Research on Node Localization Algorithm in WSN Basing Machine Learning

Abstract:

Machine learning uses experience to improve its performance. Using Machine Learing, to locate the nodes in wireless sensor network. The basic idea is that: the network area is divided into several equal portions of small grids, each gird represents a certain class of Machine Learning algorithm. After Machine Learning algorithm has learnt the parameters using the known beacon nodes, it can classify the unknown nodes location classes, and further determine their coordinates. For the SVM OneAgainstOne Location Algorithm, the results of simulation show that it has a high localization accuracy and a better tolerance for the ranging error, while it doesnt require a high beacon node ratio. For the SVM Decision Tree Location Algorithm, the results show that this algorithm is not affected seriously by coverage holes, it is suitable for the network environment of nonuniformity distribution or existing coverage holes.

• Periodical: Advanced Materials Research (Volumes 756-759)
• Pages: 3568-3573
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.756-759.3568
• Online since: September 2013
• Authors: Qing Zhang Chen, Yu Zheng Chen, Cong Ling Fan, Fan Yang, Peng Wang, Yan Jing Lei
• Keywords: Coverage Hole, Node Localization, Region Classification, Support Vector Machine (SVM), Wireless Sensor Network (WSN)

References

[1] D. Ma and Er Meng Joo. A Survey of Machine Learning in Wireless Sensor Network. Proceeding of 2007 6th International Conference on Information Communications and Signal Processing . Singapore, 2007, 1-5.

DOI: 10.1109/icics.2007.4449882

[2] H.B. Liu, S.W. Xiong, Q. Chen. Localization in Wireless Sensor Network Based on Multi-class Support Vector Machines. Proceeding of 2009 5th International Conference on Wireless Communication Networking and Mobile Computing. Beijing, 2009, 1(9): 1-4.

DOI: 10.1109/wicom.2009.5303322