Adaptive beamformer

About: Adaptive beamformer is a research topic. Over the lifetime, 4934 publications have been published within this topic receiving 93100 citations.

Adaptive detection of distributed targets with orthogonal rejection

Abstract: This study deals with the problem of detecting distributed targets in the presence of homogeneous and partially homogeneous Gaussian disturbance with unknown covariance matrix. The proposed detectors improve the adaptive beamformer orthogonal rejection test (ABORT) idea to address detection of distributed targets, which makes it possible to decide whether some observations contain a useful target or a signal belonging to the orthogonal complement of the useful subspace. At the design stage, the authors resort to either the plain generalised likelihood ratio test (GLRT) or ad hoc design procedures. Remarkably, the considered criteria lead to receivers ensuring the constant false alarm rate (CFAR) property with respect to the unknown quantities. Moreover, authors’ derivations show that the ad hoc detector for a partially homogeneous environment coincides with the generalised adaptive subspace detector. The performance assessment conducted by Monte Carlo simulation has confirmed the effectiveness of the newly proposed detection algorithms.

SINR Analysis of the Subtraction-Based SMI Beamformer

Abstract: The sample matrix inversion (SMI) beamformer suffers from performance degradation due to the finite sample size effect and the arrival angle mismatch problem. A simple technique to provide robustness to the conventional SMI beamformer is to block the desired signal from the received data before calculating the beamformer's weight vector, which leads to the subtraction-based SMI (S-SMI) beamformer. In this correspondence, closed-form approximations of the expected value of the signal-to-interference-plus-noise ratio (SINR) for the S-SMI beamformer and the SMI beamformer are derived, where the effect of both finite sample size and arrival angle mismatch are considered. Simulations show that the derived approximations are close enough to represent the true values of the SINR, when the sample size is small and the arrival direction mismatch exists.

Broadband beamforming for a microphone array

Abstract: An array of sensors can be used in conjunction with a beamformer, which processes the sensor signals, to achieve a directional response. The beamformer has to be designed such that a beam pattern with certain desired characteristics like specific main beam direction, defined main beam shape, and desired sidelobe level is formed. Conventional methods for the design of beamformers assume sensors with ideal features and do not take the disturbance of the sound field, caused by the mounting of the array, into account. Therefore the predicted theoretical polar response and the measured response often differ significantly. This paper presents a method for optimizing a beamformer for a one‐dimensional microphone array, taking into consideration nonideal features of the sensors and the mounting. Thus the actual polar response can be improved. By evaluating cross‐correlation functions of the sensor signals during a calibration procedure in an anechoic chamber and minimizing the mean squared error between the beamformer output and a prescribed response, optimum parameters for the beamformer are assigned.

Adaptive array beamforming based on an efficient technique

Abstract: This paper presents a new processing technique for adaptive array beamforming. We first properly decompose the optimal weight vector associated with the original array beamformer into several adaptive weight subvectors. An adaptive algorithm based on the conventional RLS algorithm is derived to update the weight subvectors. The required computational complexity is also evaluated for the proposed technique. The convergence rate of the proposed technique is mainly affected by the size of the largest weight subvector. Faster convergence speed and less array sensitivity to the pointing errors can be achieved by using the proposed technique. Several simulation results are included for illustration and comparison.

CSA-BF: a constrained switched adaptive beamformer for speech enhancement and recognition in real car environments

Abstract: While a number of studies have investigated various speech enhancement and processing schemes for in-vehicle speech systems, little research has been performed using actual voice data collected in noisy car environments. In this paper, we propose a new constrained switched adaptive beamforming algorithm (CSA-BF) for speech enhancement and recognition in real moving car environments. The proposed algorithm consists of a speech/noise constraint section, a speech adaptive beamformer, and a noise adaptive beamformer. We investigate CSA-BF performance with a comparison to classic delay-and-sum beamforming (DASB) in realistic car conditions using a corpus of data recorded in various car noise environments from across the U.S. After analyzing the experimental results and considering the range of complex noise situations in the car environment using the CU-Move corpus, we formulate the three specific processing stages of the CSA-BF algorithm. This method is evaluated and shown to simultaneously decrease word-error-rate (WER) for speech recognition by up to 31% and improve speech quality via the SEGSNR measure by up to +5.5 dB on the average.