Further analysis and applications of (asymmetric) crosstalk-resistant adaptive noise cancellation
Adaptive Noise Cancellation (ANC) with crosstalk has been a common problem. There have been various approaches to combat it. There are many adaptive noise cancellers with feedback structures in order to enhance the noise suppression of the desired output signal. The canceller is called Crosstalk Resistant Active Noise Canceller (CTRANC). There is another derivative structure called Asymmetric CTRANC (ACTRANC). However, there has been something lacking in the completed analysis of the system because the (A)CTRANC involves feedback filters. In this thesis, some of the previous presentations are revisited and a significant mistake is clarified. Most importantly, the analysis of a two-channel CTRANC leads to an optimal two-channelcrosstalk- canceller. The optimal two-channel-crosstalk-canceller is analyzed and further simplified as Dr. Widrow and et. al. have done in the LMS algorithm. This thesis showed results of simulations of (A)CTRANC in various adaptive noise cancellation, speech enhancement, and the (A)CTRANC. The theory limits of the (A)CTRANC were investigated. The research demonstrated that the (A)CTRANC can be used to achieve better performance under different conditions.