High-speed adaptive noise canceller with parallel block structure using input for stability

A new adaptive algorithm for a high-speed adaptive noise canceller with parallel block structure is proposed. First, an IIR bandpass filter with a variable center angular frequency using adaptive Q-factor control is shown in which the bandpass filter and the adaptive control signal generator are composed of parallel block structures. Secondly, if we use a large step size parameter for the adaptive updating algorithm for fast processing, the large step size parameter may cause a large variation after convergence of the adaptive algorithm. In order to reduce the variation after convergence of the adaptive algorithm, we propose a new adaptive updating algorithm in which the input is used for stability. This method has essentially the same convergence performance as the previous algorithm. However the variation after convergence of the new adaptive updating algorithm is always smaller than that of the previous algorithm. The new algorithm has a small increment in computational complexity. Finally, simulation results are given to demonstrate the convergence performance