적응 문턱치 알고리즘을 이용한 충격잡음 제거

Impulse Noise Cancellation Using Adaptive Threshold Algorithm

  • 이진 (삼척대학교 제어계측공학과) ;
  • 박종환 (서울시립대학교 전자.전기공학부) ;
  • 김세동 (두원공과대학 전기과) ;
  • 이영석 (청운대학교 전자공학과) ;
  • 김성환 (서울시립대학교 전자.전기공학부)
  • 발행 : 2000.11.01

초록

This paper presents a new adaptive impulse noise cancelling technique based on the adaptive nonlinear suppressing function. The proposed "adaptive threshold algorithm (ATA)" is controlled by the normalized power prior input data term, and this adaptive threshold makes the cancelling system highly robust against additive impulse noise. For the performance evaluation, we have tested the proposed algorithm with the observed signals simulated in various impulsive noise environments and real EMG signals. As a result the proposed algorithm shows superior performance of 51.7% to the available techniques in the points of SNR and MSE.

키워드

참고문헌

  1. IEEE Trans. Signal Proc. v.43 no.12 Parameter Estimation and Blind Channel Identification in Impulsive Signal Environments X. Ma;C. L. Nikias
  2. IEEE Trans. Sig. Proc. v.42 no.6 Detection in Impulsive Noise Based on Robust Whitening A. J. Efron;H. Jeen
  3. 한국 음향 학회지 v.16 no.5 임펄스 잡음 환경 하에서 FNOM과 MD를 이용한 새로운 시지연 및 시간-주파수 지연 복합 추정 방법 이진;정정균;이영석;김성환
  4. 대한전자공학회 논문지 v.36 no.3 임펄스 잡음 환경 하에서 MO-FIR 필터링을 이용한 적응 시지연 추정 이진;전광석;여송필;김성환
  5. IEEE Trans. Signal Processing v.43 no.8 Adaptive Robust Impulse Noise Filtering S. R. Kim;A. Efron
  6. IEEE Trans. Signal Proc. v.45 no.12 The Stability of LMS V. Solo
  7. IEEE Trans. Auto. Control v.42 no.6 Necessary and Sufficient Conditions for Stability of LMS L. Guo;L. Lennart;G. J. Wang
  8. Applied Optics v.35 no.20 Fractional Wiener filter Z. Zalevsky;D. Mendlovic
  9. J. Acoust. Soc. Am. v.101 no.2 Error analysis in acoustie elastography. Ⅱ. Strain estimation and SNR analysis M. Bilgen;M. F. Insana
  10. IEEE Trans. on Biomedical Eng. v.44 no.9 MSE Behavior of Biomedical Event-Related Filters A. K. Barros;N. Ohnishi
  11. Journal Am. Statis. Assoc. v.93 no.444 Time-Dependent Spectral Analysis of Nonstationary Time Series S. Adak
  12. Med. Biol. Eng. Comput. v.21 Online muscle fiber action potential conduction velocity measurements using the surface e.m.g cross-correlation technique H. Zorn;M. Naeije