Sound Zones as an Optimal Filtering Problem

Jesper Kjær Nielsen, Taewoong Lee, Jesper Rindom Jensen, Mads Græsbøll Christensen

Research output: Contribution to book/anthology/report/conference proceedingArticle in proceedingResearchpeer-review

16 Citations (Scopus)
377 Downloads (Pure)

Abstract

The design of so-called sound zones is a fairly old idea which has recently gained some research attention. The idea is to control a loudspeaker array to reproduce a desired sound field in certain regions. In this paper, we show how the sound zone control problem can be solved using techniques from speech enhancement. Specifically, we first describe in detail the recently introduced variable span linear filtering (VSLF) framework which unifies the popular optimal filtering and subspace-based approaches to speech enhancement. We then show how the sound zone control problem can be solved using the VSLF framework and how a number of well-known sound zone control methods can be viewed as special cases of this solution. We also discuss in detail the differences between the speech enhancement problem and the sound zone control problem and argue that the VSLF framework is actually even better suited for controlling sound zones than for enhancing speech.
Original languageEnglish
Title of host publication2018 52nd Asilomar Conference on Signals, Systems, and Computers
Number of pages5
PublisherIEEE
Publication date2018
ISBN (Print)978-1-5386-9218-9
ISBN (Electronic)978-1-5386-9216-5, 978-1-5386-9217-2, 978-1-5386-9219-6
DOIs
Publication statusPublished - 2018
EventAsilomar Conference on Signals, Systems and Computers (ACSSC 2018) - Pacific Grove, United States
Duration: 28 Sept 201831 Oct 2018

Conference

ConferenceAsilomar Conference on Signals, Systems and Computers (ACSSC 2018)
Country/TerritoryUnited States
CityPacific Grove
Period28/09/201831/10/2018
SeriesAsilomar Conference on Signals, Systems and Computers. Conference Record
ISSN1058-6393

Fingerprint

Dive into the research topics of 'Sound Zones as an Optimal Filtering Problem'. Together they form a unique fingerprint.

Cite this