Adaptive sound zone control using third-order tensor decomposition-based acoustic contrast control-pressure matching

Sound zone control utilizes a loudspeaker array to reproduce different audio contents in multiple regions of a shared space. It is preferred that the reproduction of audio sources can adapt to sound zone system changes. However, the existing adaptive sound zone control methods do not address their tracking/convergence capability. In this paper, we propose acoustic contrast control-pressure matching (ACCPM) based on the third-order tensor decomposition technique (ACCPM-TOT) to improve the tracking/convergence capability of adaptive sound zone control under dynamic conditions. The proposed method decomposes the ACCPM control filter into three subfilters, then updates them using separate recursive least squares (RLS) algorithms, and finally combines their coefficients. The behavior of ACCPM-TOT is evaluated using the generated and measured room impulse responses. The results show that under stable conditions, ACCPM-TOT achieves comparable performance to the nonadaptive ACCPM and the adaptive version of ACCPM with a single RLS, and in dynamic environments, ACCPM-TOT stands out with recontrol capability, faster tracking/convergence, and lower complexity due to third-order tensor decomposition.