Ad hoc and Sensor Networks Chaiporn Jaikaeo chaiporn.j@ku.ac.th - PowerPoint PPT Presentation

Wireless Embedded Systems (0120442x) Ad hoc and Sensor Networks Chaiporn Jaikaeo chaiporn.j@ku.ac.th Department of Computer Engineering Kasetsart University Materials taken from lecture slides by Karl and Willig

Typical Wireless Networks Base stations connected to wired backbone  Mobile nodes communicate wirelessly to  base stations

Ad hoc Networks Networks without pre-configured infrastructure  require no hubs, access points, base stations  are instantly deployable  can be wired or wireless  wired wireless multi-hop wireless Initially targeted for military and emergency  applications

802.11 Ad hoc Mode IEEE 802.11 already provides support for  ad hoc mode Computers can be connected without an  access point Only work with single hop 

Possible Applications for Ad hoc Networks Factory Floor Disaster recovery Car-to-car communication Automation

Characteristics of Ad hoc Networks Heterogeneity ― sensors, PDAs, laptops  Limited resources ― CPU, bandwidth, power  Dynamic topology due to mobility and/or failure  Mobile Ad hoc Networks (MANETs)  B C A

Sensor Networks Participants in the previous examples were  devices close to a human user, interacting with humans Alternative concept:  Instead of focusing interaction on humans, focus on interacting with enviro ronme nment nt Network is embedde ded d in environment  Nodes in the network are equipped with sensin ing g and  actuati ation on to measure/influence environment Nodes process information and communicate 

Traditional Sensors Network sensor field Remote monitoring Data loggers Local monitoring

Wireless Sensors Sensors communicate with data logger via  radio links Network sensor field Remote monitoring radio link

Wireless Sensor Networks Wireless sensors + wireless network  Sensor nodes (motes) deployed and forming an ad hoc  network Requires no hubs, access points  Instantly deployable  Sensor node/mote Sensor network Gateway Targeted applications Internet  Emergency responses  Remote Remote data acquisition  monitoring

WSN Platforms Most are based on IEEE 802.15.4 (Wireless  Low-Rate Personal Area Network) and many others…

WSN Application Examples Agriculture  Humidity/temperature  monitoring Civil engineering  Structural response  Disaster management  Environmental sciences  Habitat monitoring  Conservation biology 

WSN in Telemetry Applications Browser sensor sensor Information Server wireless sensor node GPRS Network Sensor field or Internet Gateway

Landslide Monitor Real deployment scenario… 

Roles of Participants in WSN Sources rces of data: Measure data, report them “somewhere”  Typically equip with different kinds of actual sensors  Sinks ks of data: Interested in receiving data from WSN  May be part of the WSN or external entity, PDA, gateway, …  Actua uato tors rs (actor ors) s) : Control some device based on data,  usually also a sink WSN = WASN

Classifying Application Types  Inte teract raction ion pa patt tter erns ns between sources and sinks classify application types  Event detection  Periodic measurement  Function approximation  Edge detection  Tracking

Deployment Options Dropped from aircraft   Rando dom m deployme oyment nt Well planned, fixed   Regular ar deploy oyme ment nt Mobi bile le sensor nodes   Can move to compensate for deployment shortcomings  Can be passively moved around by some external force (wind, water)  Can actively seek out “interesting” areas

Maintenance Options Feasible and/or practical to maintain  sensor nodes?  Replace batteries  Unattended operation  Impossible but not relevant Energy supply   Limited from point of deployment  Some form of recharging / energy scavenging

Characteristic Requirements Type of service of WSN  Not simply moving bits like another network  Rather: provide answers (not just numbers)  Geographic scoping are natural requirements  Quality of service  Fault tolerance  Lifetime: node/network  Scalability  Wide range of densities  Programmability  Maintainability 

Required Mechanisms Multi-hop wireless communication  Energy-efficient operation   Both for communication and computation, sensing, actuating Auto-configuration   Manual configuration just not an option Collaboration & in-network processing   Nodes in the network collaborate towards a joint goal  Pre-processing data in network (as opposed to at the edge) can greatly improve efficiency

Required Mechanisms Data centric networking   Focusing network design on data, not on node identifies (id-centric networking)  To improve efficiency Locality   Do things locally (on node or among nearby neighbors) as much as possible Exploit tradeoffs   E.g., between invested energy and accuracy

MANET vs. WSN - Similarities MANET – Mobile Ad hoc Network  Self-organization  Energy efficiency  (Often) Wireless multi-hop 

MANET vs. WSN - Differences Eq Equipment: ipment: MANETs more powerful  Applicati lication on-spe specific: cific: WSNs depend much  stronger on application specifics Environ En ironment ment in interacti raction: on: core of WSN,  absent in MANET Sca cale: le: WSN might be much larger (although  contestable) En Energy: rgy: WSN tighter requirements,  maintenance issues

MANET vs. WSN - Differences Dependabili pendability/QoS: ty/QoS: in WSN, individual node  may be dispensable (network matters), QoS different because of different applications Addressing: dressing: Data centric vs. id-centric  networking

Enabling Technologies for WSN Cost reduction   For wireless communication, simple microcontroller, system on chip, sensing, batteries Miniaturization   Some applications demand small size  “Smart dust” as the most extreme vision Energy scavenging   Recharge batteries from ambient energy (light, vibration, …)

Conclusion MANETs and WSNs are challenging and  promising system concepts Many similarities, many differences  Both require new types of architectures &  protocols compared to “traditional” wired/wireless networks In particular, application-specificness is a  new issue

Demonstration

Sensor Modules IWING-MRF modules from IWING LAB  250 kbps 2.4GHz IEEE 802.15.4  12MHz Atmel ATMega328P microcontroller  Additional light and temperature sensors 

Scenario Monitor station Sensor nodes measuring light intensity