Topic Description
Active noise control (ANC) is nowadays a standard feature in headsets and is used to minimize external noise. Physically, ANC is based on the principle of destructive interference of acoustic waves. ANC algorithms are implemented digitally using powerful DSPs. Methods of adaptive filters, parameter estimation and robust feedback control are used for this purpose.

Task Definition
The basic filter principles of feedforward or feedback control use digital signal processing and filter design methods ,and also machine learning. In feedforward systems, the aim is to generate the anti-noise signal with minimum latency. To do this, the acoustic path must be estimated to compensate for its influence when generating the signal, and the ANC filter can also be adapted to changes in the acoustic conditions. With feedback ANC, a filter is optimized in advance. A-priori knowledge of the path uncertainty plays an important role here, as the feedback loop must remain stable in any case.

Prerequisites

• Good programming skills in Python / Matlab
• Good knowledge of acoustics & acoustic measurement technology
• Interest in applied systems theory