In this paper, a performance analysis of FeedBack (FB) Active Noise Cancellation (ANC) systems for automotive applications is presented. Since the performance of FB ANC systems is strongly impaired by the delay induced by the physical distance between microphones and loudspeakers, an experimental setup representative of a car seat headrest has been developed. In this setup, the loudspeaker-microphone distance is on the order of few centimeters. An experimentally acquired band-limited noise source is obtained from an idling car. Noise reduction is performed by using a Filtered-x Least Mean Square (FxLMS) algorithm. Two 2×2 Multiple Input - Multiple Output (MIMO) schemes are considered—standard one and Filtered-Error (FE) based. Our experimental results show that the proposed MIMO systems enable improved performance and are able to control the perceived noise at two listening positions in a realistic automotive application.
Performance Analysis of Feedback MIMO ANC in Experimental Automotive Environment / Opinto, Alessandro; Martalo, Marco; Tripodi, Carlo; Costalunga, Alessandro; Cattani, Luca; Raheli, Riccardo. - ELETTRONICO. - (2020), pp. 1-6. (Intervento presentato al convegno International Conference on Signal Processing and Communication Systems (ICSPCS) tenutosi a Adelaide, Australia nel 14-16 Dec. 2020) [10.1109/ICSPCS50536.2020.9310039].
Performance Analysis of Feedback MIMO ANC in Experimental Automotive Environment
Opinto, Alessandro
;Martalo, Marco;Raheli, Riccardo
2020-01-01
Abstract
In this paper, a performance analysis of FeedBack (FB) Active Noise Cancellation (ANC) systems for automotive applications is presented. Since the performance of FB ANC systems is strongly impaired by the delay induced by the physical distance between microphones and loudspeakers, an experimental setup representative of a car seat headrest has been developed. In this setup, the loudspeaker-microphone distance is on the order of few centimeters. An experimentally acquired band-limited noise source is obtained from an idling car. Noise reduction is performed by using a Filtered-x Least Mean Square (FxLMS) algorithm. Two 2×2 Multiple Input - Multiple Output (MIMO) schemes are considered—standard one and Filtered-Error (FE) based. Our experimental results show that the proposed MIMO systems enable improved performance and are able to control the perceived noise at two listening positions in a realistic automotive application.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
