초록 ▼

A method is specified for the high-resolution evaluation and, if appropriate, for the recovery of narrow-band signals, received by a centrosymmetrical sensor group (AG), for directional estimation. A method is also specified for the high-resolution evaluation of superimposed non-attenuated harmonic

A method is specified for the high-resolution evaluation and, if appropriate, for the recovery of narrow-band signals, received by a centrosymmetrical sensor group (AG), for directional estimation. A method is also specified for the high-resolution evaluation of superimposed non-attenuated harmonic signals for spectral analysis in the case, if appropriate, of multi-channel observations. For directional estimation, signals which are noisy and disturbed by multipath propagation can be prepared for recovery. The method for directional estimation is particularly suitable for signal evaluation in mobile radio systems or wireless communication systems, in high-resolution radar image processing systems, sonar systems, and medical technology systems. The method for frequency estimation is suitable for image processing applications, and for a multiplicity of spectral analysis applications.

대표청구항 ▼

[ What is claimed is:] [1.] A method for high-resolution evaluation of signals for one-dimensional or two-dimensional directional estimation in a device for digital signal processing to which is assigned a centrosymmetrical sensor group having a number of sensors, comprising the steps of:storing sam

[ What is claimed is:] [1.] A method for high-resolution evaluation of signals for one-dimensional or two-dimensional directional estimation in a device for digital signal processing to which is assigned a centrosymmetrical sensor group having a number of sensors, comprising the steps of:storing sample values currently measured in the sensors in a complex measured-value matrix (x) having a dimension (M.times.N) determined by the number of the sensors and a number of sampling instances;combining an M-dimensional antidiagonal permutation matrix (II.sub.M) with the conjugate complex measured-valued matrix (x*) and an N-dimensional antidiagonal permutation matrix (II.sub.N) to form a first matrix (II.sub.M x*II.sub.N);determining a centrohermitian matrix (x II.sub.M x*II.sub.N) from the complex measured-value matrix (x) and the first matrix (II x*II.sub.N);determining via an M-dimensional left II-real, adjoint matrix (Q.sub.M.sup.H) and a 2N-dimensional left II-real matrix (Q.sub.2N), starting from the centrohermitian matrix (x II.sub.M x*II.sub.N), a second, purely real matrix (T(x)) of a form T(x)=Q.sub.M.sup.H (x II.sub.M x*II.sub.N)Q.sub.2N which contains exclusively real values, which can be assigned to the measured values and --which has double the number of elements;undertaking by processing the second, purely real matrix (T(x)), a signal subspace estimation to determine a real signal subspace matrix (E.sub.S) with d dominant vectors of the second, purely real matrix (T(x));undertaking a subgroup formation, undertaken separately for each evaluation dimension x,y of the method, of the centrosymmetrical sensor group into two subgroups displaced relative to one another, and determining two selection matrices (K.sub.1,K.sub.2,K.sub..mu.1,2,K.sub..nu.1,2) for each evaluation dimension x,y in accordance with configuration of the subgroups;calculating a solution, undertaken separately for each evaluation dimension x,y of the method, of system equations prescribed by the signal subspace matrix(E.sub.S), produced from the signal subspace estimation, and the selection matrices (K.sub.1,K.sub.2,K.sub..mu.1,2,K.sub..nu.1,2) so that in each case a solution matrix (Y,Y.sub..mu.,Y.sub..nu.) is available in accordance with K.sub.1 E.sub.S Y=K.sub.2 E.sub.S ;determining an eigenvalue matrix (.OMEGA.,.LAMBDA.) from the solution matrix (Y,Y.mu.,Y.nu.), depending on the dimension x,y of the method; andthe eigenvalues (.omega..sub.k,.lambda..sub.k) determined from the eigenvalue matrix (.OMEGA.,.LAMBDA.) representing directional or frequency estimates.