Two Sensor Array Beamforming Algorithm

Abstract

Nowadays, noise reduction in mobile communication is an increasingly important topic. This thesis discusses the possibilities of using beamforming with a two microphone array for smartphone applications. The goal of the project is to realise a beamforming algorithm for smartphones to improve on current Signal to Noise ratios attained by standard noise reduction methods. By means of a literature study, three beamformers were investigated and presented: Delay and Sum (DAS), Minimum Variance Distortionless Response (MVDR) and the Generalized Sidelobe Canceller (GSC). Due to time constraints the DAS and MVDR were chosen for further simulations and testing, based on the terms of reference. Both beamformers were extensively simulated and tested in Matlab. The results were used to compare the two beamformers based on their white noise suppression, punctuated noise source suppression, frequency range and computational complexity. Results showed that the white noise suppression for both beamformers is the same. For the suppression of punctuated noise sources however, the MVDR beamformer is more reliable, because it adapts its response to the main noise source. Furthermore, the MVDR beamformer performs better on the frequency spectrum of interest. The only advantage of the DAS beamformer over MVDR is a lower computational complexity. It is concluded that beamforming with a two microphone array is possible. Considering the terms of reference, the MVDR beamformer is the better option of the simulated two for implementation on smartphones.