The Comparison of the Performance for LMS Algorithm Family Using Asymptotic Relative Efficiency

점근상대효율을 이용한 최소평균제곱 계열 적응여파기의 성능 비교

  • Sohn, Won (School of Electronics and Information, Kyung Hee University)
  • 손원 (경희대학교 전자정보학부)
  • Published : 2000.11.25

Abstract

This paper examines the performance of adaptive filtering algorithms in relation to the asymptotic relative efficiency (ARE) of estimators. The adaptive filtering algorithms are Hybrid II and modified zero forcing (MZF) algorithms. The Hybrid II and MZF algorithms are simplified forms of the LMS algorithm, which use the polarity of the input signal, and polarities of the error and input signals, respectively. The ARE of estimators for each algorithm is analyzed under the condition of the same convergence speed. Computer simulations for adaptive equalization are performed to check the validity of the theory. The explicit expressions for the ARE values of the Hybrid II and MZF algorithms are derived, and its results have similar values to the results of computer simulation. It also revealed that the ARE values depend on the correlation coefficients between input signal and error signal.

이 논문은 최소평균제곱계열 적응여파기의 성능을 동일한 수렴속도를 가지는 조건에서 최소평균제곱 알고리즘에 대한 상대적인 성능을 점근상대효율을 이용하여 분석하였다. 분석된 최소평균제곱 계열 알고리즘은 Hybrid II 및 MZF(Modified Zero Forcing) 알고리즘이다. 이들은 최소평균제곱 알고리즘을 단순화한 형태로서 각각 입력신호의 부호정보, 오차신호와 입력신호의 부호정보를 사용한다. 각 알고리즘에 대한 추정기의 점근상대효율은 동일수렴속도 조건에서 분석되었으며, 적응등화기에 대한 모의실험이 분석결과를 확인하기 위하여 수행되었다. 각 알고리즘에 대하여 유도된 점근상대효율에 대한 명시적 표현은 모의실험결과와 유사한 결과를 가졌으며, 점근상대효율은 입력신호와 오차신호간의 상관계수 값에만 좌우된다는 것이 밝혀졌다.

Keywords

References

  1. E. L. Lehmann and G. Casella, Theory of Point Estimation. New York: Springer Verlag, 1998
  2. J. D. Gibson and J. L. Melsa, Introduction to Nonparametric Detection with Applications. New York: Academic Press. 1975, and IEEE Press Classic Reissue, 1995
  3. D. L. Duttweiler, 'Adaptive filter performance with nonlinearities in the correlation multiplier,' IEEE Trans. Acoust., Speech, Signal Processing, vol. ASSP-30, pp.578-586, August 1982
  4. E. Masry and F. Bullo, 'Convergence analysis of the sign algorithm for adaptive filtering.' IEEE Trans. Inform. Theory, vol. IT-41, pp, 489-495, March 1995 https://doi.org/10.1109/18.370150
  5. E. Eweda, 'Transient performance degradation of the LMS, RLS, sign, signed regressor, and sign sign algorithms with data correlation,' IEEE Trans. Circuits Syst., vol. CAS-46, pp. 1055-1063, August 1999 https://doi.org/10.1109/82.782049
  6. S. Koike, 'Analysis of adaptive filters using normalized signed regressor LMS algorithm,' IEEE Trans. Inform. Theory, vol. IT-47, pp. 2710-2723, October 1999 https://doi.org/10.1109/78.790653
  7. 손원, '동일 수렴속도 조건에서 점근상대효율을 이용한 LMS와 Hybrid 1 알고리즘의 성능분석,' 한국통신학회논문지, 제25권 제 2B호, 385-390쪽, 2000년 2월
  8. S. Haykin, Adaptive Filter Theory. New York: Prentice Hall. 1995
  9. B. Widrow and S. D. Stearns, Adaptive Signal Processing. New York: Prentice Hall, 1985