IJCER (www.ijceronline.com) International Journal of computational Engineerin...
Wireless Sensor Networks ,Middleware and Programming abstractions
1. Middle ware and Programming
Paradigms for WSN
Budwell Tamuka Masaiti
February 23, 2013
JNTUCEH(CSE Department)
3. Agenda
Wireless Sensor Networks :Introduction
Applications of WSNs
Main Issues in WSN
Middleware and Programming Abstractions for WSN
Goals for Middleware and Programming abstractions
OSM : Object State model (FSM)
Contiki: Event driven abstractions
TinyOs: Event driven abstractions
Conclusion
4. Wireless Sensor Networks :Introduction
A wireless sensor network (WSN) consists of
spatially distributed autonomous sensors to
monitor physical or environmental conditions,
such as temperature, sound, pressure, etc. and
to cooperatively pass their data through the
network to a main location.
5. Sensor Node
Transceiver
10Kbps-1Mbps 50-125 range
Memory
(3K-1Mb)
Embedded Processor
A 8bit 4-8Mhz
D
Sensor C
Transducer Battery
12. WSN Middleware and
Programming Abstractions
Middleware provides the software layer interfacing the
hardware and the application programs in an OS less
system
Or the software layer between the OS and the
application programs.
Programming abstractions refer to the interfaces and
data structures provided for a programming model and
language
14. Object State Model
Oliver Kasten n Kay Romer
A programming model and language for programming
sensor nodes, based on FSM
Uses states and transitions such that the invocation of an
event is now a function of both the event and the
program state
State attributes (infor sharing )
15. Object State Model(OSM)
OSM goes beyond traditional event driven
programming and draws inspiration from FSM state
charts.
OSM is a hybrid FSM implementation which tackles FSM
problems such as:
Memory Management(Global and local variables)
Event hierarchies
Manual flow control
16. OSM Main Features
Uses Parallel State Machines
Allows hierarchical composition of state(superstates)
Explicit program state(not implied)
State variables
Event queues
17. Contiki
Background:
Contiki – pioneering open source operating system for
sensor networks
Development started in 2001
Got its name from Kon Tiki
Across the Atlantic with bare minimum
resources
18. Contiki Key Features
Small memory footprint
IP networking
Hybrid threading model, protothreads
Power profiling
-measure power consumption at network scale
Network shell
- for easy command line interaction
Designed for portability
Dynamic loading
19. Contiki Communication Stack
Two communication stacks
uIP TCP/IP
Rime low overhead
Can run on top of each other
20. Contik :Event Driven Paradigm
The Contiki kernel is event-based
Invokes processes whenever something happens
Sensor events, processes starting,
Process invocations must not block
Protothreads provide sequential flow of control in
Contiki processes
Protothreads extremely lightweight and stackless
21. Contik :Event Driven
Event-driven vs multithreaded
Event-driven requires less memory
Multithreading requires per-thread stacks
Threads require per-thread stack memory
Events require one stack
Protothreads like events require one stack for as many
as there are protothreads running
22. TinyOs:Event Driven
An event driven OS specifically designed for WSN
Completely non-blocking
Programs are built out of softaware componets
Tasks are non –preemptive and run in FIFO order
Code is statically linked
23. TinyOS :Scheduling
Two level scheduling: events and tasks
Scheduler is simple FIFO
A task cannot preempt another task
Events preempt tasks (higher priority)
24. Provides a specific Service
Message Handling, Signal Processing
Implemented in a Module(code)
Wired up of other components in a Configuration
26. Conclusion
Unique challenges and opportunities
Expanding areas of applications
Contik and TinyOs
New research directions
27. references
Geoff Werner-Allen, Konrad Lorincz, and Matt Welsh et al .Monitoring Volcanic
Activity at Reventador Volcano, Ecuador with a Wireless Sensor Network
Oliver Kansten and Kay Romer. Beyond Event Handlers:Programming Wireless
Sensors with Attriunuted State machines
A Dunkels,Bjorn Gronvall ,and Thiemo Voigt ,Contiki –a Lightweight and Flexible
Operating System for Tiny Networked Sensors”inEmNeTSI,Tampa USA 2005
http://www.wsnmagazine.com/tinyos-tutorial/
http://fiji.eecs.harvard.edu/Volcano
https://www.sics.se/search/content/contiki
http://www.tinyos.net
Editor's Notes
Wireless Sensor Networks :IntroductionApplications of WSNsMain Issues in WSN Middleware and Programming Abstractions for WSNGoals for Middleware and Programming abstractionsOSM : Object State model (FSM)Contiki: Event driven abstractionsTinyOs: Event driven abstractionsConclusionReferences
Check on the range and specs of components
Local variables allocated on runtime stack, they r exited after an execution as such cn not be used for nested events Global variables lock up the lil available memory
Command-line interaction with a network of Contiki nodes● Unix-style pipelines● File system interaction: ls, write, read● Repetition: repeat, randwait● Network commands: netcmd● sense | senseconv● sense | write file | send● repeat 0 20 { randwait 20 { sense | blink | se
Rime: low-power, routing● Can run on top of each other