The Use of Orthogonal Transforms for Improving Performance of Adaptive Filters

Daniel F. Marshall, W. Kenneth Jenkins, J. J. Murphy

    Research output: Contribution to journalArticle

    125 Citations (Scopus)

    Abstract

    The least mean square (LMS) adaptive algorithm is probably the best known and the most widely used real time adaptive filtering algorithm due to its simple computational requirements. However, as VLSI digital processors become cheaper and more readily available, the question arises as to whether more effective real time algorithms can be found that take advantage of increased computational resources as they become available. It has been shown in the literature that a real time decomposition of the incoming signal into a set of partially uncorrelated components via an orthogonal transform, and a subsequent adaptation on these individual components, leads to faster convergence rates. In this paper, transform domain processing is characterized by the effect of the transform on the shape of the mean-square error surface. It is shown that the effect of an ideal transform is to convert equal error contours, that are initially hyperellipses in the parameter space, into hyperspheres. Five specific real-valued orthogonal transforms are then compared in terms of learning characteristics and computational complexity. Since the Karhunen-Loeéve transform (KLT) is the ideal transform for this application, and since the KLT is defined in terms of the statistics of the input signal, it is certain that no fixed-parameter transform will deliver optimal learning characteristics for all input signals. However, the simulations suggest that, with a little trial and error, transforms can be found which give much improved performance in a given situation.

    Original languageEnglish (US)
    Pages (from-to)474-484
    Number of pages11
    JournalIEEE Transactions on Circuits and Systems
    Volume36
    Issue number4
    DOIs
    StatePublished - Apr 1989

    Fingerprint

    Adaptive filters
    Mathematical transformations
    Adaptive filtering
    Adaptive algorithms
    Mean square error
    Computational complexity
    Statistics
    Decomposition
    Processing

    All Science Journal Classification (ASJC) codes

    • Engineering(all)

    Cite this

    Marshall, Daniel F. ; Jenkins, W. Kenneth ; Murphy, J. J. / The Use of Orthogonal Transforms for Improving Performance of Adaptive Filters. In: IEEE Transactions on Circuits and Systems. 1989 ; Vol. 36, No. 4. pp. 474-484.
    @article{f4b770e81a6442cc82ca28f8abd5a11c,
    title = "The Use of Orthogonal Transforms for Improving Performance of Adaptive Filters",
    abstract = "The least mean square (LMS) adaptive algorithm is probably the best known and the most widely used real time adaptive filtering algorithm due to its simple computational requirements. However, as VLSI digital processors become cheaper and more readily available, the question arises as to whether more effective real time algorithms can be found that take advantage of increased computational resources as they become available. It has been shown in the literature that a real time decomposition of the incoming signal into a set of partially uncorrelated components via an orthogonal transform, and a subsequent adaptation on these individual components, leads to faster convergence rates. In this paper, transform domain processing is characterized by the effect of the transform on the shape of the mean-square error surface. It is shown that the effect of an ideal transform is to convert equal error contours, that are initially hyperellipses in the parameter space, into hyperspheres. Five specific real-valued orthogonal transforms are then compared in terms of learning characteristics and computational complexity. Since the Karhunen-Loe{\'e}ve transform (KLT) is the ideal transform for this application, and since the KLT is defined in terms of the statistics of the input signal, it is certain that no fixed-parameter transform will deliver optimal learning characteristics for all input signals. However, the simulations suggest that, with a little trial and error, transforms can be found which give much improved performance in a given situation.",
    author = "Marshall, {Daniel F.} and Jenkins, {W. Kenneth} and Murphy, {J. J.}",
    year = "1989",
    month = "4",
    doi = "10.1109/31.92880",
    language = "English (US)",
    volume = "36",
    pages = "474--484",
    journal = "IEEE Transactions on Circuits and Systems",
    issn = "0098-4094",
    publisher = "Institute of Electrical and Electronics Engineers Inc.",
    number = "4",

    }

    The Use of Orthogonal Transforms for Improving Performance of Adaptive Filters. / Marshall, Daniel F.; Jenkins, W. Kenneth; Murphy, J. J.

    In: IEEE Transactions on Circuits and Systems, Vol. 36, No. 4, 04.1989, p. 474-484.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - The Use of Orthogonal Transforms for Improving Performance of Adaptive Filters

    AU - Marshall, Daniel F.

    AU - Jenkins, W. Kenneth

    AU - Murphy, J. J.

    PY - 1989/4

    Y1 - 1989/4

    N2 - The least mean square (LMS) adaptive algorithm is probably the best known and the most widely used real time adaptive filtering algorithm due to its simple computational requirements. However, as VLSI digital processors become cheaper and more readily available, the question arises as to whether more effective real time algorithms can be found that take advantage of increased computational resources as they become available. It has been shown in the literature that a real time decomposition of the incoming signal into a set of partially uncorrelated components via an orthogonal transform, and a subsequent adaptation on these individual components, leads to faster convergence rates. In this paper, transform domain processing is characterized by the effect of the transform on the shape of the mean-square error surface. It is shown that the effect of an ideal transform is to convert equal error contours, that are initially hyperellipses in the parameter space, into hyperspheres. Five specific real-valued orthogonal transforms are then compared in terms of learning characteristics and computational complexity. Since the Karhunen-Loeéve transform (KLT) is the ideal transform for this application, and since the KLT is defined in terms of the statistics of the input signal, it is certain that no fixed-parameter transform will deliver optimal learning characteristics for all input signals. However, the simulations suggest that, with a little trial and error, transforms can be found which give much improved performance in a given situation.

    AB - The least mean square (LMS) adaptive algorithm is probably the best known and the most widely used real time adaptive filtering algorithm due to its simple computational requirements. However, as VLSI digital processors become cheaper and more readily available, the question arises as to whether more effective real time algorithms can be found that take advantage of increased computational resources as they become available. It has been shown in the literature that a real time decomposition of the incoming signal into a set of partially uncorrelated components via an orthogonal transform, and a subsequent adaptation on these individual components, leads to faster convergence rates. In this paper, transform domain processing is characterized by the effect of the transform on the shape of the mean-square error surface. It is shown that the effect of an ideal transform is to convert equal error contours, that are initially hyperellipses in the parameter space, into hyperspheres. Five specific real-valued orthogonal transforms are then compared in terms of learning characteristics and computational complexity. Since the Karhunen-Loeéve transform (KLT) is the ideal transform for this application, and since the KLT is defined in terms of the statistics of the input signal, it is certain that no fixed-parameter transform will deliver optimal learning characteristics for all input signals. However, the simulations suggest that, with a little trial and error, transforms can be found which give much improved performance in a given situation.

    UR - http://www.scopus.com/inward/record.url?scp=0024646362&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=0024646362&partnerID=8YFLogxK

    U2 - 10.1109/31.92880

    DO - 10.1109/31.92880

    M3 - Article

    AN - SCOPUS:0024646362

    VL - 36

    SP - 474

    EP - 484

    JO - IEEE Transactions on Circuits and Systems

    JF - IEEE Transactions on Circuits and Systems

    SN - 0098-4094

    IS - 4

    ER -