The paper develops a unified approach to the transient analysis of adaptive filters with error nonlinearities. In addition to deriving earlier results in a unified manner, the approach also leads to new performance results without restricting the regression data to being Gaussian or white. The framework is based on energy-conservation arguments and avoids the need for explicit recursions for the covariance matrix of the weight-error vector.
Florent Gérard Krzakala, Lenka Zdeborová, Hugo Chao Cui
Victor Panaretos, Laya Ghodrati