1. Athanasios Kinalis ∗, Sotiris Nikoletseas ∗, Dimitra Patroumpa ∗, Jose Rolim† ∗ University of Patras and Computer Technology Institute, Patras, Greece †Centre Universitaire d’ Informatique, Geneva, Switzerland IEEE Globecom2009 Biased Sink Mobility with Adaptive Stop Times for Low Latency Data Collection in Sensor Networks

2. Outline Introduction Network model The proposed scheme Simulation Conclusion

3. Introduction As the sensed data are forwarded to the sink node in the WSN Settings have increased implementation complexity Sensor devices consume significant amounts of energy Sensor node Sink node

4. Introduction A approach has been introduced that shifts the burden of acquiring the data, from the sensor nodes to the sink. Help conserve energy since data is transmitted over fewer hops. Many apparent difficulties arise as well since traversing the network in a timely and efficient way is critical. high density of sensors in an area some sensors have recorded a significant amount of data

5. Introduction High delivery delays Even data loss A B

6. Goal They propose biased sink mobility with adaptive stop times, as a method for efficient data collection in wireless sensor networks. reduces latency delivery success rate

7. Network model The sink can accurately calculate its position The sink can aware of the dimensions and boundaries of the network area The sensor of sensing range R D D j j Sensor node Sink node

8. Scheme Network Traversal Deterministic Walk Biased Random Walk Calculation of Stop Time

9. Deterministic Walk j j A Sensor node Sink node

10. Biased Random Walk The probability p(f) v of visiting a neighboring vertex v 2 2 1 1 4 2 2 2 3 3 0 2 3 2 3 1 AB C D E 2 Sensor node Sink node

11. E total is the total initial energy of all the sensors in the network T total_stop represents the maximum total amount of time the sink can remain stationary. n is the number of sensors of the network Calculation of Stop Time ε i the initial energy of each sensor i. T sim is the total time that the experiment is performed E total the maximum amount of energy consumed when sending a message the average event generation rate the energy spent when the sensors remain idle

12. T total_stop_round is the maximum amount of time that the sink will remain static in each round. represents the maximum total amount of time the sink can remain stationary Calculation of Stop Time the algorithm evolves in r rounds

13. Constant stop time. Adaptive stop time. the maximum amount of time that the sink will remain static in each round Calculation of Stop Time the density in cell i

14. Calculation of Stop Time Example AB C D d = 1 d A = 1.5 d B = 1.2 d C = 0.5 d D = 0.8 1.51.2 0.50.8

15. The maximum time the sink can stay in cell i is T adapi. If the sensors empty their memory before T adapi expires, the sink leaves the cell Otherwise, it leaves at the end of T adapi, even if there exist more data to be sent. The sink waits for messages to arrive for a submultiple of T adapi if no messages arrive during that time, it leaves the cell. Calculation of Stop Time

16. Simulation Simulatorns − 2 the size of the network area200m × 200m sensor nodes300 the speed of the mobile sink s ∈ {4, 8, 10, 20}m/s The initial energy reserves of the nodes5.5 Joules The data is generated at an average rate1 message/10 sec

17. Simulation SCD (Stop to Collect Data), one of the algorithms proposed in [6]. In SCD, the mobile sink stops when receiving new data. A. Kansal, A. Somasundara, D. Jea, M. Srivastava, and D. Estrin, “Intelligent fluid infrastructure for embedded networks,” in 2nd ACM/USENIX International Conference on Mobile Systems, Applications, and Services (MobiSys04), 2004.

18. Simulation SCD (Stop to Collect Data), one of the algorithms proposed in [6]. In SCD, the mobile sink stops when receiving new data. A. Kansal, A. Somasundara, D. Jea, M. Srivastava, and D. Estrin, “Intelligent fluid infrastructure for embedded networks,” in 2nd ACM/USENIX International Conference on Mobile Systems, Applications, and Services (MobiSys04), 2004.

19. Conclusion They propose both randomized mobility and deterministic traversals. They adaptive stop times lead to significantly reduced latency and keeping the delivery success rate.

20. Thank you ~