Elsevier

Signal Processing

Volume 59, Issue 2, June 1997, Pages 221-233
Signal Processing

Regular paper
An alternative approach to coherent source location problem

https://doi.org/10.1016/S0165-1684(97)00048-0Get rights and content

Abstract

We propose a novel preprocessing scheme, referred to as (eigen) vector smoothing, as an alternative to the conventional spatial smoothing for solving the multiple source location problem involving coherent sources or a rank-deficient source covariance matrix. The essence of the technique is to preprocess the signal subspace eigenvectors rather than the covariance matrix as in spatial smoothing. It is shown by analysis and computer simulations that these two approaches are related, and that they have comparable estimation performances when employed with the MUSIC-type DOA estimators. Additionally, eigenvector smoothing offers advantages in terms of computational simplicity and flexibility. The latter is especially true with eigenstructure DOA estimators in adaptive estimation problems, i.e., when the signal subspace eigenvectors are updated using fast adaptive algorithms.

Zusammenfassung

Wir stellen eine neue Methode zur Vorverarbeitung von Sensorgruppensignalen vor. Dieses (eigen-)vector smoothing kann an Stelle des bekannten spatial smoothing bei Richtungsschätzproblemen eingesetzt werden, bei denen kohärente Quellen auftreten oder die Kovarianzmatrix der Quellen nicht vollen Rang besitzt. Der Kern der neuen Methode ist, daβ die Eigenvektoren und nicht die Kovarianzmatrix vorverarbeitet wird. Analytisch und durch Simulationen wird die Ähnlichkeit der beiden Methoden gezeigt. Sie liefern nahezu gleiche Schätzungen, wenn man sie zusammen mit MUSIC-Algorithmen zur Richtungsschätzung einsetzt. Der Vorteil der neuen Methode ist die einfache und flexible Implementierung. Letzteres ist besonders dann von Vorteil, wenn man adaptive Schätzprobleme durch Eigenvektortechniken löst, d.h., den Signalunterraum mit Hilfe von schnellen adaptiven Algorithmen aktualisiert.

Résumé

Nous proposons une nouveau schéma de pré-traitement, appelé plus loin lissage pur vecteurs propres, comme alternative au lissage spatial conventionnel pour résoudre le problème de la localisation de sources multiples impliquant des sources cohérentes ou une matrice de covariance de sources déficiente en rang. La base de la technique consiste à pré-traiter les vecteurs propres du sous-espace du signal plutôt que la matrice de covariance, comme dans le lissage spatial. L'analyse et les simulations informatiques montrent que ces deux approches sont liées, et qu'elles ont des performances d'estimation comparables lorsqu'elles sont employées avec les estimateurs DOA du type MUSIC. De plus, le lissage par vecteurs propres offre des avantages en termes de simplicité de calcul et de flexibilité. Cette dernière caractéristique est particulièrement vraie avec les estimateurs DOA à structure propre dans les problèmes d'estimation adaptative, autrement dit quand les vecteurs propres du sous-espace sont mis à jour en utilisant des algorithmes adaptatifs rapides.

References (13)

  • D. Grenier et al.

    A new spatial smoothing scheme for direction-of-arrivals of correlated sources

    Signal Processing

    (1996)
  • A.J. Barabell

    Improving the resolution performance of eigenstructure-based direction-finding algorithms

  • J. Cardzow

    A high-resolution direction-of-arrival algorithm for narrow-band coherent and incoherent sources

    IEEE Trans. Acoust. Speech Signal Process.

    (1988)
  • R.D. DeGroat

    Noniterative subspace tracking

    IEEE Trans. Signal Process.

    (1992)
  • J.E. Evans et al.

    High resolution angular spectrum estimation techniques for terrain scattering analysis and angle of arrival estimation

  • A.B. Gershman et al.

    Optimal subarray size for spatial smoothing

    IEEE Signal Process. Lett.

    (1995)
There are more references available in the full text version of this article.

Cited by (14)

  • Source localization utilizing weighted power iterative compensation via acoustic vector hydrophone array

    2021, Applied Acoustics
    Citation Excerpt :

    However, the DOA estimation of coherent sources is inevitable, especially in various underwater reflective surfaces, such as water bottom and surface, and underwater obstacle, e.g., which deteriorate the DOA estimation performance dramatically. To mitigate this issue, a large number of decorrelation techniques are proposed in [21–24]. Although these techniques improve the DOA estimation performance of coherent sources, they can’t distinguish more sources than the number of physical AVHs.

  • Two-dimensional DOA estimation of coherent signals using acoustic vector sensor array

    2012, Signal Processing
    Citation Excerpt :

    Under this circumstance, the rank of the covariance matrix is generally less than the number of incident signals. To overcome this detrimental aspect, decorrelation techniques such as the spatial smoothing (SS) [10–12], the technique developed by Kozick and Kassam [13], eigenvector smoothing (ES) [14,15], and computationally efficient subspace-based method without eigendecomposition (SUMWE) [16] have been proposed. These techniques, however, work well only for certain array configurations, e.g., uniformly spaced linear arrays.

  • Sensor Array Signal Processing Using SSA

    2023, Lecture Notes on Data Engineering and Communications Technologies
View all citing articles on Scopus

The work of A.B. Gershman was supported by the grant Bo 568/22-1 of Deutsche Forschungsgemeinschaft (DFG), and by the SASPARC project of INTAS.

View full text