Sensor Networks: Evolution, Opportunities, and ChallengesC. Y. Chong and S. P. KumarProceedings of the IEEE, 91, 8, 2003

Allison LeeCS295-1 Sensor data management

October 12, 2005

Outline

Why is sensor network exciting?Where did it come from?

HistoryExamples

Problems and challengesWhere is it going?Conclusions

Why is sensor network exciting?Business Week: one of the 21 most important technologies for the 21st centurySense, communicate, and compute

Things without our presenceMonitor and gather information

In places not easily accessible

Querying and taskingAct on up-to-date data to develop timely new strategies

Networked sensors deployed all around the worldcontrolling homes, cities, and the environment

Attributes of sensor networks

Versatility of sensor networks

Military sensingPhysical securityVideo surveillanceDistributed roboticsEnvironmental monitoringTraffic control and surveillanceBuilding and structures monitoringIndustrial and manufacturing automation

Where did it come from?

Cold WarSOSUS (Sound Surveillance System) – acoustic sensors

Detect quiet Soviet submarines

Monitor events in the ocean (e.g., seismic and animal activity)Marine Mammal Sounds

Atlantic blue whale call South Pacific blue whale call

Where did it come from? (cont.)Deep Ocean Seismicity from Hydroacoustic Monitoring

earthquake swarm occurred overnight on NE Pacific Endeavour Ridge (2/27/05)maps showing the location of the Endeavour earthquake swarm

HistoryEarly research

Adopt a hierarchical processing structure, information processed in a consecutive orderHuman operators needed

20th centuryDSN (Distributed Sensor Networks)

Provide a network allows flexible and transparent accessNetwork-centric warfare by CCRP (Command and Control Research Program)

Information sharing and collaboration among sensorsEnables self-synchronizationIncrease robustness of sensor networks

21st centurySensIT (Sensor Information Technology) by DARPA (Defense Advanced Research Projects Agency)

Suitable for highly dynamic ad hoc environmentsProvide reliable and timely information

Three Generations of Sensor Nodes

ApplicationsInfrastructure Security

Networks of sensors deployed around facilitiesDetection and tracking of possible threatsEarly warnings and rapid coordinated responses to potential threats Example: Livermore National Laboratory

correlated sensor networks can communicate with each other to ignore false alarms and detect signal not quite at threshold by correlating spatial and temporal information from other sensorsindependent sensors do not have temporal information to discriminate false alarms

More recent applicationsDistributed tracking in wireless ad hoc networks

IDSQ (information-driven sensor querying)

Each sensorComputes the predicted information of a piece of nonlocal sensor dataDetermine from which sensor to request data

Manage resource constraintsDecrease cost of transmitting information

Source tracking in a large-scale sensor network via computations over a sequence of subnetworks formed by the detecting nodes in the vicinity of the source.

Applications (cont.)

Environment and habitat MonitoringSIVAM (System for the Vigilance of the Amazon)

Provides large scale electronic surveillance of Brazil's immenseand relatively undeveloped Amazon region A suite of sensors provides data to

Three regional operations centersOne national operations centerAir surveillance centerALL NETWORKED TOGETHER

SensorsRadar (aircraft), satellite imagery (space), environmental sensors (ground)

Applications (cont.)SIVAM (cont.)

GoalsPrevention and control of epidemicsEnvironmental protection Control of land occupation and use Economical and ecological zoning Mapping Protection of indigenous populations Border surveillance and control Monitoring river navigation Monitoring forest firesLaw enforcement/ drug traffickingAir traffic control and surveillance for both cooperative and non-cooperative aircraft

Other examples of sensor networks

Industrial sensingAccess regions inaccessible by humansMonitor machine health

Spectral sensorsSensors acquire spectra as data

Environmental sensing (electromagnetic spectrum)

To monitor the atmosphere (e.g., greenhouse effect)

Hard problems and challenges

Sensors networksUncertain and dynamic environmentEnergy and bandwidth constraintsProblems in communication, sensor management, and data processing

Ad hoc network discoveryNetwork topology constructed in real time

Adapt to unpredictable environmentKnow the identity and location of neighborsKnow self location

Hard problems and challenges (cont.)

Network control and routingDeveloping self-configuring network system

Deal with changing resource requirements and adapt routing dynamically Provide adequate survival of the network in a changing environmentBalance the tradeoffs between latency, reliability, and energy

Hard problems and challenges (cont.)

Collaborative signal and information processing (balance)

Performance vs. resource utilizationMore sensors processing data, better performance, but use more communication resources (i.e. energy)

Performance vs. robustnessDesign desired algorithm to achieve robustness

E.g., highly accurate or fail-safe resultsWithout sacrificing performance

Hard problems and challenges (cont.)

Tasking and queryingIn sensor network, the data is constantly being acquired, updated, lost and then reacquiredTraditional database style querying may not be adequateNeed a flexible, reliable, and simple interface to query and task the system that account for changes in the sensor network in a timely fashion

SecurityUtmost concern in nearly all applications of sensor networkProtect military intelligence or privacyNeed to be survivable, protected against intrusion and spoofingNot easily detectable by enemies or hackers

Where is it going?Smaller and cheaper sensors, more capable and versatile (to nanobots?)More advanced wireless networks

IEEE 802.15 standardLow energy and high bandwidthLow cost sensor netsDeployed in large numbersProvide a standard way for communication

Highly dynamic ad hoc environmentDynamic and interactive querying and tasking

Wireless network of ubiquitous low-cost disposable microsensors

Smart dust

Conclusions

Sensor NetworksExciting emerging fieldApplications limited only by imaginationCan affect all aspects of human life

Networks of small, possibly microscopic sensors embedded everywhere, even on peoplePerform automated continual and discrete monitoring and tasks

Ethical considerations may become important as sensor networks become ubiquitous and invade privacy

ReferencesChong and Kumar, “Sensor networks: Evolution, opportunities, and challenges.”Proceedings of IEEE, August 2003. “21 ideas for the 21st century,” Business Week, pp. 78-167, Aug. 30, 1999.http://www.businessweek.com/magazine/content/03_34/b3846622.htmhttp://www.businessweek.com/bwdaily/dnflash/sep2004/nf2004081_1700_db_81.htm“10 emerging technologies that will change the world,” Technol. Rev., vol. 106, no. 1, pp. 33-49, Feb. 2003.C. E. Nishimura and D. M. Conlon, “IUSS dual use: Monitoring whales and earthquakes using SOSUS,” Mar. Technol. Soc. J., vol. 27, no. 4, pp. 13-21, 1994. http://www.pmel.noaa.gov/vents/acoustics.htmlProceedings of the Distributed Sensor Nets Workshop Pittsburgh, PA: Dept. Comput. Sci., Carnegie Mellon Univ., 1978. D. S. Alberts, J. J. Garska, and F. P. Stein, Network Centric Warfare: Developing and Leveraging Information Superiority, 1999 [online] http://www.dodccrp.org/html2/research_ncw.htmlS. Kumar and D. Shepherd, “SensIT: Sensor information technology for the warfighter,” in Proc. 4th Int. Conf. on Information Fusion, 2001, pp. TuC1-3-TuC1-9. R. Hills, “Sensing for danger,” Sci. Technol. Rep., July/Aug. 2001 [online] Available: http://www.llnl.gov/str/JulAug01/Hills.html.

ReferencesD. Steere, A. Baptista, D. McNamee, C. Pu, and J. Walpole, “Research challenges in environmental observation and forecasting systems,” in Proc. 6th Int. Conf. Mobile Computing and Networking (MOBICOMM), 2000, pp. 292-299.D. Jensen, SIVAM: Communication, navigation and surveillance for the Amazon, Avionics Mag, June, 2002.http://www.discover-aai.com/technology/spectral.htmB. Deb, S. Bhatnagar, and B. Nath, A topology discovery algorithm for sensor networks with applications to network management: Dept. Comput. Sci., Rutgers Univ., [print] Tech. Rep. DCS-TR-441, 2001. C. Y. Chong, S. Mori, and K. C. Chang, “Distributed multitarget multisensor tracking,”Multitarget Multisensor Tracking: Advanced Applications, Y. Bar-Shalom, Ed. Norwood, MA: Artech House, 1990, pp. 247-295. Y. Yao and J. E. Gehrke, “Query processing in sensors networks,” in Proc. 1st Biennial Conf. Innovative Data Systems Research (CIDR 2003) Asilomar, CA, 2003.IEEE 802.15 Working Group for WPAN [online] Available: http://grouper.ieee.org/groups/802/15/ J. M. Kahn, R. H. Katz, and K. S. J. Pister, “Mobile networking for smart dust,” in Proc. ACM/IEEE Int. Conf. Mobile Computing and Networking (MobiCom), 1999, pp. 271-278.C. Y. Chong, S. Mori, and K. C. Chang, “Distributed tracking in distributed sensor networks,”presented at the Amer. Control Conf. Seattle, WA, 1986.T. Pham and H. C. Papadopoulos, Distributed tracking in ad-hoc sensor networks, IEEE Workshop on Statistical Signal Processing, July 2005

Thank you!