Time-division multiple reference approach for multiple-channel active noise control system

Chung Ying Ho, Lichuan Liu, Kuo Kai Shyu, Sen M. Kuo, Cheng Yuan Chang

Research output: Contribution to journalArticlepeer-review

9 Scopus citations

Abstract

The feedforward active noise control (ANC) requires the reference sensor to obtain the knowledge of the noise source in order to generate the corresponding anti-noise to cancel it. In some practical applications, the noise source is unknown or moving. The placement of the reference sensor is critical to achieve a reasonable performance. Conventionally, the multiple-channel ANC system using the multiple-reference/multiple-output filtered-x least mean square (MRMO-FXLMS) algorithm is the basic solution, however, the computational complexity is high. Therefore, this paper proposed a time-division multiple reference (TDMR) approach for the multiple-channel ANC system, which divides the input channel into different time slots and assigns a time slot to each reference sensor. The proposed system takes only one reference signal as an input at each iteration for the ANC system, it can reduce the computational complexity, and achieve acceptable noise attenuation performance at the same time. Without the demand of prior knowledge of the noise direction, the proposed method can attenuate the noise from anywhere and reduce enormous computational complexity. The real-time experiments of the multi-channel ANC implemented on an infant incubator show the acceptable noise reduction performance of the proposed method compared with other existing approaches.

Original languageEnglish
Article number115922
JournalJournal of Sound and Vibration
Volume495
DOIs
StatePublished - 17 Mar 2021

Keywords

  • Active noise control (ANC)
  • Filtered-x least mean square (FXLMS) algorithm
  • Multiple-channel
  • Time-division multiple reference (TDMR)

Fingerprint

Dive into the research topics of 'Time-division multiple reference approach for multiple-channel active noise control system'. Together they form a unique fingerprint.

Cite this