Wideband sound reproduction in a 2D multi-zone system using a combined two-stage Lasso-LS algorithm

This paper addresses the provision of personal soundfields (zones) to multiple listeners in a space for which there are several fixed, virtual sources. For such multizone systems, optimization of speaker positions and weightings is important to reduce the number of active speakers. In this paper a two stage pressure matching optimization is proposed for wideband sound sources (such as speech signals). In the first stage, the least-absolute shrinkage and selection operator (Lasso) is used to select the speakers' positions for all sources and frequencies. A second stage then optimizes reproduction using all selected speakers on the basis of regularized least-squares (LS) or Lasso algorithm. The performance of these new, two-stage approaches is investigated for different speaker numbers, frequency range and reproduction angles. Results show that a limited arc of speakers using a two-stage optimization can give up to 38dB improvement in zone normalized squared error over a single-stage LS optimization.