Showing papers on "Filter design published in 1981"

Electronic filter design handbook

Abstract: Chapter 1: Introduction to Modern Network Theory Chapter 2: Selecting the Response Characteristic Chapter 3: Low-Pass Filter Design Chapter 4: High-Pass Filter Design Chapter 5: Bandpass Filters Chapter 6: Band-Reject Filters Chapter 7: Networks for the Time Domain Chapter 8: Refinements in LC Filter Design and the Use of Resistive Networks Chapter 9: Design and Selection of Inductors for LC Filters Chapter 10: Component Selection for LC and Active Filters Chapter 11: Normalized Filter Design Tables Chapter 12: Introduction to Digital Filters Chapter 13: Finite Impulse-Response Filters Chapter 14: Infinite Impulse-Response Filters Chapter 15: Multirate Digital Filters Chapter 16: Digital Filter Technology Chapter 17: Switched-Capacitor Filters Chapter 18: Introduction to Microwave Filters APPENDIX A: DISCRETE SYSTEMS MATHEMATICS APPENDIX B: SOFTWARE SUMMARY INDEX

Block implementation of adaptive digital filters

Abstract: Block digital filtering involves the calculation of a block or finite set of filter outputs from a block of input values This paper presents a block adaptive filtering procedure in which the filter coefficients are adjusted once per each output block in accordance with a generalized least mean-square (LMS) algorithm Analyses of convergence properties and computational complexity show that the block adaptive filter permits fast implementations while maintaining performance equivalent to that of the widely used LMS adaptive filter

A Separable Median Filter for Image Noise Smoothing

Abstract: This paper investigates some properties of the separable filter resulting from successive applications of a one-dimensional median filter on the rows and columns of an image. Although the output of this separable filter is not identical to the corresponding nonseparable two-dimensional median filter with a square window, its performance in image noise smoothing is close. In particular, its effectiveness in smoothing noise and its behavior with edges are characterized and compared with those of the two-dimensional median filter. It is shown that the separable filter has a much simpler implementation in real-time hardware (at video rates, for example).

Block implementation of adaptive digital filters

Abstract: Block digital filtering involves the calculation of a block or finite set of filter outputs from a block of input values. This paper presents a block adaptive filtering procedure in which the filter coefficients are adjusted once per each output block in accordance with a generalized least mean-square (LMS) algorithm. Analyses of convergence properties and computational complexity show that the block adaptive filter permits fast implementations while maintaining performance equivalent to that of the widely used LMS adaptive filter.

Sampling frequency converter

Abstract: A sampling frequency converter for converting a first signal sampled at a first sampling frequency f1 into a second signal sampled at a second sampling frequency f2 comprising an interpolation device supplied with the first signal, for inserting L-1 zeros (L is an integer) for every sampling time, a filter circuit for attenuating a frequency component over a frequency f/2 (f is a frequency) within an output signal of said interpolation device, where the filter circuit has a series circuit consisting of a finite impulse response digital filter and an infinite impulse response digital filter, and the frequency f is equal to the first sampling frequency f1 when f1 f2, and a decimation device for extracting every M-th (M is an integer) output signal of the filter circuit, to produce said second signal.

Error surfaces of recursive adaptive filters

Abstract: For an adaptive filter with N adjustable coefficients or weights, the "error surface" is a plot, in N + 1 dimensions, of the mean-squared error versus the N coefficient values. If the adaptive filter is nonrecursive, the error surface is a quadratic function of the coefficients. With recursive adaptive filters, the error surface is not quadratic and may even have local minima. In this correspondence we discuss the nature of the recursive error surface and give examples of conditions under which local minima may exist. We conclude with a discussion of the effects of the nonquadratic error surface on gradient-search algorithms for recursive adaptive filters.

Properties of Digital Smoothing Polynomial (DISPO) Filters

Abstract: Digital filters for spectrometric applications are compared with the classical RC filter. Properties discussed include noise reduction, line shift, and conservation of line moments. For Gaussian and Lorentzian lines, signal deformation and change of half-width as a function of time constant and line width are calculated for several filter types. Using accuracy, sensitivity, and scan speed as criteria, it is shown that a fourth-order digital smoothing polynomial (DISPO) filter is better by typically 1 or even 2 orders of magnitude than the RC filter. Since a real time implementation of these filters is possible, they can directly replace RC filters in all spectrometric applications.

The time-sequenced adaptive filter

Abstract: A new form of adaptive filter is proposed which is especially suited for the estimation of a class of nonstationary signals. This new filter, called the time-sequenced adaptive filter, is an extension of the least mean-square error (LMS) adaptive filter. Both the LMS and time-sequenced adaptive filters are digital filters composed of a tapped delay line and adjustable weights, whose impulse response is controlled by an adaptive algorithm. For stationary stochastic inputs the mean-square error, which is the expected value of the squared difference between the filter output and an externally supplied "desired response," is a quadratic function of the weights-a paraboloid with a single fixed minimum point which can be sought by gradient techniques, such as the LMS algorithm. For nonstationary inputs however the minimum point, curvature, and orientation of the error surface could be changing over time. The time-sequenced adaptive filter is applicable to the estimation of that subset of nonstationary signals having a recurring (but not necessarily periodic) statistical character, e.g., recurring pulses in noise. In this case there are a finite number of different paraboloidal error surfaces, also recurring in time. The time-sequenced adaptive filter uses multiple sets of adjustable weights. At each point in time, one and only one set of weights is selected to form the filter output and to be adapted using the LMS algorithm. The index of the set of weights chosen is synchronized with the recurring statistical character of the filter input so that each set of weights is associated with a single error surface. After many adaptations of each set of weights, the minimum point of each error surface is reached resulting in an optimal time-varying filter. For this procedure, some a priori knowledge of the filter input is required to synchronize the selection of the set of weights with the recurring statistics of the filter input. For pulse-type signals, this a priori knowledge could be the location of the pulses in time; for signals with periodic statistics, knowledge of the period is sufficient. Possible applications of the time-sequenced adaptive filter include electrocardiogram enhancement and electric load prediction.

Digital adaptive finite impulse response filter with large number of coefficients

Abstract: A digital adaptive finite impulse response (AFIR) filter having a large number of coefficients is composed of two or more separate filter units. The first filter unit computes response samples h₀ - h n+1 in response to input signal samples to provide a partial estimated response during each sampling period. After the signal samples are fully processed by the first unit, they are transferred to a second filter unit which produces response samples h n+2 - h p+1 in response thereto to provide a second partial response. The sum of the two partial responses is computed to provide the total estimated system response. The two independent filter units thus act simultaneously to provide twice as many coeffi­ cients as prior art AFIR filters in the same amount of time.

Filters comprising reactive components, and a method of determining impedances thereof

Abstract: A filter comprising reactive components which define pass and stop bands has the impedances of the components selected to minimize the sum |Zin-Z2|+|Zout-Z1| in the pass band, where Zin and Zout are respectively the input and output impedances of the filter and Z1 and Z2 are nominal termination impedance at respectively the input and output of the filter. The filter characteristic in the pass band is thereby made substantially independent of the actual termination impedances of the filter in use. The design is exemplified by a 5th order low pass filter for connection in the voice signal path of a telephone subscriber line.

Channel equalizer and method for cancelling ghosts

Abstract: A channel equalizer and ghost cancelling technique are disclosed for removing ghosts from a communications signal which includes a training signal and data. The training signal and its ghosts are processed as a finite length sequence of numbers. The equalizer includes a chain of cascaded filters, the first of which receives the training signal and its ghosts for assuming a filter condition which reduces the finite sequence by two points at the filter's output. Each successive filter in the chain receives the output of an immediately preceding filter and assumes a filter condition in which an additional two points of the finite sequence are eliminated or forced to zero. The filter conditions assumed in response to the training signal and its ghosts are maintained while the data is applied to the filter chain so that ghosts of the type experienced by the training signal are removed from the data.

New digital adaptive-filter implementation using distributed-arithmetic techniques

Abstract: In the paper a new technique for the implementation of digital adaptive filters is presented, based on the use of a so-called distributed-arithmetic filter architecture which uses no multipliers in its realisation of the filtering function. Two adaptive algorithms suitable for this type of filter structure are developed and computer simulations show the viability of both approaches. A simplified-hardware prototype module has also been constructed which realises an 8-point transversal adaptive filter. Results from this prototype are presented to demonstrate the basic feasibility of the distributed-arithmetic adaptive-filter structure.

A new class of very low sensitivity and low roundoff noise recursive digital filter structures

Abstract: A class of new structures is proposed, for the realization of the recursive digital filter with poles near the unit circle, by introducing the particular type of delay replacement scheme in the existing digital filter structure and then, inserting, delays at the. proper places to avoid the delay-free loops. By applying the proposed method to the second-order direct form structures, new structures are found with very low magnitude transfer function sensitivities with respect to multiplier coefficients and low roundoff noises. Some numerical results are also presented.

A note on the use of windows for two-dimensional FIR filter design

Abstract: Using a one-dimensional window as a prototype, a two-dimensional window may be formulated having either a square region of support or a circular one. In this paper we compare the effects of using each of these window formulations for 2-D FIR filter design and present formulas for estimating filter order in terms of design specifications, using a Kaiser window as a prototype.

Adaptive design of digital filters

Abstract: In this paper, we present a novel technique for the design of FIR and IIR digital filters. The design approach begins with the specification of a discrete set of arbitrary magnitude and phase characteristics which describe a desired filter response. These frequency domain characteristics are used to create an ideal "pseudo-filter" whose impulse response is unknown and possibly non-causal, but whose input/output characteristics can be determined for a finite sum of sinusoids. Time-domain techniques common to adaptive system identification are then used to identify a realizable FIR or IIR digital filter which best matches the pseudo-filter. The advantages of this method include the ability to specify response at arbitrarily-spaced frequencies, to use arbitrary cost weighting, and to apply (possibly non-linear) constraints to the range of the filter coefficients.

Adaptive filtering of canine electrogastrographic signals. Part 1: system design.

Abstract: The study of the relation between gastric myo-electrical activities recorded from serosal and cutaneous electrodes is hindered by the poor quality of the cutaneous signal. This hindrance could be minimised by suitable filtering of the signal. Since it is not yet clear which aspects of the cutaneous signal constitute valuable information, the filter process should not affect phase, amplitude, frequency and waveform of the gastric component, while noise components should be suppressed strongly. The system design of a modified adaptive filter that meets these requirements is described. The filter was implemented on a digital Nova 2 minicomputer. the filter performance is described and tested.

Electronically tunable band reject filter

Abstract: An electronically tunable band reject filter rejects only frequencies of a signal that are in the reject band around its center frequency The band reject filter includes an input directional coupler for sampling of the signal that is to be filtered and provides a first and second signal thereby The first signal is filtered by a bandpass filter that passes only frequencies of the signal that are around the center frequency of the bandpass filter An output directional coupler combines the second signal with the filtered signal and obtains on a difference port the difference between the second signal and the filtered signal The frequencies of the signal that are centered around the center frequency of the bandpass filter are thereby rejected

A simplified adaptive recursive filter design

Abstract: The adaptive recursive filter design first proposed by White [1] is re-examined, and a modified filter configuration is proposed which drastically simplifies the gradient generating mechanism. The performance behavior of the simplified filter design is analyzed.

Image gradient detectors operating in a partitioned low-pass channel

Abstract: An electronic image processing system, for image enhancement and noise suppression, from signals representing an array of picture elements, or pels. The system is of the kind providing a feature-enhanced output by the addition of outputs of a high-pass filter acting as image-feature detector and a complementary low-pass filter. The low-pass filter also acts as an image-feature detector and includes a prefilter (130 and FIG. 22) and a sub-sampling filter (32) based on a set of weighting patterns in the form of sparse matrices (FIG. 23). The sub-sampling filter in a bandpass channel of the low-pass filter (206 and FIG. 28) comprises a pair of gradient detectors (210, 220, 230, 240 and FIGS. 32, 33, 34, 35) arranged back to back.

A universal, digital sampling frequency converter for digital audio

Abstract: Digital Audio requires high-quality signal conversion between a variety of sampling frequencies, often in non-trivial integer ratios. In such applications, conventional methods based on analog processing or classical FIR filter rate-changing are not adequate. A novel approach to digital sampling frequency conversion based on a single, multistage filter and adapted to conversion between arbitrary, a priori unknown sampling frequency ratios is presented. Its design, implementation and control are discussed in some detail.

Sampling filter for reducing aliasing distortion

Abstract: Sampling filter which employs fractional period displacement of samples to force additional nulls in the filter characteristic response. This sampling method and cascading antialiasing filters (501, 502, 503) provides a filter structure which minimizes the complexity of the circuit yet which provides great flexibility to implement a desired filter characteristic.

Frequency-domain considerations of LSV digital filters

Abstract: The present paper develops a framework for the analysis and synthesis of linear shift-variant (LSV) digital filters in the frequency domain. First, LSV digital filters are theoretically modeled by the successive use of linear shift-invariant (LSI) filters. On the basis of the model, we present an interpretation of shift-variant spectral modification or filtering. Further, shift-variant digital filtering is discussed in relation to the notions of the short-time spectrum and the generalized frequency function. In addition, we propose an efficient implementation procedure which reduces the number of filter coefficients and the amount of computation. The effectiveness of LSV digital filters in processing time-varying signals is demonstrated by experimental verification.

Linear phase filter with self-equalized group delay

Abstract: In modern communication systems, it has become important to provide filters, and in particular bandpass filters, which can provide substantially uniform group delay across the bandwidth of the filter while still achieving good amplitude response. In this regard, it is particularly desirable to substantially reduce the large variations in the phase characteristics found in conventional bandpass filters at the nominal band edge of the filter. To accomplish this, a filter is provided having at least one pair of lattice arms coupled in parallel to one another between the input and output of the filter. Each of the lattice arms includes a plurality of resonant LC resonators, each of the resonators having a different resonant frequency than the center frequency of the filter. In particular, within the bandwidth of the filter, the exponential damping coefficients for the resonators in each arm are set to decay at the same rate. This desired decay can be accomplished by exponential sizing of the components.

A programmable transversal filter for voice-frequency applications

Abstract: A fully integrated, programmable transversal filter optimized for low-noise, low-power, voice-frequency applications is described. The filter, fabricated with a standard double-poly NMOS process, achieves convolution of an analog input signal with digital tap weightings using a structure with sample-and-hold gates for analog storage and a multiplexed MDAC for multiplication. The design of the filter eliminates fixed pattern noise usually associated with such structures and enables a dynamic range in excess of 70 dB (LPF, f/SUB o//f/SUB s/=0.08) to be achieved at an 8 kHz sampling rate with a power dissipation of less than 80 mW. This area efficient device forms the basis for a range of possible voice-band signal processing functions.

Symmetry exploitation in the design and implementation of recursive 2-D rectangularly sampled digital filters

Abstract: It is well known that exploiting the symmetries existing in the frequency response of 2-D filters results in a reduction in design and implementation complexities. In this paper we explore how to exploit a general class of symmetries in the design and implementation of recursive two-dimensional (2-D) filters. Three types of filters are being studied, namely, causal, factorizable noncausal, and unfactorizable noncausal. The capability of each of these filters to exhibit symmetries is discussed. A method is then proposed to obtain the required form of the filter transfer function for exhibiting the different types of symmetries. Examples are solved both to illustrate the proposed method and to compare the performance of these filter types for the same implementation complexity.

Frequency domain analysis of wraparound error in fast convolution algorithms

Abstract: Fast algorithms exist for computing cyclic convolutions. To obtain the linear convolution required for an FIR filter, the data records must be overlapped by at least L - 1 points, where L is the length of the filter impulse response. If the overlap is too small, wraparound error occurs. This error transforms a linear time-invariant filter into a periodic time-varying filter, whose output is periodically nonstationary for a wide-sense stationary input. The first part of this paper contains a review of the frequency domain theory of periodic filters and processes, in the second part of the paper the theory is applied to the specific periodic filter that results from wraparound error in fast convolution algorithms.

On the stability of digital filters designed using the concept of generalized-immittance convertor

Abstract: A method for designing wave digital filters using the concept of generalized-immittance convertor has been reported recently by Antoniou and Rezk. In this paper is derived a stability criterion for this new type of filter. It is shown that the requirement for ensuring the absence of limit cycles in these filters is the same as the one proposed by Fettweis and Meerkcotter for the conventional wave digital filters.

Error feedback in normal realization of recursive digital filters

Abstract: Normal realizations of narrow-band low-pass filters are shown to be suitable for the use of error feedback to reduce roundoff noise. The use of a simple error feedback structure at each summing node, requiring only a single storage register and no multipliers, reduces the high roundoff noise that is normally associated with narrow-band low-pass filters.