Apparatus for estimating direction of arrival wave

Apparatus for estimating direction of arrival wave, for estimating arrival directions of arrival waves from received data of two or more said arrival waves which are received by two or more antennas, comprising:

• means for computing correlation matrix, for computing a correlation matrix from said received data of two or more said arrival waves which are received by two or more said antennas;
• eigenvalue decomposition operating means, for implementing a Householder transformation operation on said correlation matrix which has been computed by said means for computing correlation matrix, and for decomposing said correlation matrix into eigenvectors and eigenvalues;
• matrix producing means provided at said eigenvalue decomposition operating means, for producing a matrix for obtaining said eigenvalues for said correlation matrix and a matrix for obtaining said eigenvectors therefor;
• column norm operating means provided at said eigenvalue decomposition operating means, for respectively computing column norms of elements of respective columns of said matrix which is a subject for a Householder transformation operation every implementation of said Householder transformation operation in said matrix for obtaining said eigenvalues;
• means for judging maximum value of column norm provided at said eigenvalue decomposition operating means, for judging whether or not a maximum value of said column norms of said matrix which is said subject for said Householder transformation operation, which has been computed by said column norm operating means is the same as or lower than a predetermined threshold value every implementation of said Householder transformation operation;
• means for implementing Householder transformation operation provided at said eigenvalue decomposition operating means, for repeating said Householder transformation operation on said matrix for obtaining said eigenvalues only if said means for judging maximum value of column norm judges said maximum value of said column norms of said matrix which is said subject for said Householder transformation operation, which has been computed by said column norm operating means not to be the same as or not to be lower than said predetermined threshold value;
• means for controlling implementation of transformation operation provided at said eigenvalue decomposition operating means, for breaking implementation of said Householder transformation operation by said means for implementing Householder transformation operation, if said means for judging maximum value of column norm judges said maximum value of said column norms of said matrix which is said subject for said Householder transformation operation, which has been computed by said column norm operating means to be the same as or lower than said predetermined threshold value;
• means for implementing convergence operation provided at said eigenvalue decomposition operating means, for further implementing said Householder transformation operations by said means for implementing Householder transformation operation and said means for controlling implementation of transformation operation on said matrix for obtaining said eigenvalues on which implementation of said Householder transformation operation has been broken by said means for controlling implementation of transformation operation up to a predetermined time, and for obtaining said eigenvectors and said eigenvalues of said correlation matrix on the basis of a result of said Householder transformation operations;
• means for judging signal rank, for judging a number of said arrival waves on the basis of said implementation time of said Householder transformation operations by said means for implementing Householder transformation operation;
• and
• means for setting threshold value, for setting said predetermined threshold value so as to be higher than said column norm of said matrix for obtaining said eigenvalue when only noises are caught in said received data.