Dual adaptive noise cancellation method based on Least Mean M-estimate of noise

• Published in: Proceeding of the 11th World Congress on Intelligent Control and Automation pp. 5741 - 5746
• Main Authors: Xueli Wu, Zizhong Tan, Jianhua Zhang, Wei Li
• Format: Conference Proceeding
• Language: English
• Published: IEEE 01.06.2014
• Subjects: Adaptive filters; Correlation; dual adaptive; Filtering algorithms; Finite impulse response filters; impulsive noise; M-estimate; mean square; Noise cancellation; Signal to noise ratio
• DOI: 10.1109/WCICA.2014.7053700

Adaptive noise cancellation technology has been widely applied in all fields. The techniques are ideally suited for reducing spatially varying noise. For some fields noise is generally uncorrelated, in contrast to the useful signal. Adaptive filtering algorithms exploit the correlation properties of signals to enhance the signal-to-noise ratio of the output signal. However, in the case with few prior probability, the effect of conventional adaptive filter noise cancellation is poor. on the other hand its performance would degrade dramatically if there were impulsive noise. In order to improve the signal-to-noise ratio, effectively eliminate noise, In this paper, a two-stage adaptive noise cancellation method is proposed for enhancing ideal signal submerged in noise. The overall method is based on the use of two adaptive filters with primary and reference signals. Use the least mean M-estimate algorithm for noise as objective unction instead of the mean square (MSE) updated the filter weights to achieve the optimal noise reduction effect. The computer simulation results show that the new algorithm has better de-noising effect. And still has a good noise performance when exists impulsive noise. So it has a certain application value.