An alternative implementation of the superdirective beamformer
Jöerg Bitzer, Karl Dirk Kammeyer, Klaus Uwe Simmer
Abstract
Jöerg Bitzer, Karl Dirk Kammeyer, Klaus Uwe Simmer
Abstract
In this contribution we introduce a new implementation of superdirective beamformers. The new structure has the advantage of reduced computational complexity. This advantage is due to a GSClike (Generalized Sidelobe Canceller) scheme. Unlike the conventional GSC, the filters in the sidelobe cancelling path are fixed and can be computed in advance by using the Wiener solution. The new structure yields exactly the same noise reduction performance as the superdirective beamformer does. 1. INTRODUCTION Adaptive arrays have often been used for speech enhancement in the last decade, but the reported results are not very encouraging. On the other hand, fixed superdirective beamformers have become a renewed research field in the speech processing area [1, 2]. The main advantage of fixed beamformers is that no superresolution problems occur, and the problem of sensor noise can be solved by constrained design. Cox et al. [3] have shown that the design of superdirective beamformers and Frost's a...
OpenAlex reports 56 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
In this contribution we introduce a new implementation of superdirective beamformers. The new structure has the advantage of reduced computational complexity. This advantage is due to a GSClike (Generalized Sidelobe Canceller) scheme. Unlike the conventional GSC, the filters in the sidelobe cancelling path are fixed and can be computed in advance by using the Wiener solution. The new structure yields exactly the same noise reduction performance as the superdirective beamformer does. 1. INTRODUCTION Adaptive arrays have often been used for speech enhancement in the last decade, but the reported results are not very encouraging. On the other hand, fixed superdirective beamformers have become a renewed research field in the speech processing area [1, 2]. The main advantage of fixed beamformers is that no superresolution problems occur, and the problem of sensor noise can be solved by constrained design. Cox et al. [3] have shown that the design of superdirective beamformers and Frost's a...
Key concepts: Computer science, Scheme (mathematics), Reduction (mathematics), Path (computing), Computational complexity theory, Noise reduction, Noise (video), Wiener filter