Showing papers on "Filter design published in 1994"

Unconstrained correlation filters

Abstract: A mathematical analysis of the distortion tolerance in correlation filters is presented. A good measure for distortion performance is shown to be a generalization of the minimum average correlation energy criterion. To optimize the filter's performance, we remove the usual hard constraints on the outputs in the synthetic discriminant function formulation. The resulting filters exhibit superior distortion tolerance while retaining the attractive features of their predecessors such as the minimum average correlation energy filter and the minimum variance synthetic discriminant function filter. The proposed theory also unifies several existing approaches and examines the relationship between different formulations. The proposed filter design algorithm requires only simple statistical parameters and the inversion of diagonal matrices, which makes it attractive from a computational standpoint. Several properties of these filters are discussed with illustrative examples.

Integrated continuous-time filter design /spl minus/ an overview

Abstract: The state of the art of continuous-time filter design is reviewed Several techniques are discussed and compared in terms of performance and implementation feasibility in different fabrication technologies This review does not aim at historical completeness, but rather emphasizes techniques that have proven their worth in commercial applications Brief mention is also made of experimental work which, in the opinion of the author, shows promise for the future >

Near-perfect-reconstruction pseudo-QMF banks

Abstract: A novel approach to the design of M-channel pseudo-quadrature mirror filter (QMF) banks is presented. In this approach, the prototype filter is constrained to be a linear-phase spectral-factor of a 2Mth band filter. As a result, the overall transfer function of the analysis/synthesis system is a delay. Moreover, the aliasing cancellation (AC) constraint is derived such that all the significant aliasing terms are canceled. Consequently, the aliasing level at the output is comparable to the stopband attenuation of the prototype filter. In other words, the only error at the output of the analysis/synthesis system is the aliasing error which is at the level of stopband attenuation. Using this approach, it is possible to design a pseudo-QMF bank where the stopband attenuation of the analysis (and thus synthesis) filters is on the order of /spl minus/100 dB. Moreover, the resulting reconstruction error is also on the order of /spl minus/100 dB. Several examples are included. >

Decision-based median filter using local signal statistics

Abstract: Noise removal is important in many applications. When the noise has impulsive characteristics, linear techniquesdo not perform well, and median filter or its derivatives are often used. Although median-based filters preserve edgesreasonably well, they tend to remove some of the finer details in the image. Switching schemes — where the filter isswitched between two or more filters — have been proposed, but they usually lack a decision rule efficient enough toyield good results on different regions of the image. In this paper we present a strategy to overcome this problem. Adecision rule based on the second order local statistics of the signal (within a window) is used to switch between theidentity filter and a median filter. The results on a test image show an improvement of around 4dB over the medianfilter alone, and 2dB over other techniques.Keywords: Median filter; Image enhancement; Noise removal; Impulsive noise. 1. INTRODUCTION Noise reduction is often necessary as a pre-processing step in situations where a signal is contaminated by noise.In cases where the noise can be adequately modeled as additive Gaussian noise, linear filters are normally efficiciitfor noise-reduction. However, in many cases the noise is impulsive, and in this case linear techniques do not usuallyperform well. The median filter and its derivatives are often the filter of choice for these applications.The median filter is a non-linear filter, and it has the useful property of removing (reducing) impulsive noisewithout (severely) smoothing the edges of the signal. The main drawback of the median filter is that it also modifiesthe points not contaminated by noise, therefore removing the finer details in the signal.In the past 20 years, median filters have been generalized and modified in many ways. A good overview of pastwork on generalizations of median filters can be find in the paper by Gabbouj et al.1 Examples include rank orderfilters, weighted median filters, stack filters, and linear combinations of nonlinear filters. A theory for optimal stackfilters has been developed.2 More recently, filters where the rank selected is based on the pixel rank have been alsoproposed .

Iterative reweighted least-squares design of FIR filters

Abstract: Develops a new iterative reweighted least squares algorithm for the design of optimal L/sub p/ approximation FIR filters. The algorithm combines a variable p technique with a Newton's method to give excellent robust initial convergence and quadratic final convergence. Details of the convergence properties when applied to the L/sub p/ optimization problem are given. The primary purpose of L/sub p/ approximation for filter design is to allow design with different error criteria in pass and stopband and to design constrained L/sub 2/ approximation filters. The new method can also be applied to the complex Chebyshev approximation problem and to the design of 2D FIR filters. >

Least mean p-power error criterion for adaptive FIR filter

Abstract: An adaptive FIR filter based on the least mean p-power error (MPE) criterion is investigated. First, some useful properties of MPE function are studied. Three main results are as follows: 1) MPE function is a convex function of filter coefficients; so it has no local minima. 2) When input process and desired process are both Gaussian processes, then MPE function has the same optimum solution as the conventional Wiener solution for any p. 3) When input process and desired process are non-Gaussian processes, then MPE function may have better optimum solution than Wiener solution. Next, a least mean p-power (LMP) error adaptive algorithm is derived and some application examples are presented. Consequently, when the signal is corrupted by an impulsive noise, the adaptive algorithm with p=1 is preferred. Furthermore, when the signal is corrupted by noise or interference, the adaptive algorithm with proper choice of p may be preferred. >

Orthogonal time-varying filter banks and wavelet packets

Abstract: Considers the construction of orthogonal time-varying filter banks. By examining the time domain description of the two-channel orthogonal filter bank the authors find it possible to construct a set of orthogonal boundary filters, which allows to apply the filter bank to one-sided or finite-length signals, without redundancy or distortion. The method is constructive and complete. There is a whole space of orthogonal boundary solutions, and there is considerable freedom for optimization. This may be used to generate subband tree structures where the tree varies over time, and to change between different filter sets. The authors also show that the iteration of discrete-time time-varying filter banks gives continuous-time bases, just as in the stationary case. This gives rise to wavelet, or wavelet packet, bases for half-line and interval regions. >

Optimum loop filter in hybrid coders

Abstract: Examines the role of the loop/interpolation filter in the motion compensation loop of hybrid coders. Using the Wiener filtering concept and the statistical model developed by Chen and Pang (1992), the optimum filter that minimises the prediction error has been found. The result is expressed in an explicit form in terms of a correlation parameter, /spl rho/ and an inaccuracy parameter, /spl alpha/. It explains many current practices in MPEG and H.261 coders, as well as the leakage predictor, 3-tap versus 8-tap filters and other related issues. The analysis shows that minimum bit rate can only be achieved if the loop filter matches the statistical characteristic of the motion-compensated signal. Furthermore, since the motion noise characteristic could be very different in the horizontal and vertical direction for many sequences, the decision to deploy the optimum filter should be made separately in the two directions. The paper also derives the scalar loop filter in DCT domain. The scalar filter is sub-optimal, but it requires less computational load than the spatial domain filter (64 versus 484 multiplications per 8/spl times/8 block). Experiments show that it performs almost as efficiently as the optimum 3-tap spatial domain filter, thus ascertaining that its performance has not been significantly compromised by the scalar requirement. Experimental simulations on test sequences confirm the theoretical optimum results, and indirectly show that the simple statistical model used in the derivation is adequate. >

Vibration/noise control system

Abstract: A sine wave signal generated in synchronism with a pulse signal determining a frequency of vibrations and noises generated by a vibration/noise source is input to a W filter and a C filter. The C filter selects filter coefficients dependent on the rotational speed of an engine, and generates a transfer characteristic-dependent reference signal R dependent on a transfer characteristic of a vibration/noise-transmitting path, based on the filter coefficients. Alternatively, a divisional signal is prepared by dividing a repetition period of vibrations and noises by a predetermined number, and values of a sine wave generated in synchronism with occurrence of said divisional signal is delivered to a W filter, while the transfer characteristic-dependent reference signal is delivered from the C filter storing data of the transfer characteristic identified in advance to the W filter. Alternatively, a sine wave signal and a delayed sine wave signal delayed by a quarter of a repetition period of the sine wave relative to the sine wave, as well as phase and amplitude-related information of the transfer characteristic of the path are generated and delivered in synchronism with generation of the divisional signal. These sine wave signals and the transfer characteristic-dependent reference signal (phase and amplitude-related information) are used to actively control the vibrations and noises.

High speed simulcast system using adaptive compensation

Abstract: In a simulcast communication system (26), a method and apparatus for compensating differences in propagation time, lack of synchronization in transmitters, and multipath fading to recover data transmitted to a receiving device. In a simulcast communication system (26) that comprises a plurality of transmitters (32a, 32b, 32c), a receiver (36) includes a digital signal processor (DSP) (86) that processes a demodulated received signal to adaptively compensate for changes in the channel through which a multipath signal is propagated from the transmitters to the receiver. In one embodiment, the DSP comprises a decision feedback equalizer (300). An error signal is produced by the equalizer through a comparison of the estimated symbols with symbols most likely transmitted, for use in updating filter coefficients used by the equalizer in processing the received signal. Another embodiment implements a Viterbi algorithm to make decisions of the most likely data symbols in response to estimates of the channel impulse response. Using any one of these embodiments, a linear modulated signal can be decoded to recover the data transmitted, even though the received signal has been degraded by propagation in a multipath fading channel.

Delta-sigma ADC with multi-stage decimation filter and gain compensation filter

Abstract: An analog-to-digital conversion circuit is described which includes a front end sigma-delta modulator circuit, a multi-stage digital decimation filter circuit, and a digital compensation filter circuit. An overrange detect circuit is also provided.

Design of filters to detect a noisy target in nonoverlapping background noise

Abstract: Two types of filter are proposed to detect a noisy target embedded in nonoverlapping background noise by optimization of two proposed criteria that are used in the assessment of filter design and performance. Criterion 1 is defined as the ratio of the square of the expected value of the correlation-peak amplitude to the expected value of the output-signal energy. Criterion 2 is defined as the ratio of the square of the expected value of the correlation-peak amplitude to the average output-signal variance. It is shown that, for the nonoverlapping target and scene noise models, the target window and the scene noise window affect the filter functions significantly. Computer-simulation tests of the generalized optimum filter for various kinds of noisy input image are provided to investigate filter performance in terms of peak-to-output-energy ratio, discrimination against undesired objects, and tolerance to target distortion (for example, target rotation and scaling). We compare the results with those of other filters to verify the performance of the optimum filters.

Frequency-response masking approach for digital filter design: complexity reduction via masking filter factorization

Abstract: It has been reported in several recent publications that the frequency response masking technique is eminently suitable for synthesizing filters with very narrow transition-width The major advantages of the frequency response masking approach are that the resulting filter has a very sparse coefficient vector and that the resulting filter length is only slightly longer than that of the theoretical (Remez) minimum The system of filters produced by the frequency response masking technique consists of a sparse coefficient filter with periodic frequency response and one or more pairs of masking filters Each pair of the masking filters consist of two filters whose frequency responses are similar except at frequencies near the band-edges In this paper, we present three methods for reducing the complexity of the masking filters The success of our technique is due to the fact that each pair of the masking filters can be realized as a cascade of a common subfilter and a pair of equalizers >

Digital receiver for variable data rate communications

Abstract: A digital receiver includes a tuner and a demodulator that obtains a modulated signal carried in a received analog signal. A digital-to-analog converter operates at a preselected fixed sampling rate on the modulated signal to produce a first sequence of digitized samples. The first sequence of digitized samples is processed by a digital rotator to frequency-and phase-correct the first sequence of digitized samples. A controllable digital filter processes the first sequence to produce a filter output including a second sequence of digitized samples at a symbol rate. The second sequence is processed to ascertain a symbol rate of the modulated signal. The controllable filter coefficients are automatically varied to accommodate changes in the symbol rate of the modulated signal, so that the sampling rate of the digital-to-analog converter need not change. The second sequence is processed by post detection automatic gain control to produce a receiver output including a sequence of scaled and leveled digitized samples at the symbol rate.

Method and apparatus for enhancing sharpness of a sequence of images subject to continuous zoom

Abstract: A bandpass enhancement filter (10) has a pass-band that tracks changes in spatial frequency due to the use of a zooming process upon a digital image signal The primary pass-band of the enhancement filter is derived from a combination of a plurality of secondary bandpass filter sections (30a-30d), each having a different frequency response and each responsive to a gain adjustment (34a-34d) A control signal (26) reflecting a particular zoom ratio is used in the adjustment of the gain applied to the filter sections, thereby proportioning the output of each filter section so that the combined output tracks the zooming process

Low delay FIR filter banks: design and evaluation

Abstract: The subject of this paper is the design of low and minimum delay, exact reconstruction analysis-synthesis systems based on filter banks. It presents a time domain approach to the problem of designing FIR filter banks with adjustable reconstruction delays. It is shown that using a time domain formulation for the analysis-synthesis systems, the system delay can be considered to be relatively independent of the length of the analysis and synthesis filters. After a summary of the time domain analysis and design framework, the design of low and minimum delay systems is considered in detail. Several design examples are provided in the paper to demonstrate the performance of the design algorithm. >

Stochastic gradient adaptation under general error criteria

Abstract: Examines a family of adaptive filter algorithms of the form W/sub k+1/=W/sub k/+/spl mu/f(d/sub k/-W/sub k//sup t/X/sub k/)X/sub k/ in which f(/spl middot/) is a memoryless odd-symmetric nonlinearity acting upon the error. Such algorithms are a generalization of the least-mean-square (LMS) adaptive filtering algorithm for even-symmetric error criteria. For this algorithm family, the authors derive general expressions for the mean and mean-square convergence of the filter coefficients For both arbitrary stochastic input data and Gaussian input data. They then provide methods for optimizing the nonlinearity to minimize the algorithm misadjustment for a given convergence rate. Using the calculus of variations, it is shown that the optimum nonlinearity to minimize misadjustment near convergence under slow adaptation conditions is independent of the statistics of the input data and can be expressed as -p'(x)/p(x), where p(x) is the probability density function of the uncorrelated plant noise. For faster adaptation under the white Gaussian input and noise assumptions, the nonlinearity is shown to be x/{1+/spl mu//spl lambda/x/sup 2///spl sigma//sub k//sup 2/}, where /spl lambda/ is the input signal power and /spl sigma//sub k//sup 2/ is the conditional error power. Thus, the optimum stochastic gradient error criterion for Gaussian noise is not mean-square. It is shown that the equations governing the convergence of the nonlinear algorithm are exactly those which describe the behavior of the optimum scalar data nonlinear adaptive algorithm for white Gaussian input. Simulations verify the results for a host of noise interferences and indicate the improvement using non-mean-square error criteria. >

Sidelobe reduction via adaptive FIR filtering in SAR imagery

Abstract: The paper describes a class of adaptive weighting functions that greatly reduce sidelobes, interference, and noise in Fourier transform data By restricting the class of adaptive weighting functions, the adaptively weighted Fourier transform data can be represented as the convolution of the unweighted Fourier transform with a data adaptive FIR filter where one selects the FIR filter coefficients to maximize signal-to-interference ratio This adaptive sidelobe reduction (ASR) procedure is analogous to Capon's (1969) minimum variance method (MVM) of adaptive spectral estimation Unlike MVM, which provides a statistical estimate of the real-valued power spectral density, thereby estimating noise level and improving resolution, ASR provides a single-realization complex-valued estimate of the Fourier transform that suppresses sidelobes and noise Further, the computational complexity of ASR is dramatically lower than that of MVM, which is critical for large multidimensional problems such as synthetic aperture radar (SAR) image formation ASR performance characteristics can be varied through the choice of filter order, l/sub 1/- or l/sub 2/-norm filter vector constraints and a separable or nonseparable multidimensional implementation The author compares simulated point scattering SAR imagery produced by the ASR, MVM, and MUSIC algorithms and illustrates ASR performance on three sets of collected SAR imagery >

Linear periodic systems and multirate filter design

Abstract: We present a systematic procedure for the design of filters intended for multirate systems. This procedure Is motivated by viewing the equiripple design of filters in linear time-invariant systems as a process of obtaining optimum minimax filters for a class of bounded energy input signals. The philosophy of designing optimum minimax filters for classes of input signals is extended to multirate systems, which are not time-invariant. We develop a generalized Fourier analysis appropriate for linear periodic systems and use it to derive new error criteria for multirate filter design. Using such criteria yields optimum minimax multirate filters for the input signal class. The utility of our method is demonstrated by using it to analyze several multirate systems. We give numerical results on the design of a multirate implementation of a narrowband filter and compare our work to previous work on multirate filter design. Our numerical analysis is based upon a new formulation of the design as a semi-infinite linear programming problem. >

Synthetic approach to optimal filtering

Abstract: As opposed to the analytic approach used in the modern theory of optimal filtering, a synthetic approach is presented. The signal/sensor data, which are generated by either computer simulation or actual experiments, are synthesized into a filter by training a recurrent multilayer perceptron (RMLP) with at least one hidden layer of fully or partially interconnected neurons and with or without output feedbacks. The RMLP, after adequate training, is a recursive filter optimal for the given structure, with the lagged feedbacks carrying the optimal conditional statistics at each time point. Above all, it converges to the minimum variance filter as the number of hidden neurons increases. We call such an RMLP a neural filter. Simulation results show that the neural filters with only a few hidden neurons consistently outperform the extended Kalman filter and even the iterated extended Kalman filter for the simple nonlinear signal/sensor systems considered. >

HF filter design and computer simulation

Abstract: A book for engineers who design and build filters of all types, including lumped element, coaxial, helical, dielectric resonator, stripline and microstrip types A thorough review of classic and modern filter design techniques, containing extensive practical design information of passband characteristics, topologies and transformations, component effects and matching An excellent text for the design and construction of microstrip filters

Microstrip filter design using direct EM field simulation

Abstract: For the first time, we present minimax filter design with electromagnetic (EM) simulations driven directly by a gradient-based optimizer. Challenges of efficiency, discretization of geometrical dimensions, and continuity of optimization variables are overcome by a three-stage attack: 1) efficient on-line response interpolation with respect to geometrical dimensions of microstrip structures simulated with fixed grid sizes; 2) smooth and accurate gradient evaluation for use in conjunction with the proposed interpolation; and 3) storing the results of expensive EM simulations in a dynamically updated database. Simulation of a lowpass microstrip filter illustrates the conventional use of EM simulation for design validation. Design optimization of a double folded stub bandstop filter and of a millimeter-wave 26-40 GHz interdigital capacitor bandpass microstrip filter illustrates the new technique. >

Feedback detector and suppressor

Abstract: An A/D converter converts an analog signal to a digital signal. A plurality of cascade-connected notch filters include a first notch filter which is connected to the output of the A/D converter. A D/A converter is connected to the output of the last stage notch filter for converting a digital signal to an analog signal. The output of the last stage notch filter is connected to the input of a fast Fourier transform unit for analyzing the frequency. Analysis results of the fast Fourier transform unit are supplied to a detector. A coefficient having the same center frequency as that of a peak frequency outputted from the detector is selected from a coefficient memory and it is transferred to a second coefficient memory. Thus, the frequencies of the notch filters are set to eliminate howling.

Optical wavelet matched filter.

Abstract: A shift-invariant optical continuous wavelet transform is used for pattern recognition. We propose an optical wavelet matched filter that performs optical wavelet transforms for edge enhancement and the correlation between two wavelet transforms in a single step. This new bandpass matched filter shows improved discrimination capability with respect to the conventional matched filter and improved signal-to-noise ratio with respect to the phase-only matched filter. The wavelet matched filter provides flexibility of an adaptive choice of the scale factors of the wavelets that permit the selection of size and orientation of the smoothing function used in edge enhancement and the optimization of the performance of the filter. Optical experimental results are shown.

Multilayer suspended stripline and coplanar line filters

Abstract: Using metallization patterns on both sides of a suspended stripline substrate or adding an additional dielectric and conductor layer to coplanar line circuits, additional degrees of freedom arise for filter design like an extended range of impedances, tightly coupled line structures or increased end coupling between lines of different metallization layers. In this way, very compact filter circuits with improved performances may be realized as it is shown for different types of filters using this technique, even including active elements to realize strongly frequency selective amplifiers or active filters. >

Input Filter Design for PWM Current-Source Rectifiers

Abstract: Pulse-width modulated (PWM) rectifiers are increasingly used because they allow the elimination of low-order harmonics, and therefore a reduction in input filter components. Filtering requirements for PWM current-source rectifiers are usually satisfied through the use of low-pass LC input filters. This paper offers a systematic and user-friendly approach to choosing the filter components. Design of LC filters involves the positioning of the resonant frequency to meet the harmonic attenuation requirements (THD), and introducing damping at the resonant frequency to avoid amplification of residual harmonics. The problem is further complicated by considerations related to cost, power factor, voltage attenuation, system efficiency, and filter parameter variation. The systematic approach proposed in this paper focuses on PWM rectifiers, but can easily be extended to other classes of converters. Practical design considerations are detailed and design equations derived. Simulated results are presented to validate the design approach

Filter evaluation and selection in wavelet image compression

Abstract: Choice of filter bank in wavelet compression is a critical issue that affects image quality as well as system design. Although regularity is sometimes used in filter evaluation, its success at predicting compression performance is only partial. A more reliable evaluation can be obtained by considering an L-level synthesis/analysis system as a single-input, single-output, linear shift-variant system with a response that varies according to the input location modulo (2/sup L/, 2/sup L/). By characterizing a filter bank according to its impulse response and step response in addition to regularity, the authors obtain reliable and relevant (for image coding) filter evaluation metrics. Using this approach, they have evaluated all possible reasonably short (less than 34 taps in the synthesis/analysis pair) minimum order biorthogonal wavelet filter banks. Of this group of over 4300 candidate filter banks, they have selected and presented the filters best suited to image compression. While some of these filters have been published previously, others are new and have properties that make them attractive in system design. >

Review of models for low-loss filter design and applications

Abstract: The most frequently used models for surface acoustic wave (SAW) devices are the impulse model, the equivalent circuit models, the Coupling-of-Modes model, and the matrix models. While the impulse-model is only a first order model the other models include second order effects, e.g. reflections, dispersion, and charge distribution effects. The influence of diffraction and refraction on the transfer function of a SAW filter can be described by the angular spectrum of straight-crested waves model. A survey of these different models will be given. The simulation of low-loss filters requires flexible analysis tools, which can cope with different geometries and substrates. Operating with a parameter set, which depends only on the substrate crystal and not on the specific geometry of the SAW filter, is advantageous. Due to the high insertion attenuation of conventional transversal filters the requirements on the accuracy of the analysis are focused on S21, whereas for low-loss filters all elements of the S-matrix are important. The comparison of simulations with a P-matrix model, which fulfills the above mentioned prerequisites, and measurements of different types of low-loss filters, e.g. SPUDT, DMS, and transverse-mode coupled resonator filters are presented

Arithmetic-free digital interpolation filter architecture

Abstract: A low precision Finite Impulse Response filter (FIR) is provided for filtering in a digital interpolation operation. The interpolation operation is comprised of two steps, a sampling rate conversion operation for interspersing zeroes between samples in an input sequence and a filtering step of filtering out images that result from this operation. The filtering operation utilizes a FIR filter that utilizes a low precision set of filter coefficients that are selected to tune the frequency response such that the low end frequency response including the pass band, the transition band, and the portion of the stop band immediately after the transition band provides a response equivalent to that commensurate with substantially higher precision FIR filter coefficients. A second, low pass filter section is provided for filtering the high frequency image energy at the output of the FIR filter to provide an overall filter response commensurate to that utilizing substantially higher precision FIR coefficients. The FIR filter coefficients utilized are restricted to the set of {-1, 0, +1} such that an arithmetic-free realization is provided wherein data is stored in a random access memory (68), with the non-zero coefficients for any interpolator output limited to a predetermined number. This predetermined number equals the maximum clock rate divided by the output sampling frequency. For each interpolator output, addresses of the associated data are stored in a ROM (72), which is operable to sequentially generate the addresses for accessing of data from a RAM (68). The sign is then changed, depending upon a sign change bit in the ROM (72), and then accumulated in an output accumulator (82). After all data is accessed from the RAM (68) for a given interpolator output, the accumulator (82) provides this output to the delta-sigma converter.