Mean squared error threshold prediction of adaptive maximum likelihood techniques

Research output: Contribution to journalConference article

2 Scopus citations

Abstract

Below a threshold signal-to-noise ratio (SNR), the mean squared error (MSE) performance of nonlinear maximum-likelihood (ML) estimation degrades swiftly. Threshold SNR prediction for ML signal parameter estimation requiring intermediate estimation of an unknown colored noise covariance matrix is facilitated via an interval error based method of MSE prediction. Exact pairwise error probabilities are derived, that with the Union Bound provide accurate prediction of the true interval error probabilities. A new modification of the Cramér-Rao Bound involving the analog of the Reed, Mallett, and Brennan beta loss factor appearing in the error probabilities provides excellent prediction of the asymptotic (SNR→ ∞) MSE performance of the estimator, Together, remarkably accurate prediction of the threshold SNR is obtained.

Original languageEnglish (US)
Pages (from-to)1848-1852
Number of pages5
JournalConference Record of the Asilomar Conference on Signals, Systems and Computers
Volume2
StatePublished - Dec 1 2003
Externally publishedYes
EventConference Record of the Thirty-Seventh Asilomar Conference on Signals, Systems and Computers - Pacific Grove, CA, United States
Duration: Nov 9 2003Nov 12 2003

ASJC Scopus subject areas

  • Signal Processing
  • Computer Networks and Communications

Fingerprint Dive into the research topics of 'Mean squared error threshold prediction of adaptive maximum likelihood techniques'. Together they form a unique fingerprint.

  • Cite this