AUTHORS: Ayako Arao, Hiroaki Higaki

Download as PDF

ABSTRACT: In wireless sensor networks, each sensor node observes events and records its attributes including its occurrence time. For precise analysis of the records, local clocks of all the wireless sensor nodes are required to be synchronized; however, due to their individual differences, each local clock has its own drift and offset. Hence, estimation of relative offsets and drifts among the wireless sensor nodes are mandatory. Conventionally, an offset and a drift between neighbor wireless sensor nodes are estimated by exchanges of some control messages. Since transmission delay of these control messages are not predictable due to the mechanism for collision avoidance in wireless LAN protocols, the provided precision of the offset and drift is not acceptable for various sensor network applications. In order to solve this problem, this paper proposes a novel method for estimation of the offset and drift between neighbor wireless sensor nodes based on their event observation. Since events observed both of the neighbor wireless sensor nodes should be recorded with the same occurrence time, the offset and drift are estimated by using differences between the recorded occurrence times. However, it is impossible for the wireless sensor nodes to identify which events are commonly observed. Hence, this paper proposes a novel heuristical method for estimation of commonly observed events between the neighbor wireless sensor nodes by using their sequences of recorded event occurrence times. Here, all possible pairs of an offset and a drift are evaluated by numbers of induced commonly observed events. Results of simulation experiments show that records of event occurrence times expected to include more than three commonly observed events realizes estimation of commonly observed events more precise than 99%.

KEYWORDS: Wireless sensor networks, Local clock synchronization, Records of event occurrence times, Commonly observed events.

REFERENCES:

[1] Cristian, F., “Progabilistic Clock Synchronization,” Distributed Computing, Springer, vol. 3, no. 3, pp. 146–158 (1989).

[2] Jeremy, E., Lewis, G. and Deborah, E., “FineGrained Network Time Synchronization using Reference Broadcasts,” Proceedings of the 5uh Symposium on Operating Systems Design and Implementation, pp. 147–163 (2002).

[3] Kopetz, H. and Ochsenreiter, W., “Clock Synchronization in Distributed Real-Time Systems,” ICE Transactions on Computers, vol. C-36, no. 8, pp. 933–940 (1987).

[4] Miklos, M., Branislav, K. and Kayla, S., “The Flooding Time Synchronization Protcol,” Proceedings of the 2AD International Conference on Embedded Networked Sensor Systems, pp. 39–49 (2004).

[5] Qu, Y., and Georgakopoulos, S.V., “A Distributed Area Coverage Algorithm for Maintenance of Randomly Distributed Sensors with Adjustable Sensing Range,” ICE Global Communications Conference, pp. 286–291 (2013).

[6] Saurabh, G., Ram, K. and Mani, B.S., “TimingSync Protocol for Sensor Networks,” Proceedings of the 1st International Conferenceon Embedded Networked Sensor Systems, pp. 138–149 (2003).

[7] Schick, L., Souza, W.L. and Prado, A,F., “Wireless Body Sensor Network for Monitoring and Evaluating Physical Activity,” Proceedings of the 14th ITNG (2017).

[8] Tanenbaum, A.S. and Steen, M., “Distributed Systems Principles and Paradigms,” Prentice Hall (2002).

[9] Tuba, E., Tuba, M., and Simian, D., “Wireless Sensor Network Coverage Problem Using Modified Fireworks Algorithm,” Wireless Communications and Mobile Computing Conference, pp. 696–701 (2016).

[10] Wu, Y-C., Chaudhari, Q. and Serpedin, E., “Clock Synchronization of Wireless Sensor Networks,” ICE Signal Processing Magazine, vol. 28, no. 1, pp. 124–138 (2011).