Active noise control using step size auto adjust method

The purpose of this study is to apply Active noise control (ANC) to 3-dimensional space. Direct Adaptive Algorithm (DAA) proposed by Sano et al. is an effective algorithm because preidentification of the secondary path, the path from a control source to an error microphone, property is unnecessary for this algorithm. However DAA has a problem that the control effect is affected by input noise level. Against this problem, we propose DAA with Step Size Auto Adjusting Method (DAA-SAM). This method normalizes signals into adaptive filters by reference signal. The normalized signals are used in updating adaptive filters. The validity of DAA-SAM against different input levels of white noise or unsteady noise is shown through computer simulations and experiments.