Bradley Department of Electrical & Computer EngineeringVirginia Tech

Nonlinear / Non-Wiener Effects in Adaptive Algorithms

Synopsis

The focus of this research effort is the investigation of nonlinear effects in normalized least-mean-square adaptive filtering, as manifested by performance better than that of the corresponding Wiener filter and the occurrence of time-varying weight behavior. While the existence of such effects has been known for some time, our recent progress has been to find an explanation for these effects in narrowband adaptive noise canceling and adaptive lagged prediction scenarios.

Publications
Additional Online Material

n/a

Questions & Comments