The Experts below are selected from a list of 306 Experts worldwide ranked by ideXlab platform

S F Hsieh - One of the best experts on this subject based on the ideXlab platform.

  • vlsi algorithms and architectures for complex Householder Transformation with applications to array processing
    Signal Processing Systems, 1992
    Co-Authors: C F T Tang, S F Hsieh
    Abstract:

    The Householder Transformation is considered to be desirable among various unitary Transformations due to its superior computational efficiency and robust numerical stability. Specifically, the Householder Transformation outperforms the Givens rotation and the modified Gram-Schmidt methods in numerical stability under finite-precision implementations, as well as requiring fewer arithmetical operations. Consequently, the QR decomposition based on the Householder Transformation is promising for VLSI implementation and real-time high throughput modern signal processing. In this paper, a recursive complex Householder Transformation (CHT) with a fast initialization algorithm is proposed and its associated parallel/pipelined architecture is also considered. Then, a CHT based recursive least-squares algorithm with a fast initialization is presented. Its associated systolic array processing architecture is also considered.

  • Systolic block Householder Transformation for RLS algorithm with two-level pipelined implementation
    IEEE Transactions on Signal Processing, 1992
    Co-Authors: S F Hsieh
    Abstract:

    The authors propose a systolic block Householder Transformation (SBHT) approach to implement the HT on a systolic array and also propose its application to the RLS (recursive least squares) algorithm. Since the data are fetched in a block manner, vector operations are in general required for the vectorized array. However, a modified HT algorithm permits a two-level pipelined implementation of the SBHT systolic array at both the vector and word levels. The throughput rate can be as fast as that of the Givens rotation method. The present approach makes the HT amenable for VLSI implementation as well as applicable to real-time high-throughput applications of modern signal processing. The constrained RLS problem using the SBHT RLS systolic array is also considered.

  • Systolic implementations of up/down-dating cholesky factorization using vectorized Gram-Schmidt pseudo orthoganalization
    Journal of VLSI signal processing systems for signal image and video technology, 1991
    Co-Authors: S F Hsieh
    Abstract:

    We propose a new class of hyperbolic Gram-Schmidt methods to simultaneously update and downdate the Cholesky factor of a sample covariance matrix efficiently with applications to sliding window recursive least squares (RLS) filtering problems. Several vectorized versions of this Gram-Schmidt approach are introduced, which include conventional column-updating, modified row/column-updating, and square-root-free methods. Comparisons to the existing known methods, such as Householder Transformation and Givens rotation, are also given. Upon further reformulating these algorithms, a systolic triarray structure is proposed to facilitate VLSI implementations.

  • Recursive LS filtering using block Householder Transformations
    International Conference on Acoustics Speech and Signal Processing, 1990
    Co-Authors: S F Hsieh
    Abstract:

    A systolic block Householder Transformation is proposed to implement the HT (Householder Transformation) on a systolic array as well as to apply it to the RLS algorithm. Since the data are fetched in a block manner, vector operations are in general required for the vectorized array. The approach makes the HT suitable for VLSI implementation as well as applicable to real-time high-throughput applications of modern signal processing.

Jun-ichi Imura - One of the best experts on this subject based on the ideXlab platform.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    Proceedings of the IEEE Conference on Decision and Control, 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interacted according to a network. In this method, the structure of spatially one- dimensional reaction-diffusion that a SISO linear dynamical network intrinsically has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    49th IEEE Conference on Decision and Control (CDC), 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interconnected according to a network. In this method, the structure of spatially one-dimensional reaction-diffusion that a SISO linear dynamical network has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.

Peter Strobach - One of the best experts on this subject based on the ideXlab platform.

  • The QS-Householder Sliding Window
    2020
    Co-Authors: Bi-svd Subspace Tracker, Peter Strobach
    Abstract:

    A fast algorithm for computing the sliding window Bi-SVD subspace tracker is introduced. This algorithm produces, in each time step, a dominant rank- SVD subspace approximant of an rectangular sliding window data matrix. The method is based on the (orthonormal-square) decomposition. It uses two row-Householder Transformations for updating and one nonorthogonal Householder Transformation for downdating in each time step. The resulting algorithm is long-term stable and shows excellent numerical and structural properties, as known from pure Householder-type algorithms. The dominant com- plexity is multiplications per time update, which is also the lower bound in dominant complexity for an algorithm of this kind. A completely self-contained algorithm summary is provided and a Fortran subroutine of the algorithm is available for download from http://webuser.hs-furtwangen.de/~strobach/qsh- bisvd.for.

  • The Householder Compressor Theorem and its application in subspace tracking
    Signal Processing, 2009
    Co-Authors: Peter Strobach
    Abstract:

    The energy in a symmetric (r+1)x(r+1) matrix can be transformed perfectly into a smaller rxr submatrix by means of a two-sided Householder Transformation. The parameters of this Householder Transformation are uniquely determined by the minor eigenvector of the larger matrix. This compressor is the key to a new type of square-root Householder subspace tracker which is optimal in terms of both complexity and performance. Computer experiments validate the theoretical findings.

  • The QS-Householder Sliding Window Bi-SVD Subspace Tracker
    IEEE Transactions on Signal Processing, 2009
    Co-Authors: Peter Strobach
    Abstract:

    A fast algorithm for computing the sliding window bi-SVD subspace tracker is introduced. This algorithm produces, in each time step, a dominant rank-r SVD subspace approximant of an L timesN rectangular sliding window data matrix. The method is based on the QS (orthonormal-square) decomposition. It uses two row-Householder Transformations for updating and one nonorthogonal Householder Transformation for downdating in each time step. The resulting algorithm is long-term stable and shows excellent numerical and structural properties, as known from pure Householder-type algorithms. The dominant complexity is 4Lr +3Nr multiplications per time update, which is also the lower bound in dominant complexity for an algorithm of this kind. A completely self-contained algorithm summary is provided and a Fortran subroutine of the algorithm is available for download from http://webuser.hs-furtwangen.de/~strobach/qsh-bisvd.for.

Takayuki Ishizaki - One of the best experts on this subject based on the ideXlab platform.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    Proceedings of the IEEE Conference on Decision and Control, 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interacted according to a network. In this method, the structure of spatially one- dimensional reaction-diffusion that a SISO linear dynamical network intrinsically has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    49th IEEE Conference on Decision and Control (CDC), 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interconnected according to a network. In this method, the structure of spatially one-dimensional reaction-diffusion that a SISO linear dynamical network has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.

Kenji Kashima - One of the best experts on this subject based on the ideXlab platform.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    Proceedings of the IEEE Conference on Decision and Control, 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interacted according to a network. In this method, the structure of spatially one- dimensional reaction-diffusion that a SISO linear dynamical network intrinsically has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.

  • Extraction of 1-dimensional reaction-diffusion structure in SISO linear dynamical networks
    49th IEEE Conference on Decision and Control (CDC), 2010
    Co-Authors: Takayuki Ishizaki, Kenji Kashima, Jun-ichi Imura
    Abstract:

    In this paper, we propose a model order reduction method for SISO linear dynamical networks, where a large number of subsystems are interconnected according to a network. In this method, the structure of spatially one-dimensional reaction-diffusion that a SISO linear dynamical network has is extracted by way of Householder Transformation ordering the state variables according to the distance from the source (i.e., an input) of the diffusion. Based on this structure, a model order reduction method with the diffusion structure of the system preserved is presented, which can be applied for large-scale systems. In addition, an easily-computable error bound via the proposed model reduction is derived.