By Topic

Signal Processing, IEEE Transactions on

Issue 6 • Date Jun 1995

Filter Results

Displaying Results 1 - 25 of 27
  • On the first order equalization of hidden Markov models

    Publication Year: 1995 , Page(s): 1522 - 1524
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (236 KB)  

    The article exploits some properties of the “first order equalization” technique, used to increase the recognition performance of speech. It shows that if the eigenvalues of the covariance matrix of observations are restricted, then a solution exists for this problem, although it may not be unique, View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Source number estimators using transformed Gerschgorin radii

    Publication Year: 1995 , Page(s): 1325 - 1333
    Cited by:  Papers (84)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (744 KB)  

    We introduce effective uses of Gerschgorin radii of the unitary transformed covariance matrix for source number estimation. There are two approaches, likelihood and heuristic, used for developing the detection criteria. The likelihood approach combines the Gerschgorin radii to the well-known source number detectors and improves their detection performances for Gaussian and white noise processes. It is verified that the Gerschgorin likelihood estimators (GLE) are consistent. The Gerschgorin AIC yields a consistent estimate and the Gerschgorin MDL criterion does not tend to underestimate for small or moderate data samples. The heuristic approach applying the Gerschgorin disk estimator (GDE) developed from the projection concept, overcomes the problem in cases of small data samples, an unknown noise model, and data dependency. Furthermore, the detection performances of both approaches through the suggested rotations and averaging can be further improved. Finally, the proposed and existing criteria are evaluated in various conditions by using simulated and measured experimental data View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Sinc interpolation of discrete periodic signals

    Publication Year: 1995 , Page(s): 1502 - 1503
    Cited by:  Papers (21)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (160 KB)  

    The paper introduces a method for the sinc interpolation of discrete periodic signals. The convolution of the sinc kernel with the infinite sequence of a periodic function is rewritten as a finite summation. The method is equivalent to trigonometrical interpolation by Fourier series expansion View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Processors for generalized stack filters

    Publication Year: 1995 , Page(s): 1541 - 1546
    Cited by:  Papers (12)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (332 KB)  

    New processor structures for generalized stack filters are proposed. They can be implemented using different numbers of Boolean function units. The class of pipeline-parallel structures for generalized stack filters is simple and modular in structure, and suitable for VLSI implementation. Coder and decoder networks are developed for the mutual transform of binary-weighted and unary-weighted codes as the thresholding and addition units of the processors. The area time complexity A·T of the proposed filters is O(k3) due to the feedback, instead of O(2k) for the threshold decomposition structure, where k is the number of bits for sample representation View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Bootstrap multiple tests applied to sensor location

    Publication Year: 1995 , Page(s): 1386 - 1396
    Cited by:  Papers (15)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (988 KB)  

    A method for finding optimal locations of vibration sensors within a group of sensors for detecting knock in combustion engines is proposed. It differs from other techniques in that only signal processing and statistical tests are used. The method is based on linearly predicting the in-cylinder pressure in a combustion chamber from the output signals of a group of vibration sensors being distributed on the engine wall. The irrelevancy of a sensor in the group is characterized by the closeness to zero of a “coherence gain” explained by this sensor at some frequencies of interest. We formulate multiple hypotheses, define suitable statistics, approximate their sampling distributions by the nonparametric bootstrap method, and construct the generalized sequentially rejective Bonferroni multiple test. The level of accuracy of bootstrap tests is superior to that of asymptotic tests, provided the test statistic used is asymptotically pivotal. To achieve pivoting, we transform the test statistic into a variance stable scale where we use a bootstrap technique to approximate the variance stabilizing transformation. Simulations results as well as results of an experiment performed on a test bed with a four-cylinder engine emphasize the applicability of the method View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Harmonics in multiplicative and additive noise: performance analysis of cyclic estimators

    Publication Year: 1995 , Page(s): 1445 - 1460
    Cited by:  Papers (40)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1084 KB)  

    Multiplicative noise causes smearing of spectral lines and thus hampers frequency estimation relying on conventional spectral analysis. In contrast, cyclic mean and correlation statistics have proved to be useful for harmonic retrieval in the presence of multiplicative and additive noise of arbitrary color and distribution. Performance analysis of cyclic estimators is carried through both for nonzero and zero mean multiplicative noises. Cyclic estimators are shown to be asymptotically equivalent to certain nonlinear least squares estimators, and are also compared with the maximum likelihood ones. Large sample variance expressions of the cyclic estimators are derived and compared with the corresponding Cramer-Rao bounds when the noises are white Gaussian. It is demonstrated that previously well established results on constant amplitude harmonics are special cases of the present analysis. Simulations not only validate the large sample performance analysis, but also provide concrete examples regarding relative statistical efficiency of the cyclic estimators View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Complex hybrid correlators: an extended analysis of their performance

    Publication Year: 1995 , Page(s): 1531 - 1535
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (324 KB)  

    Discrete-time hybrid correlators were devised and their accuracy was provided in a closed form for the simplest cases of instantaneous nonlinearities, in the reference case of complex Gaussian stationary processes. The article extends the performance analysis to finer phase nonlinearities View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A fast block-based nonlinear decoding algorithm for ΣΔ modulators

    Publication Year: 1995 , Page(s): 1360 - 1367
    Cited by:  Papers (7)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (712 KB)  

    Previous work has established that the digital output of a ΣΔ modulator as an A/D converter contains more information about the analog input than is extracted with conventional linear filtering. Under reasonable mathematical assumptions, optimal nonlinear decoding of the digital output can achieve significantly larger signal-to-noise ratios than linear filtering. However, the hitherto proposed decoding algorithms only demonstrate conceptual feasibility and are impractical from a computational point of view. We present a new block-based decoding algorithm that, like previous work, employs projections onto convex sets. The algorithm owes its speed to a change of projection norm, an accelerated convergence scheme, and a decimation-like subsampling; it is on the order of 104-105 times faster than one previously published algorithm for typical parameter values, and about 2-10 times slower than linear decoding. The new algorithm is applicable to all currently popular ΣΔ architectures. View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Conjugate gradient method in adaptive bilinear filtering

    Publication Year: 1995 , Page(s): 1503 - 1508
    Cited by:  Papers (15)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    The application of the conjugate gradient (CG) method for the identification of bilinear systems is investigated. An algorithm based on the CG method is developed for adaptive bilinear digital filtering. In this algorithm, the optimization is done over blocks of input and output data rather than a single pair of data. However, only one iteration and coefficient update is done for every sample of data. This, coupled with the fact that the CG method used does not require a line search makes it very efficient in computation. Simulation results show that this algorithm outperforms the LMS and RLS algorithms in terms of speed of convergence. A preconditioning technique is also applied to further accelerate the convergence in some cases View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Wavelet neural networks for function learning

    Publication Year: 1995 , Page(s): 1485 - 1497
    Cited by:  Papers (160)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (996 KB)  

    A wavelet-based neural network is described. The structure of this network is similar to that of the radial basis function (RBF) network, except that in the present paper the radial basis functions are replaced by orthonormal scaling functions that are not necessarily radial-symmetric. The efficacy of this type of network in function learning and estimation is demonstrated through theoretical analysis and experimental results. In particular, it has been shown that the wavelet network has universal and L2 approximation properties and is a consistent function estimator. Convergence rates associated with these properties are obtained for certain function classes where the rates avoid the “curse of dimensionality”. In the experiments, the wavelet network performed well and compared favorably to the MLP and RBF networks View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Constraints for maximally flat optimum broadband antenna arrays

    Publication Year: 1995 , Page(s): 1334 - 1347
    Cited by:  Papers (7)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1008 KB)  

    Frequency response shaping for the direct form pre-steered broadband (PB) antenna array processor is often achieved by imposing look direction constraints on the weights of the processor. This results in a linearly constrained optimization problem. To ensure a maximally flat spatial response of a specified order in the look direction of the PB processor, additional constraints known as derivative constraints can be further imposed on the weights. In general, derivative constraints corresponding to necessary and sufficient (NS) conditions for a maximally flat spatial power response can result in a quadratic equality constrained optimization problem. We transform the quadratic NS derivative constraints to parameterized linear forms. These parameterized linear forms allow the global optimum of the quadratic equality constrained optimization problem to be obtained easily. They also provide a general framework for deriving new sets of derivative constraints which correspond only to sufficient conditions for a maximally flat spatial power response. These sufficient derivative constraints are useful for real-time processing because of their reduced computational requirements and because they ran deliver performance comparable to the NS derivative constraints View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A recursive frequency-sampling method for designing zero-phase FIR filters by nonuniform samples

    Publication Year: 1995 , Page(s): 1461 - 1467
    Cited by:  Papers (6)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (520 KB)  

    A novel frequency-sampling method for designing zero-phase FIR filters from nonuniform samples is presented. The method is fast, simple, recursive and can be used in the design of 1D or 2D zero-phase FIR filters by imposing some mild constraints on sample locations in the 2D frequency plane. Based on a novel Newton representation of the filter transfer function the proposed method guarantees real results, saves a number of operations and produces accurate solutions even in cases of designing high-order filters or when the interpolation matrix is ill-conditioned. In the progressive case when the next sample appears, the design parameters are evaluated by updating the old ones with correction terms that could be used as indicators for convergence, approximation, or filter reduction. The method can be used in mD filter design, in LU-factorization or in inversion of cosine matrices View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On-line spectral estimation of nonstationary time series based on AR model parameter estimation and order selection with a forgetting factor

    Publication Year: 1995 , Page(s): 1519 - 1522
    Cited by:  Papers (6)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (324 KB)  

    A new method for on-line spectral estimation of nonstationary time series via autoregressive (AR) model construction is proposed. The method consists of on-line parameter estimation based on the recursive least squares ladder estimation algorithm with a forgetting factor and on-line order determination based on AIC with some modifications. The effectiveness of the proposed method is demonstrated by computer simulation study and applying to the actual data of electroencephalogram (EEG) View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Correlation estimators based on simple nonlinear transformations

    Publication Year: 1995 , Page(s): 1438 - 1444
    Cited by:  Papers (5)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    The computational cost of estimating correlations may be reduced by employing sums of simple nonlinear functions of the data. A quadruplex transformation is presented and the performance of the associated estimator is analyzed for real and complex Gaussian processes. With independent observations, the variance of the estimator is approximately 14% higher than that obtained by averaging lag products View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Pipelined adaptive DFE architectures using relaxed look-ahead

    Publication Year: 1995 , Page(s): 1368 - 1385
    Cited by:  Papers (9)  |  Patents (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1196 KB)  

    Fine-grain pipelined adaptive decision-feedback equalizer (ADFE) architectures are developed using the relaxed look-ahead technique. This technique, which is an approximation to the conventional look-ahead computation, maintains functionality of the algorithm rather than the input-output behavior. Thus, it results in substantial hardware savings as compared to either parallel processing or look-ahead techniques. Pipelining of the decision feedback loop and the adaptation loop is achieved by the use of delay relaxation and sum relaxation. Both the conventional and the predictor form of ADFE have been pipelined. Results of the convergence analysis of the proposed algorithms are also provided. The performance of the pipelined algorithms for the equalization of a magnetic recording channel is studied. It is shown that the conventional ADFE results in an SNR loss of about 0.6 dB per unit increase in the speed-up factor. The predictor form of ADFE is much more robust and results in less than 0.1 dB SNR loss per unit increase in the speed-up factor. Speed-ups of up to 8 and 45 have been demonstrated for the conventional and predictor forms of ADFE View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On multiple pattern extraction using singular value decomposition

    Publication Year: 1995 , Page(s): 1536 - 1540
    Cited by:  Papers (22)  |  Patents (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (384 KB)  

    This paper presents a new concept of decomposition of a signal into component periodic waveforms. The singular value decomposition (SVD) is used for the detection of periodicity and separation of the component signals. The signal is configured for the sequential extraction of successively weaker components with different period lengths. The approach enjoys the numerical stability associated with SVD View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Exact interpolation and iterative subdivision schemes

    Publication Year: 1995 , Page(s): 1348 - 1359
    Cited by:  Papers (5)  |  Patents (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (920 KB)  

    We examine the circumstances under which a discrete-time signal can be exactly interpolated given only every Mth sample. After pointing out the connection between designing an M-fold interpolator and the construction of an M-channel perfect reconstruction filter bank, we derive necessary and sufficient conditions on the signal under which exact interpolation is possible. Bandlimited signals are one obvious example, but numerous others exist. We examine these and show how the interpolators may be constructed. A main application is to iterative interpolation schemes, used for the efficient generation of smooth curves. We show that conventional bandlimited interpolators are not suitable in this context. A better criterion is to use interpolators that are exact for polynomial functions. We demonstrate that these interpolators converge when iterated, and show how these may be designed for any polynomial degree N and any interpolation factor M. This makes it possible to design interpolators for iterative schemes to make best use of the resolution available in a given display medium View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Estimating spectral correlations with simple nonlinear transformations

    Publication Year: 1995 , Page(s): 1525 - 1526
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (152 KB)  

    Two computationally efficient cyclic spectral analysis algorithms are compared using a computer simulation. Each algorithm replaces multiplications with sign changes and multiplexing operations. Compared with conventional frequency smoothing estimators, the quadruplex and one bit spectral correlation algorithms suffered losses in output signal to noise ratio of approximately 0.6 dB and 1.9 dB, respectively View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of multicomponent LFM signals by a combined Wigner-Hough transform

    Publication Year: 1995 , Page(s): 1511 - 1515
    Cited by:  Papers (89)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (368 KB)  

    The aim of the paper is the performance evaluation of a method for the analysis of mono- or multicomponent linear-frequency modulation (LFM) signals, based on the Hough transform of the Wigner-Ville distribution of the signals. A closed form expression is found for the signal-to-noise ratio and the parameter estimation accuracy. The overall method, as any nonlinear method, exhibits a threshold effect. Nevertheless, it is shown to be asymptotically efficient and offers a good rejection capability of the cross terms View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Analysis of the performance and sensitivity of eigendecomposition-based detectors

    Publication Year: 1995 , Page(s): 1413 - 1426
    Cited by:  Papers (28)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (888 KB)  

    A new framework is presented for the analysis of the performance of detection methods, such as AIC and MDL, which are based on the eigenvalues of the sample covariance matrix. It is shown that theoretical analysis of the probabilities of overestimation and underestimation can be much more conveniently carried out via a proposed, particularly simple, sequence of statistics. Also, the breakdown of these detection methods in the presence of model nonidealities is explored by theory, simulations, and experimentation with real array data. For example, theoretical arguments are given to demonstrate the high degree of sensitivity of the detectors to unknown deviations of the noise from whiteness View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On necessary and sufficient conditions for perfect reconstruction multidimensional delay chain systems

    Publication Year: 1995 , Page(s): 1515 - 1519
    Cited by:  Papers (1)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (464 KB)  

    We present some necessary and sufficient conditions for a perfect reconstruction (PR) modified multidimensional (MD) delay chain system. With these results, one is able to systematically check if a D-dimensional delay chain system with a D×D sampling matrix M and a D×D delay matrix L has a PR property in a simpler way than before, where the matrix module operations are avoided. Moreover, given a D×D sampling matrix M, in many cases one can determine all possible D×D delay matrices L so that the delay chain systems with the sampling matrix M have a PR property. Several examples are provided. We also present a method to generate D×D sampling and delay matrices M and L such that their corresponding traditional MD delay chain systems are PR View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Intensity estimation from shot-noise data

    Publication Year: 1995 , Page(s): 1527 - 1531
    Cited by:  Papers (3)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (340 KB)  

    The estimation of the intensity function of a Poisson-driven shot-noise process is addressed using a regularization technique, where the data is modeled as a signal term plus a signal-dependent noise term. A new data-based method for selecting a pair of regularization parameters is presented and compared with the minimum unbiased risk method. The detail in the intensity function can be recovered by both methods, but the new method does a better job at suppressing spurious oscillations View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Performance analysis of the MVDR spatial spectrum estimator

    Publication Year: 1995 , Page(s): 1427 - 1437
    Cited by:  Papers (19)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (768 KB)  

    The performance of the minimum variance distortionless response spectrum estimator is analyzed. Finite data effects and the sensitivity of the method to random perturbations in the signal model and in the noise covariance matrix are studied. The snapshots are assumed to be complex independent identically distributed Gaussian vectors. An expression for the exact model asymptotic bias is also derived. Analytical expressions for the variance and the bias of the estimator are derived and compared with simulation results. These expressions are then employed to study the characteristics of the estimator View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A general characterization of lowpass FIR frequency response in the transition band

    Publication Year: 1995 , Page(s): 1508 - 1510
    Cited by:  Papers (2)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (236 KB)  

    A general model is derived for the frequency response of lowpass FIR filters in the transition band which has applications to filter bank aliasing calculations and convergence of the LMS algorithm. It is shown that as a class, these filters exhibit a frequency response whose squared magnitude rolls off approximately as (ω0-ω)q, where q is a positive exponent that is expressed generically in terms of the filter order and the angular stopband edge frequency ω0 View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Steering vector estimation in uncalibrated arrays

    Publication Year: 1995 , Page(s): 1397 - 1412
    Cited by:  Papers (16)
    Save to Project icon | Request Permissions | Click to expandQuick Abstract | PDF file iconPDF (1296 KB)  

    This paper presents an iterative algorithm for estimating the signal steering vectors and associated power levels received by an array of uncalibrated isotropic sensors. The inputs are assumed to consist of narrowband, uncorrelated directional signals in the presence of additive white noise. An iterative algorithm is employed to successively search for estimates that simultaneously satisfy a signal subspace and an orthogonality condition. These conditions are shown to be both necessary and sufficient for identification of the underlying steering vectors in the case when the data covariance matrix is known exactly, i.e. for the case of infinite data observation. The iterative method employs minimum distance criterion (projections) to successively map the solution between three constraint sets until a stable point is determined. Two examples are presented which illustrate the application of the algorithm in direction finding and beamforming View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.

Aims & Scope

IEEE Transactions on Signal Processing covers novel theory, algorithms, performance analyses and applications of techniques for the processing, understanding, learning, retrieval, mining, and extraction of information from signals

Full Aims & Scope

Meet Our Editors

Editor-in-Chief
Sergios Theodoridis
University of Athens