Adaptive Digital Methods for Active Noise Cancellation Headphones

The goal of the proposed PhD thesis is to improve active noise cancellation for headphones by digital means. Firstly, the analogue cancellation path can be digitally analyzed. Depending on the analysis, the analogue cancellation signal will either be amplified to gain a better cancellation or attenuated if the analysis predicts a oscillation and hence a destabilization of the cancellation system. Secondly, the cancellation signal can be digitally manipulated before it is converted back to an analogue signal that is played back by the loudspeaker. This allows for dynamic algorithms that adapt the noise cancellation to every individual situation. However, the digitalization and the following analogue conversion consume time. The time consumption corresponds to a loss of information that deteriorates the cancellation performance. Consequently, algorithms have to be found that predict the development of the noise in order to compensate for the loss of time.