Building a Microphone Array for Isolating Multiple Sound Sources

Abstract

With devices such as Alexa or ”hands-free” calling, voice recognition systems are becoming more integrated into my daily lives. As time goes on, the industry for voice recognition systems continues to expand. Even with strong growth many devices are subject of noise which can be a powerful source of interference. The ability for these systems to isolate sound sources from external noise is essential in order to improve the effectiveness of voice recognition systems. The goal of this project is to develop a device that could be used for beam-forming. The idea is to build a microphone array with only commonplace parts. Previous projects have done something similar with hybrid systems made from TFTs and CMOS technology, but this device will only require the use of CMOS components. The product will be composed of an array of 12 microphone channels each equipped with their own readout circuit that will output into an MSP430. The MSP430 will digitize all of the channels and send it to an external hard drive via a bluetooth transmitter. Once the data is collected, it could be used for such things as beam-forming where the data can be used to determine where sound is coming from.