Abstract
In this paper, a recently-proposed adaptive FIR filtering algorithm basedupon an anti-Hebbian learning scheme is analyzed. Approximate evolution equationsfor the mean coefficient vector and coefficient correlation matrix are derived assumingnoise-corrupted input and desired response signals. From our analysis, it is shown thatthe proposed algorithm can achieve an unbiased estimate of the unknown system in asystem identification task if the input noise is uncorrelated with the same variance ...

This publication has 11 references indexed in Scilit: