Distributed Delay and Sum Beamformer in Regular Networks Based on Synchronous Randomized Gossip

Konferenz: IWAENC 2012 - International Workshop on Acoustic Signal Enhancement
04.09.2012-06.09.2012 in Aachen, Germany

Tagungsband: IWAENC 2012

Seiten: 4Sprache: EnglischTyp: PDF

Persönliche VDE-Mitglieder erhalten auf diesen Artikel 10% Rabatt

Zeng, Yuan; Hendriks, Richard C. (Signal and Information Processing Lab, Delft University of Technology, The Netherlands)

In this paper, we review and propose an improvement on an earlier proposed distributed delay and sum beamformer (DDSB). In the original DDSB, each pair of neighboring nodes updates their estimates asynchronously, which may yield a relatively slow convergence, especially when the size of network is large. When multiple pairs of neighboring nodes update simultaneously, the algorithm converges potentially faster. We propose an improved synchronous DDSB in a regular network based on synchronous randomized gossip. Moreover, we provide a comparison of the convergence rate of the asynchronous DDSB with the original and with the improved synchronous DDSB. The simulation results show the faster convergence results and show that the DDSB using the different updating schemes converges to the centralized beamformer. Index Terms — Distributed delay and sum beamformer, synchronous randomized gossip, speech enhancement, sensor networks