Overview
- 1. Oveview of Wireless Sensor Networks KD Kang
- 2. Overview • What is a sensor network? – Sensing – Microsensors – Constraints, Problems, and Design Goals • Overview of Research Issues and Challenges
- 3. Applications
- 4. Applications • Interface between Physical and Digital Worlds – Many applications • Military – Target tracking/Reconnaissance – Weather prediction for operational planning – Battlefield monitoring • Industry: industrial monitoring, fault-detection… • Civilian: traffic, medical… • Scientific: eco-monitoring, seismic sensors, plume tracking…
- 5. Microsensors for in-situ sensing • Small • Limited resources – Battery powered – Embedded processor, e.g., 8bit processor – Memory: KB—MB range – Radio: Kbps – Mbps, tens of meters – Storage (none to a few Mbits)
- 6. Redwood trees: An application
- 7. Mica2 Mote Chipcorn 128KB Instruction CC1000 UART Flash EEPROM Radio Transciever Memory Max 38Kbps 128KB – 512KB - Lossy transmission 4KB Data RAM Atmega 128L microprocessor 7.3827MHz UART, ADC 51 pin expansion connector
- 8. Objective • Large-scale, fine-grained, heterogeneous sensing – 100s to 1000s of nodes providing high resolution – Spaced a few feet to 10s of meters apart – In-situ sensing – Hetegerogeneous sensors
- 9. Properties • Wireless – Easy to deploy: ad hoc deployment – Most power-consuming: transmiting 1 bit ≈ executing 1000 instructions • Distributed, multi-hop – Closer to phenomena – Improved opportunity for LOS – radio signal is proportional to 1/r4 – Centralized apporach do not scale – Spatial multiplexing • Collaborative – Each sensor has a limited view in terms of location and sensor type – Sensors are battery powered – In-network processing to reduce power consumption and data redundancy
- 10. Basic Terminology and Concepts • Phenomenon: the physical entity being monitored • Sink or base station: a collection point to which the sensor data is disseminated – Relatively resource rich node • Sensor network periodically samples phenomena in space and time • Sink floods a query
- 11. Typical Scenario Deploy Wake/Diagnosis Self-Organize Disseminate
- 12. Other variations • Sensors mobile or not? • Phenomena discrete or continuous? • Monitoring in real-time or for replay analysis? • Ad hoc queries vs. long-running queries
- 13. Protocol Stack
- 14. Alternative, more data-centric model
- 15. Protocol Stack: Physical Layer • Frequency selection • Carrier frequency generation • Signal detection • Modulation • Not the focus of this class – We will focus on the link layer and above
- 16. Protocol Stack: Physical Layer • Issues – Hardware cost • How do we get down to $1/node? – Radio • Ultrawideband? – Very low powered, short pulse radio spread over several GHz – 40Mbps ~ 600Mbps
- 17. Protocol Stack: Physical Layer • Radio (Cont.) – Zigbee/IEEE 802.15.4 • 2.4GHz radio band (= 802.11.b & Bluetooth) • 250Kbps • Up to 30 meters – Pico radio • 100Kbps • Limit power consumption to 100 uW – Other? (infra red, passive elements …)
- 18. Protocol Stack: Data Link Layer • Multiplexing of data streams • Data frame detection • Medium access • Error control
- 19. Data Link Layer • Goals: – Creation of the network infrastructure – Fair and efficient sharing of of communication resources between sensor nodes • Existing solutions? – Cellular: single hop network is impractical for sensor networks – Ad hoc MACs, e.g., 802.11 or Bluetooth: Power consuming – Scale – Data centric operation – Security • WEP for 802.11 is broken
- 20. Data Link Layer: Medium Access Control • Basic strategy: turn off radio transceiver as much as possible, while receiving and transmitting data • Techniques: TDMA, application-layer transmission scheduling, SMAC, ZMAC, BMAC, ...
- 21. Protocol Stack: Network Layer • Design principles – Power efficiency – Data-centric – Data aggregation when desired and possible – Attribute-based addressing and location awareness: no IP address
- 22. Minimum Energy Routing • Maximum power abailable route • Minimum energy route • Minimum hop (MH) route
- 23. Directed Diffusion • Route based on attributes and interests • To be covered later in the semester
- 24. Network Layer • Data-centric routing – Directed Diffusion – Data Aggregation • Flooding • Gossiping/non-uniform dissemination • Geographic routing
- 25. Transport Layer • End-to-end Reliability – Multi-hop retransmission: worth it? – Congestion: relatively little related work • End-to-end security – Like SSL: authentication, encryption, data integrity – Good? What about data aggregation?
- 26. Protocol Stack: Application Layer • Actual WSN applciations • Sensor database – TinyDB – Cougar – SINA • Virtual machines/middleware
- 27. Other Important Issues • Operating system – TinyOS – MANTIS OS • Localization, Time Synchronization, and Calibration • Aggregation/Data Fusion • Security – Encryption – Authentication – Data integrity – Availability: DOS attacks • Privacy
- 28. Time and Space Problems • Timing synchronization • Node Localization • Sensor Coverage
- 29. Time Synchronization • Time sync is critical at many layers in sensor nets – Aggregation, localization, power control Ref: based on slides by J. Elson
- 30. Sources of time synchronization error • Send time – Kernel processing – Context switches – Transfer from host to NIC • Access time – Specific to MAC protocol • E.g. in 802.11, sender must wait for CTS (Clear To Send) • Propagation time – Dominant factor in WANs • Router-induced delays – Very small in LANs • Receive time • Common denominator: non-determinism
- 31. Conventional Approaches • GPS at every node (around 10ns accuracy) – doesn’t work indoo – cost, size, and energy issues • NTP – Primary time servers are synchronized via atomic clock – Pre-defined server hierarchy – Nodes synchronize with one of a pre-specified time servers – Can support coarse-grain time synchronization • Inefficient when fine-grain sync is required – Sensor net applications, e.g., localization, beamforming, TDMA – Discovery of time servers – Potentially long and varying paths to time-servers – Delay and jitter due to MAC and store-and-forward relaying
- 32. Localization • Why each node should find its location? – Data meaningless without context – Geographical forwarding/addressing • Why not just GPS at every node? – Large size and expensive – High power consumption – Works only outdoors with LOS to satellites – Overkill: Often only relative position is needed
- 33. What is Location? • Absolute position on geoid • Location relative to fixed beacons • Location relative to a starting point – e.g. inertial platforms • Most applications: – location relative to other people or objects, whether moving or stationary, or the location within a building or an area
- 34. Techniques for Localization • Measure proximity to beacons – Near a basestation in a room • Active Badge for indoor localization – Infrared basestations in every room – Localizes to a room as room walls act as barriers • Most commercial RF ID Tag systems – Strategically located tag readers – Beacon grid for outdoor localization • Estrin’s system for outdoor sensor networks – Grid of outdoor beaconing nodes with know position – Position = centroid of nodes that can be heard – Problem • Not location sensing but proximity sensing • Accuracy of location is a function of the density of beacons
- 35. Localization • Measure direction of landmarks – Simple geometric relationships can be used to determine the location by finding the intersections of the lines-of- position – e.g. Radiolocation based on angle of arrival (AoA) • can be done using directional antennas or antenna arrays • need at least two measurements BS φ2 BS φ1 MS φ3 BS
- 36. Localization: Range-based • Measure distance to beacons – Measure signal-strength or time-of-flight – Estimate distance via received signal strength • Mathematical model that describes the path loss attenuation with distance • Use pre-measured signal strength contours around fixed beacon nodes – Distance via Time-of-arrival (ToA) • Distance measured by the propagation delay – Distance = time * c • Active vs. passive – Active: receiver sends a signal that is bounced back so that the receiver know the round-trip time – Passive: receiver and transmitter are separate » time of signal transmission needs to be known – N+1 BSs give N+1 distance measurements to locate in N dimensions
- 37. Radiolocation via ToA and RSSI x2 BS d2 BS x1 SENSOR d1 d3 BS x3
- 38. Many other issues • What about errors? Collisions? No LOS? • If sensors are mobile, when should we localize? • Multi-hop localization?
- 39. Sensor Network Coverage GATEWAY MAIN SERVER CONTROL CENTER • The Problem: – Given: • Ad hoc sensor field with some number of nodes with known location • Start and end positions of an agent – Want: • How well can the field be observed? • Example usage – Commander • Weakest path: what path is the enemy likely to take? – Network manager • Weakest path: where to deploy additional nodes for optimum coverage? – Soldier in the battlefield • Strongest path: what path to take for maximum coverage by my command? • Weakest path: how to walk through enemy sensor net or through minefield? Ref: based on slides by Seapahn Megerian
- 40. Exposure Model of Sensors • Likelihood of detection by sensors is a function of time interval and distance from sensors. • Minimal exposure paths is worst case scenarios in a field: Ref: based on slides by Seapahn Megerian
- 41. Other Issues • Coverage for continuous phenomena • Role: Sensor as a source of information and as a router – What if we route through a critical node and drain its energy compromising future coverage?
- 42. Data Management Problems sensors • Observer interested in phenomena with certain tolerance – Accuracy, freshness, delay • Sensors sample the phenomena • Sensor Data Management – Determining spatio-temporal sampling schedule observer • Difficult to determine locally – Data aggregation and fusion phenomena • Interaction with routing – Network/Resource limitations • Congestion management • Load balancing • QoS/Realtime scheduling
- 43. Spatio-Temporal Sampling • How to express interests and translate into actions • How often should a given sensor report? – Collected data should meet application goals at reasonable load to the network – Data/event driven – Locally difficult to determine appropriate sampling/reporting rate – Collaboration needed to improve local estimate
- 44. Data Aggregation and Fusion • Aggregate related data from multiple sensors – Reduce data size and overall load – Provide more comprehensive estimate of data importance to manage sensors better • Support for effective data aggregation – Routing and MAC support – Sampling schedules should be coordinated – Tradeoff between data quality and resource demand should be exposed to the application
- 45. Summary: Key Design Challenges • Energy efficiency – Sensor nodes should run for several years without battery replacement – Energy efficient protocols are required – More efficient batteries • But, efficient battery development is always slower than processor/memory development – Energy harvesting
- 46. Key Design Challenges • Responsiveness – Periodic sleep & wake-up can reduce the responsiveness of sensors • Important events could be missed – In real-time applications, the latency induced by sleep schedules should be kept within bounds even when the network is congested
- 47. Key Design Challenges • Robustness – Inexpensive sensors deployed in a harsh physical environment could be unreliable • Some sensor could be faulty or broken – Global performance should not be sensitive to individual sensor failures – Graceful performance degradation when there are faulty sensors
- 48. Key Design Challenges • Synergy – Moore’s law apply differently • Sensors may not become more powerful in terms of computation and communication capability • Cost reduction is the key to a large number of sensor deployment – A WSN as a whole needs to be much more capable than a simple sum of the capabilities of the sensors • Extract information rather than raw data – Also support efficient collaborative use of computation, communication, and storage resources
- 49. Key Design Challenges • Scalability – 10,000 or more nodes for fine-granularity sensing & large coverage area – Distributed, localized communication – Utilize hierarchical structure – Address fundamental problems first • Failure handling • In-situ reprogramming, e.g., Deluge • Network throughput & capacity limits?
- 50. Key Design Challenges • Heterogenity – Heterogeneous sensing, computation, and communication capabilities – e.g., a small number of devices of higher computational capabilities & a large number of low capability nodes -> two-tier WSN architecture – Best architecture exist for all application? – How to determine a right combination of heterogeneous devices for a given application?
- 51. Key Design Challenges • Self-configuration – WSNs are unattended distributed systems – Nodes have to configure their own network topology; • Localize, synchronize & calibrate; and • Coordinate communications for themselves
- 52. Key Design Challenges • Self-optimization & adaptation – WSNs cannot be optimized a priori – Environment is unpredictble, and may change drastically – WSN protocols should be adaptive & adapt themseleves online
- 53. Key Design Challenges • Systematic design – Tradeoff btwn two alternatives • (1) Fine-tuning to exploit application specific characteristics to improve performance • (2) More flexible, easy-to-generalize design approaches sacrificing some performance – Systematic design methodologies for reuse, modularity & run-time adaptation are required
- 54. Key Design Challenges • Security & Privacy – Security support for critical applications, e.g., battlefield monitoring – Avoid sabotage in, e.g., structural monitoring – Support privacy of medical sensor data – Severe resource limitations, but challenging security & privacy issues
- 55. Questions?