Abstract—In this paper a novel approach derived from the Mapped Real Transform (MRT) is proposed for improving the convergence rate and reducing the computational complexity of the Least Mean Square (LMS) Beamformer. In contrast to the conventional LMS beamformer which applies the LMS algorithm directly on the data received from the sensors, the proposed method utilizes a transformed version of the received data for the LMS adaptation. This transformed version is obtained by applying a One Dimensional Reduced Mapped Real Transform algorithm to the received data. Simulations show that the computational complexity is reduced and the convergence rate is improved when compared with the conventional LMS beamformers. The One Dimensional Reduced Mapped Real Transform based LMS algorithm can be used in radar, seismology, sonar, biomedical applications etc. As a future scope the One Dimensional Reduced Mapped Real Transform can be applied to other adaptive beamformers like Normalized Least Mean Square beamformer and Recursive Least Square beamformer.

 

Index Terms—LMS beamformer, mapped real transform, one dimensional reduced mapped real transform, convergence rate, computational complexity

Cite: P. R. Mini, S. Mridula, Binu Paul, and P. Mohanan, "A Novel Approach for Improving Performance of LMS Beamformer," International Journal of Signal Processing Systems, Vol. 4, No. 2, pp. 118-123, April 2016. doi: 10.12720/ijsps.4.2.118-123