Title :
Eigenstructure beamspace root estimator bank with interpolated array
Author :
Gershman, Alex B. ; Böhme, Johann F.
Author_Institution :
Signal Theory Group, Ruhr-Univ., Bochum, Germany
Abstract :
A beamspace root modification of pseudorandom joint estimation strategy (PR-JES) is developed. The essence of PR-JES is to generate the eigenstructure-based estimator bank for given sample covariance or data matrix. Combining the results of “parallel” underlying estimators, PR-JES removes the outliers and improves the threshold performance. In the case of a non-uniform array, the interpolated array approach is used to enable the application of root underlying estimators. Simulations and results of real ultrasonic data processing show that the proposed beamspace root implementation significantly outperforms spectral elementspace PR-JES and achieves a performance similar or better than that of the stochastic ML method
Keywords :
acoustic signal processing; array signal processing; covariance matrices; direction-of-arrival estimation; eigenvalues and eigenfunctions; random processes; spectral analysis; ultrasonics; beamspace root modification; data matrix; eigenstructure beamspace root estimator bank; interpolated array; nonuniform array; parallel underlying estimators; pseudorandom joint estimation strategy; sample covariance matrix; simulations; spectral elementspace PR-JES; spectral estimators; stochastic ML method; threshold performance; ultrasonic data processing; Computational efficiency; Covariance matrix; Data processing; Degradation; Direction of arrival estimation; Eigenvalues and eigenfunctions; Multiple signal classification; Roentgenium; Sensor arrays; Stochastic processes;
Conference_Titel :
Acoustics, Speech and Signal Processing, 1998. Proceedings of the 1998 IEEE International Conference on
Conference_Location :
Seattle, WA
Print_ISBN :
0-7803-4428-6
DOI :
10.1109/ICASSP.1998.681538
