Showing papers on "Linear phase published in 1992"

Multirate Systems and Filter Banks

Abstract: 1. Introduction 2. Review of Discrete-Time Systems 3. Review of Digital Filters 4. Fundamentals of Multirate Systems 5. Maximally Decimated Filter Banks 6. Paraunitary Perfect Reconstruction Filter Banks 7. Linear Phase Perfect Reconstruction QMF Banks 8. Cosine Modulated Filter Banks 9. Finite Word Length Effects 10. Multirate Filter Bank Theory and Related Topics 11. The Wavelet Transform and Relation to Multirate Filter Banks 12. Multidimensional Multirate Systems 13. Review of Discrete-Time Multi-Input Multi-Output LTI Systems 14. Paraunitary and Lossless Systems Appendices Bibliography Index

Applications of simulated annealing for the design of special digital filters

Abstract: The authors describe the salient features of using a simulated annealing (SA) algorithm in the context of designing digital filters with coefficient values expressed as the sum of power of two. A procedure for linear phase digital filter design, using this algorithm, is presented and tested, yielding results as good as those for known optimal methods. The algorithm is then applied to the design of Nyquist filters, optimizing at the same time both frequency response and intersymbol interference, and to the design of cascade form finite-impulse-response (FIR) filters. The drawback of using SA is that the computation time is on the order of 1-2 h for each filter design, on the Sun 3/60. However, this was more than compensated by the versatility of the new algorithm, which can be used to design filters with multiple constraints. >

Design of quadrature mirror filters with linear phase in the frequency domain

Abstract: The design of quadrature mirror filter (QMF) banks whose analysis and synthesis filters have linear phase is considered. Because the design problem in the frequency domain is a highly nonlinear optimization problem, a linearization technique is proposed. An analytical solution formula is obtained, leading to a very efficient procedure. Computer simulations show that the design technique achieves better results in fewer iterations than conventional approaches when starting at the same preset initial guess. Moreover, the technique produces almost the same good results in six iterations if it starts at a better initial guess compared to the preset initial guess. By incorporating the technique with a weighted least squares, (WLS) algorithm, the design of QMF banks whose overall reconstruction error is minimized in the minimax sense over the entire frequency band is facilitated. Computer simulations for illustration and comparison are provided. >

Design of a bipolar 10-MHz programmable continuous-time 0.05 degrees equiripple linear phase filter

Abstract: A bipolar seventh-order 0.05 degrees equiripple linear phase (constant group delay) transconductance-capacitor (g/sub m/-C) low-pass filter with a cutoff frequency (f/sub c/) tunable between 2 and 10 MHz is presented. Programmable equalization up to 9 dB at f/sub c/ is also provided. Total harmonic distortion at 2 V/sub p-p/ is less than 1%, with a dynamic range equal to 49 dB. Nominal power consumption from a single 5-V supply is 135 mW. The circuit also has a low-power mode ( >

A new variable fractional sample delay filter with nonlinear interpolation

Abstract: A finite impulse response (FIR) filter that can synthesize any fractional sample delay by a nonlinear interpolation technique is presented. Analytically closed-form solutions for the tap weights of such an FIR filter and their frequency responses are also presented. >

Reduction of quantum fluctuation and suppression of the Gordon-Haus effect with phase-sensitive linear amplifiers.

Abstract: Compared with the use of phase-insensitive linear quantum amplifiers of the same gain G, the use of phase-sensitive linear amplifiers on phase coherent classical light sources in an amplifier-attenuator chain reduces the total quantum noise power by a factor of 4, the homodyne noise variance by 2, and the photon number variance by 2-8 and suppresses the Kerr effect linear phase fluctuation variance as well as the soliton timing-error variance by 2G(2)

A class of linear-phase regular biorthogonal wavelets

Abstract: A class of biorthogonal systems leading to linear-phase wavelets is presented. A notable feature of this structure is that the wavelets are derived from a filter bank where the lowpass analysis filter is constrained to be a halfband filter. The authors derive finite impulse response (FIR) biorthogonal solutions from a pair of Lagrange halfband filters. They also consider infinite impulse response (IIR) biorthogonal solutions based on a pair of zero-phase halfband filters derived from Butterworth halfband filters. >

Phase-locked loop with automatic phase offset calibration

Abstract: A phase-locked loop having automatic internal phase offset calibration includes a voltage-controlled oscillator circuit for generating a recovered data signal in response to an error signal. A phase detector determines the phase difference between the recovered data signal and a reference data signal. The phase-locked loop further includes a charge pump circuit, coupled to the phase detector, for generating an error signal in response to the detected phase difference. The charge pump circuit includes first and second pump generators for respectively providing first and second sets of pump signals, with the pump generators being interconnected to facilitate generation of the error signal. The phase-locked loop is designed to alternate between operation in phase correction and phase calibration cycles. In each phase correction cycle an error signal is synthesized as described above on the basis of the most recent phase comparison. During each intervening phase calibration cycle a calibration network operates to adjust the second charge pump generator such that the first and second sets of pump signals are precisely balanced when the reference and recovered data signals have a predefined phase relationship. In a preferred embodiment the predefined phase relationship corresponds to that of the reference and recovered data signals being matched in phase. In this way inconsistencies in the operating characteristics of the pump generators are precluded from engendering steady-state phase alignment errors between the reference and recovered waveforms.

Why does linear phase shift cause mode locking

Abstract: The authors examine the character of signals in a simulated wide band oscillator. It is shown how a continuously changing phase shift (a fixed frequency shift) in the feedback path produces a response which resembles mode locking. The phase insertion or frequency shifting prevents the continuity of waves and the dominance of a single frequency component. It stimulates the growth of noise into wave trains which are easily synchronized by nonlinearity or by gain saturation, which of itself is insufficient to lock. Long term periodicities and fixed spectrum components are observed. >

Least p-power error design of FIR filters

Abstract: The authors examine various iterative reweighted least squared error algorithms to obtain an L/sub p/ approximation for designing a linear phase FIR (finite impulse response) filter. These methods consider 1 >

Ladder structures for Multidimensional Linear Phase Perfect Reconstruction filter banks and Wavelets

Abstract: The design of multidimensional filter banks and wavelets have been areas of active research for use in video and image communication systems. At the same time efficient structures for the implementation of such filters are of importance. In 1-D, the well known lattice structure and the recently introduced ladder structure are attractive. However, their extensions to higher dimensions (m-D) have been limited. In this paper we reintroduce the ladder structure, with the purpose of transforming the structure into m-D using the McClellan transform.© (1992) COPYRIGHT SPIE--The International Society for Optical Engineering. Downloading of the abstract is permitted for personal use only.

Design of complex all-pass networks using Remez algorithm

Abstract: A new method for designing complex all-pass digital filters is introduced. Phase error is regarded as the amplitude of complex error between the designed and the desired all-pass function. Then the Remez exchange algorithm is applied to the amplitude of complex error, and it is approximated to be equiripple. Although equiripple solutions to phase approximation problems are not necessarily optimum in the Chebyshev sense, considerable design experience shows that equiripple approximations are often quite satisfactory results. Such cases are considered, and it is shown that the best uniform phase approximation to an arbitrarily specified phase response can be found. In this algorithm, a satisfactory solution is obtained through a few iterations without any initial guess of the solution. Furthermore, as one of the complex all-pass digital filter applications, a large class of real coefficient doubly complementary IIR digital filter pairs is introduced by using a single complex all-pass digital filter, which has approximately linear phase. >

Position, rotation, and scale invariant recognition of images using higher-order spectra

Abstract: A new approach to recognition of images using invariant features based on higher-order spectra is presented. Higher-order spectra are translation invariant because translation produces linear phase shifts which cancel. Scale and amplification invariance are satisfied by the phase of the integral of a higher-order spectrum along a radial line in higher-order frequency space because the contour of integration maps onto itself and both the real and imaginary parts are affected equally by the transformation. Rotation invariance is introduced by deriving invariants from the Radon transform of the image and using the cyclic-shift invariance property of the discrete Fourier transform magnitude. Results on synthetic and actual images show isolated, compact clusters in feature space and high classification accuracies. >

Position, rotation, and scale invariant recognition of images using higher-order spectra

Abstract: A new approach to recognition of images using invariant features based on higher-order spectra is presented. Higher-order spectra are translation invariant because translation produces linear phase shifts which cancel. Scale and amplification invariance are satisfied by the phase of the integral of a higher-order spectrum along a radial line in higher-order frequency space because the contour of integration maps onto itself and both the real and imaginary parts are affected equally by the transformation. Rotation invariance is introduced by deriving invariants from the Radon transform of the image and using the cyclic-shift invariance property of the discrete Fourier transform magnitude. Results on synthetic and actual images show isolated, compact clusters in feature space and high classification accuracies

Performance analysis of a split-path LMS adaptive filter for AR modeling

Abstract: A split-path adaptive filter is proposed for extracting the model parameters of an autoregressive process. The structure is composed of two linear phase filters connected in parallel, one antisymmetric and the other symmetric. The two filters are adapted independently on a sample-by-sample basis using the least-mean-square (LMS) algorithm. The performance of the system in terms of convergence speed and excess mean square error is analyzed in detail, and comparisons with the conventional transversal structure are made. Theoretical analysis and experimental results show that the model can provide a much faster rate of convergence at the expense of only a moderate increase in computation. Two methods for choosing control parameters for the split-path adaptive filter are also suggested to improve further the convergence behavior. >

Phase/frequency comparator for timing recovering circuit

Abstract: The phase/frequency comparator comprises a known phase comparator for comparing the phases of an digital input signal with a clock signal of which the period varies to within one time interval with respect to the period of the input signal. Such a phase comparator produces a first error signal whose sign, and preferably magnitude, vary as a function of the difference between the phases. The phase/frequency comparator is intended to produce a second error signal which replaces the first error signal and whose sign, and preferably magnitude, vary as a function of the difference between the periods. According to a preferred embodiment, a phase shift assessing circuit detects a predetermined phase shift, e.g. substantially equal to 0, between the clock signal and the input signal during a clock period. A sign detecting circuit detects the sign of the difference between the periods for the predetermined phase shift, and a logic circuit derives two signals having pulse widths as a function of said periods thereby deriving the second error signal by differentiation. The phase/frequency comparator is particularly intended for a phase locked loop for recovering the timing of the input signal.

Hybrid composite filter banks for distortion-invariant optical pattern recognition

Abstract: In the past, several different approaches to synthetic discriminant function (SDF) filter design have been proposed, including conventional SDFs, which control the correlation values at the origin; minimum variance SDFs (MVSDFs), which minimize the noise sensitivity of the filters; minimum average correlation energy (MACE) filters, which maximize the peak sharpness; and linear phase coefficient composite (LPCC) filters, which are obtained as the sum of training images weighted by linear phase coefficients. We introduce a new family of SDF filters of which all the above are special cases. Each filter in this family is characterized by two parameters α1 and α2. Various choices of (α1 , α2) lead to the above special filters. For example, α1 = 1 and α2 = 0 leads to MACE LPCC filters, which are hybrid versions of MACE and LPCC filters. This family of filters is evaluated using the minimum probability of error (MPE) criterion and a database of aircraft images. These simulation experiments confirm the superior performance of this filter family. Also, we observe the interesting result that the MPE is at its lowest not for one of the four special filters listed above, but for a combination of them.

Signal controlled phase shifter

Abstract: A clock is phase shifted by an amount controlled by the value of a control signal by establishing at least several discrete delay times to be imposed on the clock. The control signal value controls selection of the imposed discrete delay time. An analog-to-digital converter of a phase locked loop responds to intelligence representing variable phase bits and the selected phase shifted clock to control the signal value. The selected replica is derived by at least several cascaded substantially equal time delay units. In one embodiment, a multiplexer responds to the clock, and the signal value, which is Gray coded, to control connections from one of the delay units to an output terminal. In another embodiment, the number of cascaded delay units interposed between the clock and an output terminal is controlled by the signal value.

Phased array control using phase-locked-loop phase shifters

Abstract: The work investigates the use of phaselocked loops (PLLs) as phase shifters in the control of a phased antenna array. The phase error of a PLL can be altered by adding a DC offset voltage to the output of the phase detector. If a digital phase detector, with its inherently linear response, is used, then the phase error, and hence the relative phase shift, will vary linearly with applied control voltage. Such phase shifters can act on their own at VHF or they can provide an IF reference signal to a mixer to create a linear phase shifter at microwave frequencies. Four prototype phase shifters, fed in series, have been built, driving a linear array of four monopoles, in transmit mode at 1.5 GHz. By use of a single common control voltage, a single main beam can be swept through 180° with an accuracy of better than±2° within the range±70°.

A 27 MHz programmable bipolar 0.05 degrees equiripple linear-phase lowpass filter

Abstract: A continuous-time transconductance-C filter which meets the requirement of the read channel of disk drives based on conventional peak detection pulse qualification is presented. To accommodate constant-density recording with variable data rates tip to 48 Mb/s, the cut-off frequency is tunable between 9 and 27 MHz. The filter is placed inside the AGC (automatic gain control) loop and its differential output signal amplitude is typically held at 1 V/sub PP/. Its primary role is to lower the achievable error rates by bandlimiting the noise originating in the magnetic media and the preamplifier. Its second objective is to equalize the bit stream. i.e. to slim the data pulses, allowing higher densities. To minimize pulse peak shifts in time, an accurate linear phase (or constant group delay) response over the signal bandwidth is essential. The transfer function implemented is seventh-order 0.05 degrees equiripple linear-phase. By allowing negligibly small ripple, the region of constant delay can be extended to about twice the cut-off frequency f/sub c/, compared to 1.5 f/sub c/ for a Bessel filter of equal order, traditionally used in this application. The tradeoff, however, is a nearly twofold increase in sensitivity to transconductance excess phase. >

Reflected boundary conditions for multirate filter banks

Abstract: Several methods for applying perfect reconstruction quadrature mirror filter (PR QMF) banks to finite-length signals are described and compared. Although simple periodization produces a transform that does not increase the size of the transformed signal, it has the disadvantage of introducing a jump discontinuity at the signal's boundary. Various methods of transforming smoother extensions are considered and analyzed in terms of their ability to conserve data storage costs and reproduce the signal in a numerically efficient manner. A complete classification of two-channel schemes based on periodizing symmetric (reflected) signal extensions and using linear phase filters is described, for both even- and odd-length signals. More general techniques based on transforming linear signal extrapolations and truncating the resulting subbands to conserve data size are also presented. An example using reflected boundary extension is discussed. >

An l/sub 1/-approximation based method for synthesizing FIR filters

Abstract: A method is proposed for the synthesis of digital filters having approximately the desired linear phase frequency responses. A mathematical optimization problem is formulated from the synthesis objective, and a theorem from the theory of l/sub 2/-approximation is used to convert the optimization problem such that it can be solved by the linear programming technique. The method is successfully applied to the synthesis of a digital FIR equalizer for a given analog antialiasing filter. >

Time-discrete stereo decoder

Abstract: A receiver having a signal path incorporating a tuner, a demodulator circuit for supplying a stereo multiplex signal comprising a baseband stereo sum signal (L+R), a 19 kHz stereo pilot and a stereo difference signal (L-R) which is double sideband amplitude-modulated on a suppressed 38 kHz subcarrier, a sampler for converting an analog signal into a time-discrete signal and a stereo decoder for time-division multiplex decoding of a time-discrete stereo multiplex signal into time-discrete left and right stereo signals. In order to realise an effective selection of the stereo multiplex signal by means of an easily integrable low-pass filter, the stereo decoder comprises a time-discrete halfband low-pass filter circuit having a finite impulse response, with a transition band which is substantially located in the frequency range of said modulated stereo difference signal and with a half-value transfer which is located at the frequency of said 38 kHz stereo subcarrier, which filter circuit has a substantially linear amplitude transfer characteristic as well as a linear phase transfer characteristic in the transition band, said receiver also including an interpolation circuit which is coupled to an output of the filter circuit for converting time-sequential even and odd sampling values of the output signal of the filter circuit into time-sequential pairs of simultaneously occurring sampling values, and a dematrixing circuit coupled to the interpolation circuit for a linear combination of said pairs of sampling values and a compensation of the crosstalk between the left and right stereo signals caused by the filter circuit.

Symbol timing recovery circuit

Abstract: For recovering symbol timing, the instantaneous phase of a received intermediate frequency M-ary PSK signal is sampled at successive phase sampling points of each symbol interval to produce a series of instantaneous phase values so that the phase sampling points divide the interval into first and second half sections. From the instantaneous phase values a phase angle of each half section is derived and a difference between successive phase angles is then detected for each symbol interval. The phase sampling points are controlled with the difference so that it reduces to zero. Data sampling points are determined from the controlled phase sampling points. In a modification, the instantaneous phase of the PSK signal is sampled at successive phase sampling points which are offset on the opposite sides of the data sampling point to produce a pair of instantaneous phase values, which are then converted to corresponding phase deviations with respect to signal points of the PSK signal. A difference between the phase deviations is detected for controlling the phase sampling points.

Digital transmission performance of standard analog filters

Abstract: This work analyzes the performances of Bessel, Gaussian, Butterworth, and Chebyshev (0.1-dB ripple) filters for synchronous baseband digital transmission. Numerical results showing the effects of system parameters such as signal rate, filter bandwidth, filter order, and pulse duty-cycle are presented. For rectangular inputs, linear phase filters perform better than those with nonlinear phase, from the viewpoint of the Nyquist 1 criterion. In terms of the Nyquist 2 criterion, linear phase filters perform better when the input duty-cycle is unity. For lower duty-cycles, there are symbol rate ranges over which nonlinear phase filters perform better. From the viewpoint of symbol time synchronization, the performances of the two types of filter are essentially the same. Linear phase filters are more robust to sampling clock jitter than nonlinear ones when the symbol rate is higher than the bandwidth. Furthermore, it is shown that for linear phase filters it is possible to increase the transmitting rate while keeping the filter bandwidth constant and with only a minor increase in degradation, by using low duty-cycle inputs. >

Psychophysical derivation of the impulse response through generation of ultrabrief responses: complex inverse estimation without minimum-phase assumptions.

Abstract: A new technique is presented for the psychophysical measurement of the response phase that requires no assumptions of the relation between the phase and amplitude components of the frequency response. The amplitude attenuation is measured by a standard threshold paradigm and is then compensated for by increasing harmonic strength in proportion to the visual attenuation. Then the phase of each frequency component is adjusted to maximize the internal response amplitude. The amplitude and phase values may be converted by inverse Fourier transformation into the estimated visual impulse response in any stable set of stimulus conditions and are insensitive to a wide variety of response nonlinearities. A key aspect of the phase measurement technique is the generation of a stimulus to obtain the minimum time spread of the internal response, creating a briefer response than for any other stimulus.

Efficient design of 2-D linear phase fir filters using singular value decomposition (SVD)

Abstract: An efficient technique for the design of linear phase FIR filters using the SVD method is shown. By properly choosing the order of the 1-D subfilters used in the parallel legs of the filter structure, a sizable reduction in the number of multiplier coefficients is achieved.

Frequency synthesizer having phase control loop with plural branches

Abstract: The invention relates to a circuit arrangement for frequency synthesis with a phase control circuit (1) which comprises a first phase discriminator (3) for receiving a reference signal and an output signal supplied by a first frequency divider (6) with a division ratio k, a low-pass filter (4) coupled to the output of the first phase discriminator (3), and an oscillator (5) coupled to the output of the low-pass filter (4) for generating an output signal which can be supplied to the first frequency divider (6). At least one further branch (2) with a further phase discriminator (8) and a further frequency divider (9) which is to be released and which has a division ratio k is present. The further phase discriminator (8) coupled to the input of the low-pass filter (4) is designed for receiving the reference signal delayed by a delay element (10) and the output signal of the further frequency divider (9) provided for receiving the output signal of the oscillator (5). Each delay element (10) has a different delay time corresponding to a fraction of a period T=1/(n*f ref ) of the reference signal, n being the number of branches (1,2) and f ref being the frequency of the reference signal. A control unit (11) releases the frequency divider (9) approximately after the delay time of the delay element (10) assigned to the relevant frequency divider (9).