Adaptive filtering and prediction based on Hopfield neural networks

Abstract

Adaptive filters have been successfully used in the solutions of several practical problems such as echo and noise cancelers, line enhancers, speech coding, equalizers, etc. Due to that, intensive research have been carried out to develop more efficient adaptive filter structures and adaptation algorithms, almost all of them implemented in a digital way. This is because with the advance of digital technology it is possible to implement more sophisticated and efficient adaptive filter algorithms. However the adaptive digital filters still present several limitations when required to handle frequencies higher than those in the audio range. Recently the interest on adaptive analog filters has grow because they have the ability to handle much higher frequencies, and their size and power requirements are potentially much smaller than their digital counterparts. This paper propose an analog adaptive structure for filtering and prediction whose coefficients are estimated in a continuous time way by using an artificial Hopfield neural network. Simulation results are given to show the desirable features of the proposed structure.