Back to results
US 5,495,256 AGranted
Method of efficiently computing maximum likelihood bearing estimator
OwnerJohn E. Piperinventor
Life of this patent
Estimating the term…
Abstract
The maximum likelihood function can be simplified by exploiting the symmetry in the equations. This new representation leads to a very significant decrease in the computational load for bearing estimation and spectral analysis applications.
Claims3
A method of determining bearing estimation parameters from a uniform array of sensors, said method comprising:
- (a) sampling and digitizing outputs from said uniform array of sensors;
- (b) forming a sample covariance matrix from said outputs of said uniform array of sensors;
- (c) averaging diagonals of said sample covariance matrix for obtaining a sample covariance vector (R.sub.m);
- (d) providing a projection vector (P.sub.m) where ##EQU16## for K.sub.1 to be a wavenumber for a first source, k.sub.2 to be a second wavenumber for a second source and k=k.sub.1 -k.sub.2 ;
- (e) computing a likelihood function (L) by multiplying said sample covariance vector with said projection vector;
- (f) solving a maxima of said likelihood function with respect to said wavenumber and said second wavenumber;
- and
- (g) resulting said wavenumber and said second wavenumber at said maxima for providing a best estimate of bearings.
5 more elements
A method of determining frequency estimation parameters, said method comprising:
- (a) forming a time series vector by sampling and digitizing outputs from a sensor;
- (b) forming a sample covariance matrix from said outputs of said time series vector;
- (c) averaging diagonals of said sample covariance matrix for obtaining a sample covariance vector (R.sub.m);
- (d) providing a projection vector (P.sub.m) where ##EQU17## for k.sub.1 to be a digital frequency for a first source, k.sub.2 to be a second digital frequency for a second source and k=k.sub.1 -k.sub.2 ;
- (e) computing a likelihood function (L) by multiplying said sample covariance vector with said projection vector;
- (f) solving a maxima of said likelihood function with respect to said wavenumber and said second wavenumber;
- and
- (g) resulting said wavenumber and said second wavenumber at said maxima for providing a best estimate of frequencies.
5 more elements
Dependencies are inferred from the claim text.