Hybrid Data Fusion Method Using Bayesian Estimation and Fuzzy Cluster Analysis for WSN
Data fusion is the process of combining data from multiple sensors in order to minimize the amount of data and get an accurate estimation of the true value. The uncertainties in data fusion are mainly caused by two aspects, device noise and spurious measu
- PDF / 1,795,951 Bytes
- 8 Pages / 439.37 x 666.142 pts Page_size
- 34 Downloads / 186 Views
Abstract Data fusion is the process of combining data from multiple sensors in order to minimize the amount of data and get an accurate estimation of the true value. The uncertainties in data fusion are mainly caused by two aspects, device noise and spurious measurement. This paper proposes a new fusion method considering these two aspects. This method consists of two steps. First, using fuzzy cluster analysis, the spurious data can be detected and separated from fusion automatically. Second, using Bayesian estimation, the fusion result is got. The superiorities of this method are the accuracy of the fusion result and the adaptability for occasions. Keywords Data fusion data
Fuzzy cluster analysis Bayesian estimation Spurious
H. Fu Y. Liu (&) Z. Zhang S. Dai School of Electronic and Information Engineering, Beijing Jiaotong University, Beijing 100044, China e-mail: [email protected] H. Fu e-mail: [email protected] Z. Zhang e-mail: [email protected] S. Dai e-mail: [email protected] H. Fu Y. Liu Z. Zhang Key Laboratory of Communication and Information Systems, Beijing Municipal Commission of Education, Beijing Jiaotong University, Beijing 100044, China
Y.-M. Huang et al. (eds.), Advanced Technologies, Embedded and Multimedia for Human-centric Computing, Lecture Notes in Electrical Engineering 260, DOI: 10.1007/978-94-007-7262-5_91, Springer Science+Business Media Dordrecht 2014
809
810
H. Fu et al.
Introduction Sensor networks usually have a large number of sensor nodes to observe the interest parameter in the environment. This often results in data redundancy and repetition. As a consequence of that, the energy of the network is wasted. It is important to minimize the amount of data transmission so that the lifetime can be extended and the bandwidth can be saved. Data fusion is the process of combining data from multiple sensors in order to minimize the amount of data [1]. Many algorithms about data fusion have been proposed in literatures and books include arithmetic mean, Kalman filter, Bayesian estimation, Entropy theory, DempsterShafer theory, fuzzy logic and neural network [2–8]. As the energy of the network is limited, the data fusion algorithm for WSN is required to be easy. The objective of multisensor data fusion is to obtain an accurate, consistent and meaningful information. This information cannot be achieved by any single sensor in the network because of the uncertainties in the network. These uncertainties are mainly caused by two aspects, device noise and spurious measurement. Device noise here includes device inaccuracy and the noise in the environment. Spurious measurement is caused by sensor failure or even security attack. If any of the aspects is not considered during data fusion, it might lead to an inaccurate or erroneous result. Hence, a good data fusion algorithm should consider both of the aspects. Considering the device noise and the spurious measurement, this paper proposes a new method for data fusion. The two aspects have different performance on the obs
Data Loading...
