1. 1 A Real-Time Communication Framework for Wireless Sensor-Actuator Networks Edith C.H. Ngai 1, Michael R. Lyu 1, and Jiangchuan Liu 2 1 Department of Computer Science and Engineering, The Chinese University of Hong Kong 2 School of Computer Science, Simon Fraser University IEEE Aerospace Conference, Mar 4-11, 2006, Big Sky, MT

2. 2 Outline Introduction Related Work Real-time Communication Framework Event Detection and Report Actuation Coordination and Reaction Conclusion

3. 3 Introduction Wireless sensor network (WSN) formed by a group of sensors monitors the environments passive, without interactions Wireless sensor-actuator network (WSAN) includes both actuators and sensors becomes an extension to WSN

4. 4 WSAN Actuators resource-rich devices equipped with more energy, stronger computation power, longer transmission range, and usually mobile make decisions and perform appropriate actions in response to the sensor measurements Sensors small and low-cost devices with limited energy, sensing, computation, and transmission capability passive devices for collecting data only and not interactive to the environments

5. 5 WSAN (3) Event Reporting (1) Event Happen (2) Data Aggregation (4) Response to the Event

6. 6 WSAN Sensors and actuators collaborate sensors perform sensing and report the sensed data to the actuators actuators then carry out appropriate actions in response Applications environmental monitoring sensing and maintenance in large industrial plants military surveillance, medical sensing, attack detection, and target tracking, etc.

7. 7 Our Focus Real-time communications and reactions e.g. fast reaction in a fire Self-organized and distributed Event-driven applications e.g. fire, leakage of gas, attack Provide fast and effective response to an event

8. 8 Related Work Real-time communications in WSN SPEED real-time unicast, real-time area-multicast and real- time area-anycast for WSN achieved by using a combination of feedback control and non-deterministic QoS-aware geographic forwarding with a bounded hop count

9. 9 Related Work Real-time communications in WSN MMSPEED Multi-Path and Multi-Speed Routing Protocol for probabilistic QoS guarantee in WSN multiple QoS levels are provided in the timeliness domain by guaranteeing multiple packet delivery speed options supported by probabilistic multipath forwarding in the reliability domain

10. 10 Related Work Distributed coordination framework for WSAN based on an event-driven clustering paradigm all sensors in the event area forward their readings to the appropriate actors by the data aggregation trees provides actuator-actuator coordination to split the event area among different actuators assumes immobile actuators that can act on a limited area defined by their action range

11. 11 Our Work A real-time communication framework for WSN Event-reporting algorithm divides the event area into pieces of maps data fusion layered data representation Actuator coordination algorithm supports mobile actuators under sparse deployment

12. 12 A Real-time Communication Framework for WSN Event reporting Detection of an event Formation of map and data aggregation Data transmission Actuator coordination Combination of maps Location update

13. 13 Event Detection and Report

14. 14 Formation of Maps To reduce the network traffic, the sensor will aggregate event reports and perform data fusion from the neighboring nodes The sensors r, which detected an event the earliest, start the formation of maps

15. 15 Formation of Maps Algorithm 1 Formation of Maps for nodes r detected an event if (data aggregation not yet started) Broadcast DetectEvt (r, 0, e) msg. to n r end if end for for nodes v receive DetectEvt msg. from v if (h<max_hop && (v.event && ! v.reported)) forward DetectEvt(v, h+1, e) msg. to n v else reply ReplyEvt (meets boundary) msg. to v end if end for for nodes v receive ReplyEvt msg. if (msg.==meets boundary) reply ReplyEvt(x v, y v, data v, e) msg. to parent else concat own data and reply ReplyEvt msg. to parent end if end for

16. 16 Data Aggregation When a node receives the replies from its descendent nodes, it concatenates its own reply and forwards them to the previous hop Nodes with even number of depth h concatenate the reply with its own coordinates and sensed data Nodes with odd number of depth h aggregate the data from their immediate descendents before forwarding them.

17. 17 Data Aggregation Algorithm 2 Data Aggregation for nodes receive ReplyEvt msg. if (h==odd) //node in odd no. of depth gather all data from its descendents v hj in h+1 remove datav hj from ReplyEvt msg. concat {meanv h, xv h, yv h, e} to ReplyEvt msg. forward ReplyEvt msg. to parent in depth h-1 else concat {xv h, yv h, datav h, e} to ReplyEvt msg. forward ReplyEvt msg. to parent in depth h-1 end if end for aggregate

18. 18 Layered Data Transmission Data are divided into the base layer and the refinement layer The base layer contains the type of event the time when the event is first detected the location of the map mean value of the collected data The refinement layer contains all the means calculated by nodes with odd number of depth and their corresponding locations

19. 19 Base Layer and Refinement Layer Base Layer Refinement Layer …… Transmission Sequence of Refinement Layer : data from the node located at C : data from nodes with distance d max from C : data from nodes with distance d max/2 from C : … …… mean 0 mean dmax mean dmax/2 mean dmax/4 mean dmax*3/4

20. 20 Actuator Coordination and Reaction

21. 21 Combination of Maps After an actuator receives the data in the base layer from the sensor r, it gets one piece of map in the event area It then combines multiple maps if it receives more than one report on same type of event happening in the same area within time period t e It starts communicating with other actuators located closely to the event area as well Actuators exchange information for combining their maps and approximating the size of the event

22. 22 Combination of Maps Algorithm 3 Combination of Maps for each actuator a on event e, if (received multiple S r ) Gather the B r in grid coordinates from all S r Remove the redundant B r Remove the connected B r Store the remaining B r in B a end if Exchange the B a with other actuators Remove the redundant B a Remove the connected B a Estimate the B a by finding lower-left and upper- right grids and end for Sr Ba

23. 23 Location Update Update the location of actuator to sensors Plan the optimal location of the actuators for efficient reactions

24. 24 Future Work Complete detailed operations Enhance the efficiency and reliability of the current approach Provide performance analysis with mathematical models Evaluate the solution with simulations

25. 25 Conclusion A real-time communication for WSAN is presented It provides an efficient event-reporting algorithm that reduces network traffic It considers layered data transmission to minimize the delay It provides an actuator coordination algorithm with combination of maps for effective reaction It offers a distributed, self-organized, and comprehensive solution for real-time event reporting and reaction for WSAN

26. 26 Q & A