An adaptive stochastic filter with no positive real condition

A new adaptive stochastic filter is introduced which eliminates the need for strict passivity condition required by existing filters. The proposed filter will also reduce the bias in the estimated parameters by using an appropriate factor to obtain an improved estimate of the unknown white noise. This filter is then applied to the problems of adaptive line enhancement, sinusoidal detection, and spectral estimation to illustrate its usefulness and improvement over existing algorithms.