A subspace-based tracking algorithm is proposed for high-resolution DOA estimation, using the generalized singular value decomposition (GSVD) of the sample data matrices. A GSVD updating procedure is presented. With this procedure, a new approximate decomposition can be computed from previous one, with finite operations at each iteration. Combined with exponetial weighting technique, this algorithm can solve DOA estimation tracking problems efficiently.
Roy R,Kailath T. ESPRIT-estimation of signal parameters via rotational invariance techniques.[2]IEEE Trans. on ASSP, 1989, ASSP-37(7): 984-995.[3]Moonen M, Van Dooren P, Vandewalle J. Updating singular value decomposition: A parallel im-[4]plementation, SPIE 1152, Advanced algorithms and architectures for signal processing IV, 1989, 80-91.[5]Moonen M, Van Dooren P, Vandewalle J. A singular value decomposition algorithm for subspace tracking, SIAM J[J].Matrix and Anal. Appl.1992, 13(4):1015-1038[6]Moonen M, Van Dooren P, Vandewalle J. A systolic array for SVD updating, SIAM J[J].Matrix Anal. Appl.1993, 14(2):353-374[7]Paige C C, Saunders M A. Towards a generalized singular value decomposition, SIAM J[J].Numer. Anal.1981, 18(2):398-405[8]Paige C C. Computing the generalized singular value decomposition, SIAM J[J].Sci. Stat. Comput.1986, 7(4):1126-1146[9]Golub G H, Van Loan C F. Matrix Computations, Baltimore MD: The Johns Hopkina University Press, 1983, Chapter 12.