries can sleep Sentries – Provide coarse monitoring & backbone communication network – Sentries “wake up” non-sentries for finer sensing Sentry rotation – Even energy distribution – Prolong system life
University of Virginia Tripwire-based Surveillance Partition sensor network into multiple sections. Turn off all the nodes in dormant sections. Apply sentry-based power management in tripwire sections Periodically, sections rotate to balance energy. Road Dormant Active DormantActive Dormant
University of Virginia Lifetime Analysis Network Life Time Number of Tripwires (10 regions, 30% sentry, 7 day life) AA Batteries50 days70 days105 days210 days 4 AA Batteries100 days140 days210 days420 days
University of Virginia Internet (Global) Scale WSN Internet Local Transport Protocol Local Transport Protocol Programming Station Server Nodes
University of Virginia System Architecture Internet Local Transport Protocol Local Transport Protocol Programming Station Server Nodes Information about Services, Interfaces Location
University of Virginia System Architecture Internet Programming Station Server Nodes Local Transport Protocol Local Transport Protocol High level Programming Language EXE High Level Virtual Machine High Level Virtual Machine Low Level Virtual Machine Low Level Virtual Machine
University of Virginia System Architecture Internet Local Transport Protocol Local Transport Protocol Programming Station Server Nodes Responsible for Resource management User access rights
University of Virginia System Architecture Internet Local Transport Protocol Local Transport Protocol Programming Station Server Nodes Omnix Physical Network Omnix Physical Network The Physicalnet
University of Virginia Summary - Research Approach Fundamental and Important Problems –Not incremental Application Driven (current) –Military –Medical –Environmental Experimental Systems Research –Some analysis techniques Build Testbeds and Real Systems Cooperate with other Univ.
University of Virginia Summary - Our Research Areas Wireless Networking Realities Localization Real-Time Hardware Privacy Security The crowded spectrum - Multi-frequency systems OS for WSN Spatial Temporal Systems
University of Virginia Our Research Areas (cont.) Power Management Analysis Programming Languages –Across networks of networks Acoustic Streaming and other High Level Services Real-Time Data Sharing Self-Healing Data Association Auto-calibration Pervasive Computing WSN Streams on the Internet
University of Virginia Research Partners UIUC Harvard Univ. of Minnesota UVA Medical School Microsoft Consortium of Universities in Korea
University of Virginia Breakout Session Thurs 4:30 – my office 238E