Analysis on the MDL source under-enumeration based on critical eigenvalue method

PDF(2071 KB)

Journal of Vibration and Shock ›› 2016, Vol. 35 ›› Issue (21) : 113-119.

Du Fei Li Yi-bo Jin Shi-jiu Zeng Zhou-mo

Author information +

History +

Abstract

A method based on the critical eigenvalue is proposed to evaluate the underestimation problem of minimum description length criterion (MDL) source enumeration. The uniqueness of critical eigenvalue has been proved and the solution expression is presented by the analysis on the enumeration algorithm. The critical eigenvalue method can predict the boundary of underestimation for different numbers of undetected sources and access the influences of array parameters on the performance of source enumeration comparing with the existing methods. An approximation method is proposed to reduce the calculation complexity of solving the critical eigenvalues for large numbers of undetected sources and the factors on the approximation errors are also analyzed. Simulation results indicate that the critical eigenvalue method can exactly describe the boundary of underestimation of source enumeration. Similar analysis can be also performed on other source enumeration methods based on information criteria.

Key words

minimum description length criterion / source enumeration / underestimation / array signal processing

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

Du Fei Li Yi-bo Jin Shi-jiu Zeng Zhou-mo. Analysis on the MDL source under-enumeration based on critical eigenvalue method[J]. Journal of Vibration and Shock, 2016, 35(21): 113-119

References

[1] Krim H, Viberg M. Two decades of array signal processing research: The parametric approach [J]. IEEE Signal Processing Magazine, 1996, 13(4): 67-94.
[2] Aouada S, Zoubir A M, See C M S. A comparative study on source number detection [C]// IEEE. Signal Processing and Its Applications, 2003. Proceedings. Seventh International Symposium on. USA: IEEE, 2003. 173-176.
[3] Akaike H. Information theory and an extension of the maximum likelihood principle [C]// Petrov B N, Csaki F. 2nd Int. Symp. Inform. Theory, suppl. Problems of Control and Inform. Theory. Budapest: Akadémiai Kiadó. 1973. 267-281.
[4] Rissanen J. Modeling by shortest data description [J]. Automatica, 1978, 14(5): 465-471.
[5] Schwarz G. Estimating the dimension of a model [J]. Annals of Statistics, 1978, 6(2): 461-464.
[6] Wax M, Kailath T. Detection of signals by information theoretic criteria [J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1985, 33(2): 387–392.
[7] Zhang Q T, Wong K M, Yip P C, et al. Statistical analysis of the performance of information theoretic criteria in the detection of the number of signals in array processing [J]. IEEE Transactions on Acoustics, Speech, and Signal Processing, 1989, 37(10): 1557-1567.
[8] Kritchman S, Nadler B. Non-parametric detection of the number of signals: Hypothesis testing and random matrix theory [J]. IEEE Transactions on Signal Processing, 2009, 57(10): 3930-3941.
[9] Nadler B. Nonparametric detection of signals by information theoretic criteria: Performance analysis and an improved estimator [J]. IEEE Transactions on Signal Processing, 2010 58(5): 2746–2756.
[10] Lu Z, Zoubir A M. Flexible Detection Criterion for Source Enumeration in Array Processing [J]. IEEE Transactions on Signal Processing, 2013, 61(6):1303-1314.
[11] Hannan E J, Quinn B G. The determination of the order of an autoregression [J]. Julia Campos Fernández, 1979, 41(2): 190-195.
[12] Wu T J, Sepulveda A. The weighted average information criterion for order selection in time series and regression models [J]. Statistics & Probability Letters, 1998, 39(1): 1-10.
[13] Nadakuditi R R, Edelman A. Sample eigenvalue based detection of high-dimensional signals in white noise using relatively few samples[J]. IEEE Transactions on Signal Processing, 2008, 56(7): 2625-2638.