Multichannel LMS-Based Large Area Sound Inhibition in Home Environment: Preliminary Simulation and Experiments

In this paper, we report the methodology and preliminary results of a set of research work towards large area sound inhibition in home environment. The major algorithmic structure is multichannel LMS (Least Mean Square) adaptive filtering, with filtered-X and IIR (Infinite Impulse Response) as basic refinements. The simulation part of work consists both of the acoustic environment and LMS signal processing. The former deals with sound reflection paths against all inner surfaces of a room between all source-receiver pairs, whereas the latter mainly verifies the specific LMS filter settings with given numbers and locations of the sound source, the inhibition loudspeakers, and the microphones. The experimental part of work mainly consists of setting of sound sources and receivers, and a DSP-based (Digital Signal Processor) real-time processing unit running the DSP algorithms. The preliminary yet promising results in both simulation and real-time experiments provide the future direction of improvement in this research work.