Transient Signal Detection in Non-Gaussian Noise Using Stochastic Resonance and Wavelet Domain Power-Law Detector

Abstract:

The detection of weak transient signal buried in non-Gaussian noise is investigated. Non-Gaussian noise is modeled by Gaussian mixture distribution. 3-level quantizer is used as a nondynamic stochastic resonance method to enhance SNR of weak signal. NL-length samples of signal are arranged into a matrix. Every column of the matrix is calculated into M-level decomposition. Based on the squared value of the detail and approximation coefficients, a novel Power-Law detector in wavelet domain is established. Numerical experiments and comparison show that, on the same SNR and false alarm rate, proposed method could provide higher detection probability.

• Periodical: Advanced Materials Research (Volumes 945-949)
• Pages: 2043-2047
• DOI: https://doi.org/10.4028/www.scientific.net/AMR.945-949.2043
• Online since: June 2014
• Authors: Hua Yu Dong*, Zhi Yang
• Keywords: DWT, Non-Gaussian Noise, Power-Law Detection, Stochastic Resonance, Transient Detection

References

[1] B. Q. Chen, Z. S. Zhang, Y. Y. Zi, Z. J. He and C. Sun: Mechanical Systems and Signal Processing, Vol. 40 (2013), p.1.

[2] R. G. Waghmare, S. L. Nalbalwar and Arnab Das: Procedia Engineering, Vol. 30 (2012), p.129.

[3] P. Ravier, P. O. Amblard: Signal Processing, Vol. 81(2001), p. (1909).

[4] S. Zozor, P. O. Amblard: Signal Processing, Vol. 82 (2002), p.353.

[5] A. A. Saha, G. V. Anand: Signal Processing, Vol. 83 (2003), p.1193.

[6] M. Sany, K. Bart: Proceedings of the IEEE, Vol. 86(1998), p.2152.

[7] R. Raghu, US AD Report (1995).

[8] A. Nuttall, NUWC-NPT Tech. Rep., Newport, RI, 10751. (1994).

[9] W. Zhen, K. W. Peter: IEEE Transactions on Signal Processing, Vol. 49(2001), p.2454.