By Topic

1991., IEEE International Sympoisum on Circuits and Systems

11-14 June 1991

Filter Results

Displaying Results 1 - 25 of 800
  • 1991 IEEE International Sympoisum on Circuits and Systems (Cat. No.91CH3006-4)

    Publication Year: 1991
    Request permission for commercial reuse | PDF file iconPDF (16 KB)
    Freely Available from IEEE
  • Homotopy continuation methods for neural networks

    Publication Year: 1991, Page(s):2483 - 2486 vol.5
    Cited by:  Papers (5)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (290 KB)

    The application of the homotopy continuation method for finding the global minimum during the training phase of a multilayer neural network is presented. A brief description of the theory of the homotopy continuation methods is given. The backward error propagation algorithm used for training neural networks is summarized. The reformulation of the error minimization problem in the learning algorit... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On the design of a multiprocessor architecture for neural network simulation

    Publication Year: 1991, Page(s):2538 - 2541 vol.5
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (414 KB)

    Criteria are outlined for sizing the memory in a parallel architecture for NN (neural networks) simulation. Although a particular topology and a certain learning algorithm are considered, the design guidelines illustrated can be seen as being generally valid. The performance parameters qualifying an NN simulator are mainly speed and its capacity to represent networks of extended size. It is shown ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Systolic multiple-output multipliers for digital signal processing and computer vision

    Publication Year: 1991, Page(s):2633 - 2636 vol.5
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (321 KB)

    Word-level, bit-serial and bit-level designs for multiplier-less systolic multiple-output multipliers are described. They are characterized in terms of execution speed, hardware requirement and propagation of signals. Their applications to digital signal processing and computer vision are discussed and exemplified respectively by digital filters and a Hough transform.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • On SVD for estimating generalized eigenvalues of singular matrix pencil in noise

    Publication Year: 1991, Page(s):2780 - 2783 vol.5
    Cited by:  Papers (7)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (369 KB)

    Several algorithms for estimating generalized eigenvalues (GEs) of singular matrix pencils perturbed by noise are reviewed. The singular value decomposition (SVD) is explored as the common structure in three basic algorithms: direct matrix pencil algorithm, Pro-ESPRIT, and TLS-ESPRIT. It is shown that several SVD-based steps inherent in those algorithms are equivalent to the first-order approximat... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Gradient algorithms for FIR adaptive filtering - A tutorial

    Publication Year: 1991, Page(s):560 - 563 vol.1
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (432 KB)

    A unified approach for generating fast block- and sequential-gradient LMS FIR (least mean square finite impulse response) adaptive algorithms is presented. The general formulation leads to three classes of adaptive algorithms. It is shown how each class of algorithms is obtained from the previous one by a simple trade-off between adaptation performance and computational complexity. Implementation ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A novel CMOS analog counter for Boltzmann machines

    Publication Year: 1991, Page(s):1311 - 1314 vol.3
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (329 KB)

    A CMOS analog counter for artificial neural networks which is based on a novel principle of charge injection is described. It features an excellent linearity and controllability of its step in a large dynamic range. This counter circuit has a small area and a high robustness against device dispersion.<> View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Temporal models for neural nets for integrated circuit implementations

    Publication Year: 1991
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (111 KB)

    Summary form only given. The possibility of proposing models that have temporal characteristics with a view toward electronic circuit implementation is investigated. Similar to the Hodgkin-Huxley membrane model, a model using Josephson junction circuits can be developed. The advantage of the Josephson junction model lies in its high speed and low power dissipation properties. In addition, such a m... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Transistor-level CMOS gate models for timing analysis and simulation

    Publication Year: 1991, Page(s):2156 - 2159 vol.4
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (244 KB)

    The authors developed a transistor-level CMOS gate model which predicts output transition and delay times with an accuracy which is usually within 10% of SPICE. The model uses constant grounded capacitors to account for storage effects, which makes it applicable in timing analysis and simulation programs which use point-relaxation solution techniques. The authors present the model for a CMOS inver... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Eigenvector method for adaptive interference nulling

    Publication Year: 1991, Page(s):2794 - 2797 vol.5
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (192 KB)

    An new approach is presented to adaptively null out interference which arrives from unknown directions. The constraints forced in both look direction and jammer directions are placed in the new algorithm. From the eigenvectors of a specially constructed matrix, the optimum weight vector is found by a minimum norm criterion. It can create exact nulls in the interference directions while maintaining... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Adaptively trained neural networks and their application to electric load forecasting

    Publication Year: 1991, Page(s):1125 - 1128 vol.2
    Cited by:  Papers (9)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (332 KB)

    A training procedure that adapts the weights of a trained layered perceptron type artificial neural network to training data originating from a slowly varying nonstationary process is proposed. The procedure for adaptive dating assures proper response to previous training data by seeking to minimize a weight sensitivity cost function while, at the same time, minimizing the mean square error normal... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • BiCMOS defect-modelling and fault analysis

    Publication Year: 1991, Page(s):2152 - 2155 vol.4
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (228 KB)

    Common defects are modeled for BiCMOS logic gates. Only 54% of the defect set simulated in the BiCMOS test circuits manifested themselves in terms of a failure in the logical behavior of the gate. This would suggest that test vector sets designed on the premise of testing for logical failures would be less appropriate for BiCMOS than for CMOS gates where 69% of the defect set caused logical failur... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of a linear 'sum-of-product' circuit for Boltzmann machines

    Publication Year: 1991, Page(s):1299 - 1302 vol.3
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (316 KB)

    A novel analog sum-of-product CMOS circuit which is based on a linear voltage-to-current converter is described. A synaptic cell is derived which features low current and small area, high linearity and high operation speed. It is well-suited to Boltzmann machines and any neural network with signed multivalued weights and binary neuron states View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Silicon implementation of an auto-adaptive network for real-time separation of independent signals

    Publication Year: 1991, Page(s):2971 - 2974 vol.5
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (232 KB)

    A nonlinear autoadaptive filter for the separation of independent signal sources is described. The filter takes the form of a recurrent network which uses N simple processing units interconnected by N(N-1) inhibitory synapses. Each synapse determines its weight from a local Hebbian-like learning rule. Criteria are specified which the learning rule must satisfy, and examp... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • New structure of oversampling converters [delta-sigma modulator]

    Publication Year: 1991, Page(s):1649 - 1652 vol.3
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (312 KB)

    A novel structure of delta-sigma modulator (DSM) is proposed where a two-stage second-order loop is modified to a third-order loop by using a differentiator. The structure is simple and is shown to achieve 16-bit performance for Hi-Fi audio-band applications when the sample rate is 3 MHz. For the applications in A/D, the sample rate for this structure is two times of that for the double-loop DSM. ... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Implementation of super high definition image processing on HiPIPE

    Publication Year: 1991, Page(s):348 - 351 vol.1
    Cited by:  Papers (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (448 KB)

    In order to manipulate SHD (super high definition) images, a novel parallel processing unit called HiPIPE (Highly Parallel Image Processing Engine) is developed as a project of NOVI-II. The engine is connected to an SHD image display unit and an image data storage unit. Extremely high computational power is obtained by a multicomputer type parallel processing technique. 128 processing elements are... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Lossy transmission line effects in a thin-film multichip module

    Publication Year: 1991, Page(s):2304 - 2307 vol.4
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (276 KB)

    The author reports results of an investigation of the chip-to-chip signal propagation properties in a thin-film multichip module. The lossy transmission effects on the CMOS interface were analyzed by SPICE simulation and compared with an emitter-coupled logic interface. It is proved that CMOS-based multichip modules are less affected by the line resistance and have less power dissipation. An examp... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • State vector estimation of high dimensional systems based on functional spline operators

    Publication Year: 1991, Page(s):2788 - 2793 vol.5
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (428 KB)

    An estimation procedure based on the functional spline interpolation for dealing with high-dimensional bilinear dynamical models is described. At first, the nonlinear dynamics of a given system is transformed through Carleman bilinearization into a bilinear form. Decoupled bilinear models are then constructed, with input-output mappings expressible in closed form and with dimension determined by t... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Low distortion VLSI compatible self-tuned continuous-time monolithic filters

    Publication Year: 1991, Page(s):1448 - 1451 vol.3
    Cited by:  Papers (11)  |  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (392 KB)

    The design and implementation of three highly linear monolithic continuous-time filters is described. The filters operate from a single 5-V supply and are integrated in a standard doubly-polysilicon CMOS process. Each filter is automatically tuned to a corner frequency of 78 kHz±5% using on-chip circuitry and an external reference clock. All filters are single-ended and have a lowpass respo... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of a binary neural network for security classification in power system operation

    Publication Year: 1991, Page(s):1121 - 1124 vol.2
    Cited by:  Papers (3)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (384 KB)

    The authors present a method for designing a neural network (NN) for potential application in real-time system security analysis. Specifically, the authors formulate the contingency classification problem as a pattern recognition problem and then design a NN to classify the system states (i.e., normal, alert and emergency). A two-layered NN with a fully-connected asynchronous binary model for each... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Design of two-dimensional recursive digital filters based on the optimal decomposition of magnitude specifications

    Publication Year: 1991, Page(s):614 - 617 vol.1
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (260 KB)

    The authors present a novel decomposition technique called the optimal decomposition (OD) for decomposing 2-D magnitude specifications into 1-D ones. By using the OD, 2-D magnitude specifications can be decomposed into 1-D ones which are always nonnegative, and the root mean square (RMS) decomposition error is minimum. As a result, the problem of designing a 2-D digital filter can be reduced to on... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • A module selection algorithm for high-level synthesis

    Publication Year: 1991, Page(s):1777 - 1780 vol.3
    Cited by:  Papers (15)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (348 KB)

    A heuristic approach to the module selection problem in high-level synthesis is presented. In contrast to the common assumption made by most high-level synthesis systems, which consider one available resource for each type of operation, the authors assume that several resources with different delays and areas are available in a functional-block library. The proposed algorithm solves the scheduling... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • New results on fault covering in RRAMs

    Publication Year: 1991, Page(s):2148 - 2151 vol.4
    Cited by:  Papers (2)  |  Patents (2)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (284 KB)

    The authors present new results for the fault covering problem and propose an algorithm that integrates these new results. They present a fast greedy algorithm for constructing repair solutions and establish a simple criterion under which they can guarantee that the algorithm will find a repair solution. This, in turn, leads to a simple test for repairability. The authors present a new test for de... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Biased estimates in the order recursive forward-backward lattice algorithm

    Publication Year: 1991, Page(s):1295 - 1298 vol.2
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (192 KB)

    Deals with the accuracy problem of the estimated reflection coefficients. The algorithm is studied in comparison to the time recursive forward-backward lattice method (FBLT) and the exact order recursive LS lattice algorithm (LSLO). Some mathematical formulations are derived to analyze the bias in estimation. Both theoretical and simulation results show that the estimated reflection coefficients o... View full abstract»

    Full text access may be available. Click article title to sign in or learn about subscription options.
  • Survivor sequence memory management in Viterbi decoders

    Publication Year: 1991, Page(s):2967 - 2970 vol.5
    Cited by:  Papers (19)  |  Patents (5)
    Request permission for commercial reuse | Click to expandAbstract | PDF file iconPDF (336 KB)

    This work extends previous trace-back approaches. A new one-pointer trace-back algorithm for survivor sequence memory management that is particularly well-suited to a VLSI implementation is described. Memory size, latency and implementational complexity of the survivor sequence management are analyzed for both uniprocessor and multiprocessor realizations of Viterbi decoders View full abstract»

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