Variable Step-Size and Variable Regularization Adaptive Filtering Algorithms
博士 === 元智大學 === 電機工程學系 === 101 === Numerous variable step-size and variable regularization algorithms have been derived to solve the dilemma of fast convergence rate or low excess mean-square error in the past two decades. This thesis proposes new nonparametric time-domain variable step-size normali...
| Main Authors: | , |
|---|---|
| Other Authors: | |
| Format: | Others |
| Language: | en_US |
| Online Access: | http://ndltd.ncl.edu.tw/handle/42422157543841638980 |
| Summary: | 博士 === 元智大學 === 電機工程學系 === 101 === Numerous variable step-size and variable regularization algorithms have been derived to solve the dilemma of fast convergence rate or low excess mean-square error in the past two decades. This thesis proposes new nonparametric time-domain variable step-size normalized least mean-square (VSS-NLMS) and variable-regularization NLMS (VR-NLMS) algorithms. The VSS-NLMS employs the mean-square error (MSE) and the estimated system noise power to control the step-size update, and the VR-NLMS uses the input signal power, MSE and the estimated system noise power to control the variable regularization parameter. Extending the ideas of our VSS-NLMS algorithm, this thesis also presents a variable step-size control method for the partitioned frequency-domain block LMS (PFBLMS) adaptive filter.
|
|---|