Transcript Document
Sensor Network Overview Taekyoung Kwon [email protected] For starters • The problems of engineering education – Problem solving – English – Communication skills For starters • What you can achieve by taking this course – Problem solving • Problem definition – Topics in the wireless/sensor network • Idea • Verify/evaluate – sensor network • Ubiquitous computing • standardization Evolution (size and number) Confluence of technologies Ubiquitous computing • 21st century computers – Embedded in our world (ubiquitous, pervasive) • They weave themselves into the fabric of everyday life until they are indistinguishable from it [Mark Weiser, 1991] • The anti-thesis of “virtual reality” • Like motor technology, embedding computers everywhere and having them “disappear in the background” is easy Wired vs. wireless • Bandwidth • Reliability • CSMA/CD vs CSMA/CA Wireless networks • • • • • • Wireless network ad hoc network Ad hoc network sensor network? Wireless WAN: Cellular Wireless MAN: IEEE 802.16 Wireless LAN: IEEE 802.11 series Wireless PAN: IEEE 802.15 family What is sensor? • Sensor: a transducer that converts a physical, chemical, or biological parameter into an electrical signal • Actuator: a transducer that accepts an electrical signal and converts it into a physical, chemical, or biological action • Transducer: a device converting energy from one domain into another. The device may either be a sensor or an actuator Sensor network Internet, Satellite, etc Sink Sink Task Manager • Tens of thousand nodes – Densely deployed Sensor node hardware Mobilizer Location Finding System Processor Transceiver Sensor ADC • • • • • Memory • • Power Unit Small Low power Low bit rate High density Low cost (dispensable) Autonomous Adaptive Sensor network • Power constraint – Battery powered mains powered – Energy harvest • Light(solar), vibration, temperature • Tradeoff between energy and QoS – Prolong network lifetime by sacrificing application requirements • Delay, throughput, reliability, data fidelity,… – Still QoS is attractive • Deterministic or probabilistic bound Sensor network • Traffic type: streaming, periodic, event • Low cost, Low bit rate, low duty cycle • IEEE 802.15.4: 250Kbps Data Link Layer Physical Layer Task Management Plane Network Layer Mobility Management Plane Transport Layer Power Management Plane Application Layer Ad hoc vs. sensor • Number of sensor nodes can be several orders of magnitude higher • Sensor nodes are densely deployed and are prone to failures • The topology of a sensor network changes very frequently due to node mobility and node failure • May leverage broadcasting than point-to-point communications • May operate in aggregate fashion • In-network processing • Sensor nodes are limited in power, computational capacities, and memory • May not have global ID like IP address • Need tight integration with sensing tasks Design issues • Fault tolerance – Battlefield application • Scalability – Node density: (NR^2)/A (transmission) • Production costs • Hardware constraints • Topology – Deployment phase – Post-deployment phase • Environment • Transmission media: ISM, IR • Power consumption: sensing, processing, communication PHY layer • Sync • Self-organization – Beacon scheduling (periodic) • Directional/smart antenna • Ultra-wideband (UWB) • Transmit-only device – pros: cost, energy – Cons: uncontrollable, communications/networking overhead MAC layer • TDMA vs. CSMA – TDMA: inter-cluster, scalability – CSMA: idle listening, overhearing • Sleep cycle • Coordination – Spatial correlation – Clustering (MAC vs NWK) • Additional control channel – FDMA or TDMA • Location awareness – Exposed terminal problem network layer • Attribute-based addressing – Information-centric delivery • Routing – Route discovery • Data aggregation/coordination • Location awareness – Directional antenna (AOA) – UWB (distance measure via signal flight time) – GPS routing • Route discovery (AODV, DSR,…) – Route selection metric: hop count – Metric can be generalized to cost • Hierarchical tree routing • Gradient routing: data broadcasting Transport layer • Goodput decreases drastically as the offered traffic exceeds the network capacity • Flow control vs. Congestion control – open loop vs closed loop – Proactive vs. reactive Transport layer • Reliability concept should be relaxed – Event-to-sink reliability • Not all event-sensing nodes need to report • N reception among M transmission might be OK (M > N) • Hop-by-hop approaches Middleware/Language/Appl. • query/advertisement – Publish/subscribe • nesC, Mate, SQTL – Declarative rather than procedural – TEDS (IEEE 1451) Some of the commercial applications – Industrial automation (process control) – Defense (unattended sensors, real-time monitoring) – Utilities (automated meter reading), – Weather prediction – Security (environment, building etc.) – Building automation (HVAC controllers). – Disaster relief operations – Medical and health monitoring and instrumentation What to consider: application requirements • Energy-saving • QoS – Throughput/Goodput – Reliability – timeliness • Traffic/application scenario – Amdahl’s law – Every possible case • Self-organization What to consider: enabling technologies • Directional (smart, MIMO) antenna – Multi-hop reachability – AoA – Hidden node problem • Heterogeneous node type – E.g., Transmit-only device • GPS: too costly • UWB (distance measurement) – Location aware • Energy harvesting device • Additional (separate) control channel Possible approaches • • • • Conservative vs. aggressive Pessimistic vs. opportunistic vs. optimistic Proactive (a priori) vs reactive (on demand) Information amount vs. performance (better control/decision) – History – Neighbors within some hops • Deterministic (e.g. threshold) vs. probabilistic – N * p = 1? • Reservation vs. random access • Heterogeneous functionalities – E.g, cluster head, member Possible enhancements: • Flexibility vs. efficient – adaptivity • Stability vs. throughput (utilization) – Goodput • Reliable vs. fault-tolerant vs. error-resilient vs. robust • fairness • Legacy-system support, standard-compliant, backward compatibility Final goal • • • • Tradeoff Quantitative trend Qualitative feature How to verify? – Analysis – Simulation – Implementation analysis • • • • • • assumptions Whole system vs key element Steady state probability Upper/lower bound Worst/average case Complexity: O() – Temporal vs. spatial Simulation • Arbitrary level of detail • Still too many ambiguities – Follow the norm, other reference • How to emphasize the strength? • Also show the weakness Implementation • Most time and energy consuming • Good luck! Leverage other techniques • Algorithm • Combination theory • AI – e.g., self-learning • Operations Research – optimization • Network Flow, scheduling theory • Probability – Queuing theory Let’s make team!