HongJie Wan
Information Engineering Department, Beijing University of Chemical Technology, China
Xuegang Chen
Department of Computer science, Xiangnan University, China
HaoJiang Deng
National Network New Media Engineering Research Center, Institute of Acoustics, Chinese Academy of Sciences, China
Xiang Cui
Information Engineering Department, Beijing University of Chemical Technology, China
ABSTRACT
Due to long time continuous operation and the high sampling rate, the distributed optical fiber pre-warning system will generate vast amounts of data. Compressed sensing can sample and compress the signal at the same time thus reduce the data amount to be stored or transferred. Therefore, this article uses compressed sensing approach to compress the optical fiber pre-warning data. In compressed sampling phase, the signal is classified using sparse detection method, then more measurements will be taken for the segment containing threatening event than the segment of normal operation, through this the amount of data can be further reduced. In the signal reconstruction phase, the signal compressed sampling process is modeled by the relevant vector machine and the signal recovery is implemented by probabilistic parameter estimation methods. Using Bayesian framework, the sparsity of the signal and the noise each is modeled by a prior. The sparsity and the noise can be estimated in parameter estimation process and the sparsity needn't be given in advance. The optical fiber pre-warning system is long-running system and the sparsity of the signal will change with time, so the automatic sparsity determination ability is superior to other existing recovery methods. Experimental results show that, under the same measurement, the proposed method can reconstruct the signal with high quality and the reconstructed signal will not affect the positioning result.
PDF References Citation
How to cite this article
HongJie Wan, Xuegang Chen, HaoJiang Deng and Xiang Cui, 2013. RVM Based Compressed Sensing for Optical Fiber Pipeline Data Compression. Journal of Applied Sciences, 13: 1877-1882.
DOI: 10.3923/jas.2013.1877.1882
URL: https://scialert.net/abstract/?doi=jas.2013.1877.1882
DOI: 10.3923/jas.2013.1877.1882
URL: https://scialert.net/abstract/?doi=jas.2013.1877.1882
REFERENCES
- Tsaig, Y. and D.L. Donoho, 2006. Extensions of compressed sensing. Signal Process., 86: 549-571.
CrossRefDirect Link - Needell, D. and J.A.Tropp, 2009. CoSaMP: Iterative signal recovery from incomplete and inaccurate samples. Analysis, 26: 301-321.
Direct Link - Candes, E.J. and T. Tao, 2007. The Dantzig selector: Statistical estimation when p is much larger than n. Ann. Statist., 35: 2313-2351.
Direct Link - Tropp, J.A. and A.C. Gilbert, 2007. Signal recovery from random measurements via orthogonal matching pursuit. IEEE Trans. Inform. Theory, 53: 4655-4666.
CrossRef - Tipping, M.E., 2001. Sparse Bayesian learning and the relevance vector machine. J. Machine Learn. Res., 1: 211-244.
Direct Link - Babacan, S.D. R. Molina and A.K. Katsaggelos, 2010. Bayesian compressive sensing using laplace priors. IEEE Trans. Image Process., 19: 53-63.
CrossRef - Mallat, S.G. and Z. Zhang, 1993. Matching pursuits with time-frequency dictionaries. IEEE Trans. Signal Process., 41: 3397-3415.
CrossRef - Ji, S., Y. Xue and L. Carin, 2008. Bayesian compressive sensing. IEEE Trans. Signal Process., 56: 2346-2356.
CrossRef - Chen, S.S., D.L. Donoho and M.A. Saunders, 1998. Atomic decomposition by basis pursuit. SIAM J. Sci. Comput., 20: 33-61.
CrossRefDirect Link - Dai, W. and O. Milenkovic, 2009. Subspace pursuit for compressive sensing signal reconstruction. IEEE Trans. Inform. Theory, 55: 2230-2249.
CrossRef