1. Data funneling : routing with aggregation and compression for wireless sensor networks Petrovic, D.; Shah, R.C.; Ramchandran, K.; Rabaey, J. ; SNPA 2003

2. Outline Introduction Data funneling Simulation result Coding by ordering Conclusion

3. Introduction There is a multiplicity of scenarios in sensor networks –Environmental control in office building –Monitoring of seismic activity –Smart home providing security –Interactive museum

4. Introduction Energy consumption determines the life time of a sensor network Communication wirelessly consumes more power at the nodes than other activity We want to minimize the amount of communication required by the sensor nodes

5. Introduction Two methods are discussed to improve the lifetime –Packet aggregation technique –Data compression

6. Data funneling The network environment –Sensors Numerous Sense physical phenomena Generate readings –Controllers Fewer in number Observe the readings from multiple sensors

7. Data funneling Sensors may –Report to the controller at approximately the same time –Have similar headers Savings may be realized by combining different packets into one large packet with a single header

8. Data funneling It reduces the overhead of packet headers Decreases the probability of packet collision –It allows the same amount of information to be transmitted by fewer nodes

9. Data funneling

13. Data funneling creates clusters within the sensor network –The clusters it creates have a dynamic hierarchy –There is not a single cluster head Border nodes take turns acting as cluster head Spreading out the responsibility and the load

14. Simulation result OpNet network simulator Each sensor sends it reading to the controller every 10 seconds If the average number of sensor readings per packet is 7 –The energy expected on packet header is reduced by 6/7=86%

15. Simulation result α is the ratio of bits in a packet header to the total number of bits in a packet m is the average number of sensor readings per transmitted packet Total energy reduced by –α*((m-1)/m)*100%

16. Simulation result

17. Coding by ordering The border node receives the packets from n sensors and make up a super-packet Super-packet –Contain each node ’ s ID Payload

18. Coding by ordering The border node has the freedom to choose the ordering of the packets within the super-packet The border node is allowed to choose to suppress some of the packets –Not to include them in the super-packet

19. Coding by ordering For example –Four node with ID 1,2,3,and 4 –Each generates an independent reading which is a value from the set {0, …,5} –The border node can choose To suppress the packet from node 4 An appropriate ordering among the 3!=6 –Possible orderings of the packets from nodes 1,2,3 indicate the value generated by node 4

20. Coding by ordering

21. n : the number of packets present at the encoder k : the range of possible values generated by each sensor(2 k ) d : the range of node ID ’ s of the sensor nodes l : the largest number of packet that can be suppressed

22. Coding by ordering- achievable with simple codec To alleviate this problem, Stiring ’ s approximation is used to convert (1)

23. Conclusion This work proposes a routing algorithm-Data Funneling It can reduce the amount of energy spent on communication It also reduces the probability of packet collision