Interval Arithmetic based Adaptive Filtering Technique for Removal of Noise in Audio Signal
Akshay V. Nagashetti1, Soumya S. Patil2, Rajashekar B. Shettar3
1Akshay Nagashetty, Student, School of Electronics and Communication Engineering, KLE Technological university Hubballi, Karnataka, India.
2Soumya S Patil, Assistant Professor, School of Electronics and Communication Engineering, KLE Technological university Hubballi, Karnataka, India.
3Rajashekar B Shettar, Professor, School of Electronics and Communication Engineering, KLE Technological university Hubballi, Karnataka ,India.
Manuscript received on April 30, 2020. | Revised Manuscript received on May 06, 2020. | Manuscript published on May 30, 2020. | PP: 1900-1905 | Volume-9 Issue-1, May 2020. | Retrieval Number: E6311018520/2020©BEIESP | DOI: 10.35940/ijrte.E6311.059120
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: Active noise cancellation is one of the fundamental problems in acoustic signal processing. The proposed work focuses on the enhancement of audio signal quality by cancelling the noise using interval analysis (arithmetic). An adaptive filters basically works on the concept of optimal weight calculations which is an optimization problem. This optimization problem can be more effectively solved using interval analysis. Interval analysis gives the boundary of the weight co officiants. Using interval Newton method, the weight co officiants are found. This algorithm is tested for noise cancellation of speech signal. The three adaptive filters algorithm used for comparison with the obtained results are Least Mean Square (LMS), Recursive Mean Square (RMS) filters and Kernel based filters. It is observed that the parameters mean square error is very less. The speed of convergence and signal to noise ratio is improved as compared to kernel methods. But processing time is very high and computational cost is doubled, as interval data includes infimum and supremum values. This algorithm can be used in noise cancelling headphones.
Keywords: Adaptive noise cancellation (ANC), LMS algo- rithm, Kernel adaptive filter, RLS algorithm, Interval analysis.
Scope of the Article: Algorithm Engineering