MobileIPMobileIP11Sensor NetworksSensor Networks CEN 5531CEN 5531Dr. Sumi Helal & Jeff KingDr. Sumi Helal & Jeff KingComputer & Information Science & Engineering DepartmentComputer & Information Science & Engineering DepartmentUniversity of Florida, Gainesville, FL 32611University of Florida, Gainesville, FL 32611Phone: (352) 392-6845Phone: (352) [email protected] [email protected] Slides adopted from presentations by Kirill Mechitov, David Culler,Joseph Polastre, Robert Szewczyk, Cory Sharp.MobileIPMobileIP22OutlineOutlineWhat are Sensors and Sensor What are Sensors and Sensor Networks?Networks?MotivationMotivationThe Berkeley MOTEThe Berkeley MOTE–TinyOS & NesCTinyOS & NesCThe UF Sensor PlatformThe UF Sensor PlatformLab DemoLab DemoMobileIPMobileIP33Wireless “Smart” SensorsWireless “Smart” SensorsRef: Kirill MechitovMobileIPMobileIP44Network Embedded Network Embedded SystemsSystemsEmbedded systems are the future of Embedded systems are the future of computingcomputing–Low-cost hardwareLow-cost hardware–Many application areasMany application areas–Already more embedded processors than Already more embedded processors than desktops, and the gap is increasingdesktops, and the gap is increasingMost embedded processors act Most embedded processors act independentlyindependently–Industrial control systems, u-controllers in cars, Industrial control systems, u-controllers in cars, etc.etc.Connecting embedded systems into Connecting embedded systems into networks enables new class of applicationsnetworks enables new class of applicationsRef: Kirill MechitovMobileIPMobileIP55Network Embedded Network Embedded SystemsSystemsLow-power, inexpensive embedded Low-power, inexpensive embedded processors cannot perform very processors cannot perform very complex taskscomplex tasksBut a network of such systems can be But a network of such systems can be very powerfulvery powerfulExample: sensor networksExample: sensor networks–Each processor is equipped with a sensorEach processor is equipped with a sensor–Becomes a “smart” sensor nodeBecomes a “smart” sensor nodeRef: Kirill MechitovMobileIPMobileIP66Sensor NetworksSensor NetworksData from multiple sensors is processed and Data from multiple sensors is processed and combined into “big picture”combined into “big picture”Sensor coverageSensor coverage–Sensors can be deployed to cover a large areaSensors can be deployed to cover a large areaReliabilityReliability–Redundant sensor readingsRedundant sensor readings–Resiliency to failure of individual sensorsResiliency to failure of individual sensorsCostCost–Many inexpensive sensors can be cheaper than Many inexpensive sensors can be cheaper than one powerful sensorone powerful sensorRef: Kirill MechitovMobileIPMobileIP77Wireless Sensor NetworksWireless Sensor NetworksSensor networks where nodes communicate Sensor networks where nodes communicate over wireless channelsover wireless channelsEase of deploymentEase of deployment–No infrastructure requiredNo infrastructure required–Once placed in the area of interest, sensors Once placed in the area of interest, sensors automatically assemble into an “ad hoc” networkautomatically assemble into an “ad hoc” networkReliabilityReliability–No wires that can be damaged or cutNo wires that can be damaged or cutRef: Kirill MechitovMobileIPMobileIP88WSN in the labWSN in the labRef: Kirill MechitovMobileIPMobileIP99WSN in the FieldWSN in the FieldRef: Kirill MechitovMobileIPMobileIP1010WSN EnvironmentWSN EnvironmentLarge-scale systems where:Large-scale systems where:Nodes and links have limited capabilities.Nodes and links have limited capabilities.Real-time requirements must be met in the Real-time requirements must be met in the absence of a predefined global clock. absence of a predefined global clock. Faults are common.Faults are common.failuresDelaysRef: Kirill MechitovMobileIPMobileIP1111ChallengesChallengesDistributed processingDistributed processing–Need algorithms that are not centralized, i.e., do Need algorithms that are not centralized, i.e., do not require all of the datanot require all of the dataLow bandwidth communicationLow bandwidth communication–Efficiently move large amounts of sensor data for Efficiently move large amounts of sensor data for processingprocessingLarge scale coordinationLarge scale coordination–Many independent sensor nodes need to act in Many independent sensor nodes need to act in concert with one anotherconcert with one anotherReal-time computationReal-time computation–New data is always coming, so it must be New data is always coming, so it must be processed faster than it is generatedprocessed faster than it is generatedRef: Kirill MechitovMobileIPMobileIP1212Routing and Group Routing and Group CommunicationCommunicationRouting delivers messages to a specific node in Routing delivers messages to a specific node in the networkthe network–Multi-hop, ad hocMulti-hop, ad hoc–Old problem, but needs newOld problem, but needs newapproach in the sensor network environmentapproach in the sensor network environmentGroup communication (multicast) delivers Group communication (multicast) delivers messages to a subset of nodes in the networkmessages to a subset of nodes in the network–Needed to communicate to groups of sensorsNeeded to communicate to groups of sensorsParameters: reliability, efficiency,Parameters: reliability, efficiency,power consumptionpower consumptionRef: Kirill MechitovMobileIPMobileIP1313Data AggregationData AggregationCombines data from many sensors into a Combines data from many sensors into a more compact form before forwarding to a more compact form before forwarding to a location for processinglocation for processingNeeded to handle the large amount of data Needed to handle the large amount of data generated in sensor networksgenerated in sensor networksParameters: efficiency, speedParameters: efficiency, speedtraffic vs. distance from sinkwithout data aggregationAggregationForwardingvs.Ref: Kirill MechitovMobileIPMobileIP1414Clock SynchronizationClock SynchronizationSensors’ clocks drift slightly over timeSensors’ clocks drift slightly over timeNeed to periodically adjust the local clocks so Need to periodically adjust the local clocks so that time is consistent throughout the networkthat time is consistent