"This dissertation proposes four new algorithms based on fractionally lower order statistics for adaptive filtering in a non-Gaussian interference environment. One is the affine projection sign algorithm (APSA) based on L₁ norm minimization, which combines the ability of decorrelating colored input and suppressing divergence when an outlier occurs. The second one is the variable-step-size normalized sign algorithm (VSS-NSA), which adjusts its step size automatically by matching the L₁ norm of the a posteriori error to that of noise. The third one adopts the same variable-step-size scheme but extends L₁ minimization to Lp minimization and the variable step-size normalized fractionally lower-order moment (VSS-NFLOM) algorithms are generalized. Instead of variable step size, the variable order is another trial to facilitate adaptive algorithms where no a priori statistics are available, which leads to the variable-order least mean pth norm (VO-LMP) algorithm, as the fourth one. These algorithms are applied to system identification for impulsive interference suppression, echo cancelation, and noise reduction. They are also applied to a phased array radar system with space-time adaptive processing (beamforming) to combat heavy-tailed non-Gaussian clutters. The proposed algorithms are tested by extensive computer simulations. The results demonstrate significant performance improvements in terms of convergence rate, steady-state error, computational simplicity, and robustness against impulsive noise and interference"--Abstract, page iv.
Zheng, Y. Rosa
Stanley, R. Joe
Grant, Steven L.
Moss, Randy Hays, 1953-
Electrical and Computer Engineering
Ph. D. in Electrical Engineering
United States. Air Force. Office of Scientific Research
Missouri University of Science and Technology
Journal article titles appearing in thesis/dissertation
- Affine projection sign algorithm robust against impulsive interference
- Variable step-size normalized sign algorithm for acoustic echo cancellation
- Variable step-size normalized fractionally low-order moment algorithms for non-Gaussian interference environment
- Variable step-size LMP algorithm for heavy-tailed interference suppression in phased array radar
- Fast-converging space-time adaptive processing algorithm for non-Gaussian clutter suppression
xii, 115 pages
© 2011 Tiange Shao, All rights reserved.
Dissertation - Open Access
Adaptive control systems
Echo suppression (Telecommunication)
Space time codes
Print OCLC #
Electronic OCLC #
Link to Catalog Record
Shao, Tiange, "Robust adaptive filtering algorithms for system identification and array signal processing in non-Gaussian environment" (2011). Doctoral Dissertations. 1953.