Active noise barrier efficiency improvement using multirate signal processing

Abstract

In this paper, implementation of multirate signal processing techniques in an active noise barrier system was considered. Such a modification of a conventional multiple-error filtered-x LMS algorithm is supposed to give improved overall system performance, by separately controlling two different frequency ranges. The control system consists of one detection microphone, three secondary sources, and three error sensors. The primary signal was a random noise signal with a spectrum that coincides a typical road traffic noise spectrum. Locations of components were the ones found to be optimal in our previous work. Computer simulation was developed to compare the attenuation achieved with a control system that uses a multiple-error filtered-x LMS algorithm and the one involving multirate signal processing. Results of the computer simulation confirmed that attenuation is increased using multirate signal processing.