Showing papers on "Noise (signal processing) published in 1997"

Stochastic Approximation Algorithms and Applications

Abstract: Applications and issues application to learning, state dependent noise and queueing applications to signal processing and adaptive control mathematical background convergence with probability one - Martingale difference noise convergence with probability one - correlated noise weak convergence - introduction weak convergence methods for general algorithms applications - proofs of convergence rate of convergence averaging of the iterates distributed/decentralized and asynchronous algorithms.

Fundamentals of nonlinear digital filtering

Abstract: Nonlinear Signal Processing Signal Processing Model Signal and Noise Models Fundamental Problems in Noise Removal Algorithms Statistical Preliminaries Random Variables and Distributions Signal and Noise Models Estimation Some Useful Distributions 1001 Solutions Trimmed Mean Filters Other Trimmed Mean Filters L-Filters C-Filters (Ll-Filters) Weighted Median Filters Ranked-Order and Weighted Order Statistic Filters Multistage Median Filters Median Hybrid Filters Edge-Enhancing Selective Filters Rank Selection Filters M-Filters R-Filters Weighted Majority with Minimum Range Filters Nonlinear Mean Filters Stack Filters Generalizations of Stack Filters Morphological Filters Soft Morphological Filters Polynomial Filters Data-Dependent Filters Decision-Based Filters Iterative, Cascaded, and Recursive Filters Some Numerical Measures of Nonlinear Filters Discussion Statistical Analysis and Optimization of Nonlinear Filters Methods Based on Order Statistics Stack Filters Multistage and Hybrid Filters Discussion Exercises Bibliography Index

Sensorless field orientation control of an induction machine by high frequency signal injection

Abstract: This paper describes a new scheme to find the rotor flux angle from stator voltages and currents by injecting a high-frequency signal. The signal is not a rotating one but a fluctuating one at a reference frame rotating synchronously to the fundamental stator frequency. When the estimated rotor flux angle coincides with the actual angle, the proposed method makes virtually no ripple torque, no vibration and less audible noise caused by the injected signal. The difference of impedances between the flux axis and the quadrature axis at high-frequency signal injection on the rotor flux angle is explained by the equivalent circuit equation of the induction machine. The difference is verified by experiments on test motors under various testing conditions. A sensorless field orientation algorithm is proposed and experimental results clarify the satisfactory operation of the algorithm with one hundred and fifty percent load torque at zero stator frequency.

Speech intelligibility as a function of the number of channels of stimulation for signal processors using sine-wave and noise-band outputs

Abstract: Vowels, consonants, and sentences were processed through software emulations of cochlear-implant signal processors with 2-9 output channels. The signals were then presented, as either the sum of sine waves at the center of the channels or as the sum of noise bands the width of the channels, to normal-hearing listeners for identification. The results indicate, as previous investigations have suggested, that high levels of speech understanding can be obtained using signal processors with a small number of channels. The number of channels needed for high levels of performance varied with the nature of the test material. For the most difficult material--vowels produced by men, women, and girls--no statistically significant differences in performance were observed when the number of channels was increased beyond 8. For the least difficult material--sentences--no statistically significant differences in performance were observed when the number of channels was increased beyond 5. The nature of the output signal, noise bands or sine waves, made only a small difference in performance. The mechanism mediating the high levels of speech recognition achieved with only few channels of stimulation may be the same one that mediates the recognition of signals produced by speakers with a high fundamental frequency, i.e., the levels of adjacent channels are used to determine the frequency of the input signal. The results of an experiment in which frequency information was altered but temporal information was not altered indicates that vowel recognition is based on information in the frequency domain even when the number of channels of stimulation is small.

On errors of digital particle image velocimetry

Abstract: The goal of the present study is to quantify and reduce, when possible, errors in two-dimensional digital particle image velocimetry (DPIV). Two major errors, namely the mean bias and root-mean-square (RMS) errors, have been studied. One fundamental source of these errors arises from the implementation of cross correlation (CC). Other major sources of these errors arise from the peak-finding scheme, which locates the correlation peak with a sub-pixel accuracy, and noise within the particle images. Two processing techniques are used to extract the particle displacements. First, a CC method utilizing the FFT algorithm for fast processing is implemented. Second, a particle image pattern matching (PIPM) technique, usually requiring a direct computation and therefore more time consuming, is used. Using DPIV on simulated images, both the mean-bias and RMS errors have been found to be of the order of 0.1 pixels for CC. The errors of PIPM are about an order of magnitude less than those of CC. In the present paper the authors introduce a peak-normalization method which reduces the error level of CC to that of PIPM without adding much computational effort. A peak-compensation technique is also introduced to make the mean-bias error negligible in comparison with the RMS error. Noise in an image suppresses the mean-bias error but, on the other hand, significantly amplifies the RMS error. A digital video signal usually has a lower noise level than that of an analogue one and therefore provides a smaller error in DPIV.

Thresholding for Change Detection.

Abstract: Image differencing is used for many applications involving change detection. Although it is usually followed by a thresholding operation to isolate regions of change there are few methods available in the literature specific to (and appropriate for) change detection. We describe four different methods for selecting thresholds that work on very different principles. Either the noise or the signal is modeled, and the model covers either the spatial or intensity distribution characteristics. The methods are as follows: (1) a Normal model is used for the noise intensity distribution, (2) signal intensities are tested by making local intensity distribution comparisons in the two image frames (i.e., the difference map is not used), (3) the spatial properties of the noise are modeled by a Poisson distribution, and (4) the spatial properties of the signal are modeled as a stable number of regions (or stable Euler number).

A real-time microprocessor QRS detector system with a 1-ms timing accuracy for the measurement of ambulatory HRV

Abstract: The design, test methods, and results of an ambulatory QRS detector are presented. The device is intended for the accurate measurement of heart rate variability (HRV) and reliable QRS detection in both ambulatory and clinical use. The aim of the design work was to achieve high QRS detection performance in terms of timing accuracy and reliability, without compromising the size and power consumption of the device. The complete monitor system consists of a host computer and the detector unit. The detector device is constructed of a commonly available digital signal processing (DSP) microprocessor and other components. The QRS detection algorithm uses optimized prefiltering in conjunction with a matched filter and dual edge threshold detection. The purpose of the prefiltering is to attenuate various noise components in order to achieve improved detection reliability. The matched filter further improves signal-to-noise ratio (SNR) and symmetries the QRS complex for the threshold detection, which is essential in order to achieve the desired performance. The decision for detection is made in real-time and no search-back method is employed. The host computer is used to configure the detector unit, which includes the setting of the matched filter impulse response, and in the retrieval and postprocessing of the measurement results. The QRS detection timing accuracy and detection reliability of the detector system was tested with an artificially generated electrocardiogram (EGG) signal corrupted with various noise types and a timing standard deviation of less than 1 ms was achieved with most noise types and levels similar to those encountered in real measurements. A QRS detection error rate (ER) of 0.1 and 2.2% was achieved with records 103 and 105 from the MIT-BIH Arrhythmia database, respectively.

Externally linear, time-invariant systems and their application to companding signal processors

Abstract: This paper considers externally linear, time-invariant systems which can be internally nonlinear and/or time-varying. Several examples of the synthesis of such systems are presented, and it is argued that they offer advantages in comparison to internally linear, time-invariant systems. In particular, their use in signal processing is considered, and it is shown that they can be designed to reduce the undesirable effects of overloading and noise corruption through the use of companding. It is shown that a variety of previously proposed schemes for this purpose are all related. Several practical problems are discussed, as are performance criteria and types of measurements needed for the evaluation of the systems discussed.

Ultrasensitive laser measurements without tears

Abstract: Several easily implemented devices for doing ultrasensitive optical measurements with noisy lasers are presented. They are all-electronic noise cancellation circuits that largely eliminate excess laser intensity noise as a source of measurement error and are widely applicable. Shot-noise-limited optical measurements can now easily be made at baseband with noisy lasers. These circuits are especially useful in situations where strong intermodulation effects exist, such as current-tuned diode laser spectroscopy. These inexpensive devices ~parts cost ’$10! can be optimized for particular applications such as wideband or differential measurements. Although they cannot eliminate phase noise effects, they can reduce amplitude noise by 55‐70 dB or more, even in unattended operation, and usually achieve the shot-noise limit. With 1-Hz signal-to-noise ratios of 150‐160 dB, they allow performance equal or superior to a complex heterodyne system in many cases, while using much simpler dual-beam or homodyne approaches. Although these devices are related to earlier differential and ratiometric techniques, their noise cancellation performance is much better. They work well at modulation frequencies from dc to several megahertz and should be extensible to ’100 MHz. The circuits work by subtracting photocurrents directly, with feedback applied outside the signal path to continuously adjust the subtraction for perfect balance; thus the excess noise and spurious modulation ideally cancel at all frequencies, leaving only the shot noise. The noise cancellation bandwidth is independent of the feedback bandwidth; it depends only on the speeds of the photodiodes and of the bipolar junction transistors used. Two noise-canceled outputs are available; one is a high-pass filtered voltage proportional to the signal photocurrent and the other is a low-pass filtered voltage related to the log ratio of the signal and comparison photocurrents. For reasonable current densities, the noise floors of the outputs depend only on the shot noise of the signal beam. Four variations on the basic circuit are presented: low noise floor, high cancellation, differential high power, and ratio-only. Emphasis is placed on the detailed operation and design considerations, especially performance extension by compensation of the nonideal character of system components. Experience has shown that some applications advice is required by most users, so that is provided as well. © 1997 Optical Society of America

Method and system for simultaneously broadcasting and receiving digital and analog signals

Abstract: A system and method for transmitting digital information through a medium such as atmospheric free-space includes a transmitter which generates a signal based on a basis set of mutually orthogonal, spectrally-shaped, sequences of substantially equal length and having predetermined autocorrelation values. The sequences may resemble noise in at least some of their characteristics. The orthogonality or cross-correlation characteristics, the autocorrelation characteristics and the resemblance to noise are due to features derived from sequences of pseudo-random numbers which themselves resemble noise in at least some of their characteristics. The waveform set based on the sequences is modulated digitally. The modulated set may be summed together along with a wideband reference signal of reduced amplitude and optionally an FM analog signal to form a composite signal which is broadcast typically through free space to at least one receiver. The receiver separates the analog FM signal from the digital signal and thereafter demodulates the digital data-carrying waveforms and outputs a stream of digital data. It has been determined to be resistant to multipath degradation.

Maximum likelihood for blind separation and deconvolution of noisy signals using mixture models

Abstract: An approximate maximum likelihood method for blind source separation and deconvolution of noisy signal is proposed. This technique relies upon a data augmentation scheme, where the (unobserved) input are viewed as the missing data. In the technique described, the input signal distribution is modeled by a mixture of Gaussian distributions, enabling the use of explicit formula for computing the posterior density and conditional expectation and thus avoiding Monte-Carlo integrations. Because this technique is able to capture some salient features of the input signal distribution, it performs generally much better than third-order or fourth-order cumulant based techniques.

Generalized multichannel image-filtering structures

Abstract: Recent works in multispectral image processing advocate the employment of vector approaches for this class of signals. Vector processing operators that involve the minimization of a suitable error criterion have been proposed and shown appropriate for this task. In this framework, two main classes of vector processing filters have been reported in the literature. Astola et al. (1990) introduce the well-known class of vector median filters (VMF), which are derived as maximum likelihood (ML) estimates from exponential distributions. Trahanias et al. (see ibid., vol.2, no.4, p.528-34, 1993 and vol.5, no.6, p.868-80, 1996) study the processing of color image data using directional information, considering the class of vector directional filters (VDF). We introduce a new filter structure, the directional-distance filters (DDF), which combine both VDF and VMF in a novel way. We show that DDF are robust signal estimators under various noise distributions, they have the property of chromaticity preservation and, finally, compare favorably to other multichannel image processing filters.

Frequency estimation by demodulation of two complex signals

Abstract: This paper presents a method for frequency estimation in a power system by demodulation of two complex signals. In power system analysis, the /spl alpha//spl beta/-transform is used to convert three phase quantities to a complex quantity where the real part is the in-phase component and the imaginary part is the quadrature component. This complex signal is demodulated with a known complex phasor rotating in opposite direction to the input. The advantage of this method is that the demodulation does not introduce a double frequency component. For signals with high signal to noise ratio, the filtering demand for the double frequency component can often limit the speed of the frequency estimator. Hence, the method can improve fast frequency estimation of signals with good noise properties. The method loses its benefits for noisy signals, where the filter design is governed by the demand to filter harmonics and white noise. The method has been previously published, but not explored to its potential. The paper presents four examples to illustrate the strengths and weaknesses of the method.

Electrocardiogram noise reduction using multi-dimensional filtering

Abstract: Noise is reduced from a received ECG signal representative of activity of the heart of a patient. A collection of beats from the ECG signal is selected and transformed into a multi-dimensional representation. A multi-dimensional filter function is applied to the multi-dimensional representation to enhance a signal-to-noise ratio of the collection of beats.

Method and apparatus for accurately measuring the saturated oxygen in arterial blood by substantially eliminating noise from the measurement signal

Abstract: A pulse oximeter includes a light-emitting device for repeating red light emission, infrared light emission, and no light emission with respect to an object to be measured every data sampling cycle in a measurement of a saturated oxygen in arterial blood, a light-receiving device for outputting a light-receiving signal obtained by receiving transmitted light or reflected light from the object to be measured; a noise level detecting device for detecting from the light-receiving signal a noise signal level at the time of no light emission in the light-receiving device, and a light-receiving signal generating device for obtaining a light-receiving signal of a level corresponding only to the red light emission and the infrared light emission by subtracting the noise signal level from each of light-receiving signal levels of the red light emission and the infrared light emission.

Maximum likelihood array processing in spatially correlated noise fields using parameterized signals

Abstract: This paper deals with the problem of estimating signal parameters using an array of sensors. This problem is of interest in a variety of applications, such as radar and sonar source localization. A vast number of estimation techniques have been proposed in the literature during the past two decades. Most of these can deliver consistent estimates only if the covariance matrix of the background noise is known. In many applications, the aforementioned assumption is unrealistic. Recently, a number of contributions have addressed the problem of signal parameter estimation in unknown noise environments based on various assumptions on the noise. Herein, a different approach is taken. We assume instead that the signals are partially known. The received signals are modeled as linear combinations of certain known basis functions. The exact maximum likelihood (ML) estimator for the problem at hand is derived, as well as computationally more attractive approximation. The Cramer-Rao lower bound (CRB) on the estimation error variance is also derived and found to coincide with the CRB, assuming an arbitrary deterministic model and known noise covariance.

Methods for generating comfort noise during discontinuous transmission

Abstract: An improved method for generating comfort noise (CN) in a mobile terminal operating in a discontinuous transmission (DTX) mode. In one embodiment the invention provides an improved method for comfort noise generation, in which a random excitation is modified by a spectral control filter so that the frequency content of comfort noise and background noise become similar. In another embodiment the transmitter identifies speech coding parameters that are not representative of the actual background noise, and replaces the identified parameters with parameters having a median value. In this manner the non-representative parameters do not skew the result of an averaging operation.

Method and apparatus for cancellation of unwanted signals in MWD acoustic tools

Abstract: An apparatus and method are disclosed for eliminating a noise signal from at least one source during an acoustic measurement of a subsurface geological formation or borehole. The apparatus includes a longitudinal body for positioning in the borehole and a transmitter supported by the body for transmitting acoustic signals into the formation and borehole. A sensor, substantially isolated within the body, is used to detect one or more noise signals and a receiver is carried by the body for receiving acoustic signals traversing the formation and borehole, and for receiving one or more noise signals. A processor is connected to the sensor and receiver for processing the acoustic signals and noise signals coupled from the receiver and the noise signal coupled from the sensor into a preferred formation or borehole signal by determining the noise signal received at the receiver using the noise signal received in the sensor and a propagation factor for the noise signal between the sensor and receiver. The determined noise signal is used to identify and eliminate the noise signals from the acoustic signals traversing the formation and borehole.

Robust Nonlinear H Synchronization of Chaotic Lur'e Systems

Abstract: In this paper, we propose a method of robust nonlinear H master-slave synchronization for chaotic Lur'e systems with applications to secure communication. The scheme makes use of vector field modulation and either full static state or linear dynamic output error feedback control. The master-slave systems are assumed to be nonidentical and channel noise is taken into account. Binary valued continuous time message signals are recovered by minimizing the -gain from the exogenous input to the tracking error for the standard plant representation of the scheme. The exogenous input takes into account the message signal, channel noise and parameter mismatch. Matrix inequality conditions for dissipativity with finite -gain of the standard plant form are derived based on a quadratic storage function. The controllers are designed by solving a nonlinear optimization problem which takes into account both channel noise and parameter mismatch. The method is illustrated on Chua's circuit.

Cisoid parameter estimation in the colored noise case: asymptotic Cramer-Rao bound, maximum likelihood, and nonlinear least-squares

Abstract: The problem of estimating the parameters of complex-valued sinusoidal signals (cisoids, for short) from data corrupted by colored noise occurs in many signal processing applications. We present a simple formula for the asymptotic (large-sample) Cramer-Rao bound (CRB) matrix associated with this problem. The maximum likelihood method (MLM), which estimates both the signal and noise parameters, attains the performance corresponding to the asymptotic CRB, as the sample length increases. More interestingly, we show that a computationally much simpler nonlinear least-squares method (NLSM), which estimates the signal parameters only, achieves the same performance in large samples.

Matching pursuits with a wave-based dictionary

Abstract: The method of matching pursuits utilizes a nonlinear iterative procedure to project a given waveform onto a particular dictionary. For scattering problems, the most appropriate dictionary is composed of wave objects that are consistent with the underlying wave phenomenology. A signal scattered from most targets of interest can be decomposed in terms of wavefronts, resonances, and chirps-and each of these subclasses can be further subdivided based on characteristic wave physics. Here, we investigate the efficacy of applying the method of matching pursuits with a wave-based dictionary for the processing of scattering data. The performance of this algorithm is examined for scattering data with and without additive noise.

Data embedding employing degenerate clusters of data having differences less than noise value

Abstract: A method of embedding auxiliary information into a set of host data, such as a photograph, television signal, facsimile transmission, or identification card. All such host data contain intrinsic noise, allowing pixels in the host data which are nearly identical and which have values differing by less than the noise value to be manipulated and replaced with auxiliary data. As the embedding method does not change the elemental values of the host data, the auxiliary data do not noticeably affect the appearance or interpretation of the host data. By a substantially reverse process, the embedded auxiliary data can be retrieved easily by an authorized user.

Voice activity detector

Abstract: A voice activity detector suitable for deployment in a mobile phone apparatus is disclosed. An advantage of the voice activity detector is that it is better able to provide a decision (79) as to whether an input signal (19) consists of noise (which it is not desired to transmit) or comprises speech or information tones (which are required to be transmitted), especially in noisy environments. The voice activity detector includes a number of components, in particular an auxiliary voice activity detector (3). The auxiliary voice activity detector (3) distinguishes between noise and speech on the basis that the spectrum of speech changes more rapidly than that of noise. This results in the auxiliary detector (3) rarely mistaking a speech signal to be a noise signal. Hence, a very reliable noise template (421) is obtained. For this reason, the auxiliary detector (3) is also useful in noise reduction applications. The voice activity detector also uses a neural net classifier (7).

Noise cancelling circuit for pixel signals and an image pickup device using the noise cancelling circuit

Abstract: The elimination of noise can be positively achieved by a simpler arrangement. For this reason, a capacitor holds a difference between a level at a first time point of an input signal and a reference potential of a given level. A capacitor holds a difference between a level at a second time point of the input signal and a threshold value of an inverter serving as a comparator. The first and second capacitors are switched to a series-connected state. With a combined voltage of holding voltages on the first and second capacitors as an input-side voltage to the inverter, a reference voltage Vref varies as a bias through the input-side voltage. When the output of the inverter is inverted during a portion of a time in which the reference voltage varies, a controller enables the holding of a measurement value corresponding to the reference voltage, the measurement value being used as a variation corresponding to the input signal.

Noise-enhanced capacity via stochastic resonance in an asymmetric binary channel

Abstract: A nonlinear system is considered where an aperiodic binary input signal is added to an arbitrarily distributed noise and compared to a fixed threshold to determine the binary output signal. Noise enhancement of the transmission of the aperiodic signal via stochastic resonance is demonstrated and studied in this nonlinear information channel. The characterization developed goes up to the calculation of the information capacity of the channel, defined as the maximal achievable input-output transinformation occurring when the statistics of the input signal is matched to the noise. It is then demonstrated that a regime exists where the information capacity of the channel can be increased by means of an increase of the noise, up to an optimal noise level where the capacity resonates at a maximum value. The influence on this resonance of the noise distribution is also studied.

Subharmonic quadrature sampling receiver and design

Abstract: A receiver for down-converting a modulated carrier into its in-phase (I) and quadrature (Q) components for further processing is proposed. This is accomplished using a sampling method in which the signal is sampled directly using a sampling circuit which is driven by a single sampling clock frequency substantially lower than the carrier frequency while allowing the I and Q components to be precisely obtained. Each of the signal samples comprises sub-samples taken successively which represent the in-phase, quadrature, negative in-phase and negative quadrature components of the signal. The negative components permit flexible application of the invention in several modes, including differential mode for the removal of common-mode noise. The invention is useful because it provides an integrated circuit means for precisely obtaining I and Q components of a very high frequency modulated carrier. This greatly eases the difficulty of implementing receiver architectures such as direct down-conversion or low-IF receivers, which permit on-chip integration of traditionally difficult-to-integrate components such as IF filters and VCO circuits while eliminating the need for image-rejection filters.

System and method for reducing cumulative noise in a distributed antenna network

Abstract: A communications system and method are provided for reducing noise in a distributed antenna network. The system includes a plurality of remote antenna units (100n) with each remote antenna unit including a receiver for receiving input signals, a signal strength processor (130n) for determining whether a valid signal is present at its respective remote antenna unit (102n) and an output controller (142n) for switching off the network connection when no valid signal is present. As a result, the cumulative noise generated by uplink stages connected to the remote antenna units (100n) is reduced by switching off the network conncection of remote antennas that do not have to be connected. The communications system and method may further include an output controller for controlling the output for its respective remote antenna unit based on the comparison by the signal strength processor. The output controller may control the network connection of the remote antennas by switching off their connection to the network when the signal strength processor determines that the signal strength level is below a noise threshold level. The output controller may control the output of the remote antenna units by switching off their connection to the network when the signal strength processor determines that the signal strength level is below a noise threshold level and by switching on the connection to the network when the input signal strength level is greater than or equal to the noise threshold level.

A closed-form solution to bearings-only target motion analysis

Abstract: Bearings-only target motion analysis is a nonlinear state estimation problem in which the noise corrupted angle of arrival measurements of an emitted signal are used to obtain estimates of the source's range, bearing, course, and speed. The estimation process is complicated by unusual observability properties that render the quality of the estimate highly dependent on both the measurement noise levels and the source-observer geometry. Solutions that use recursive Kalman filtering approach or batch-style algorithms have been reported. The nonlinear batch style estimators for this process require iterative solution methods and under certain scenarios can be sensitive to initial conditions. Pseudolinear solutions that alleviate some of the difficulties with the iterative batch algorithms have been proposed. Although early versions of the pseudolinear filter suffered from biased estimates, subsequent improvements appear to have reduced the bias problem. This paper discusses a new pseudolinear solution based on the observable parameters from individual data segments defined by periods of constant observer velocity (termed "legs"). This solution is a true closed-form solution to the bearings-only target motion analysis problem. Although theoretically interesting, the technique does suffer under conditions of poor observability. A practical pseudolinear estimate, that does not suffer from the same observability problems, is developed and related to the first solution. Algorithm performance results, obtained from computer simulation, are presented. For the scenarios examined, the technique provides good state estimates under conditions of high observability. As observability conditions deteriorate, the solution does develop biases. However, it may still be useful for initializing an iterative nonlinear batch-style estimation algorithm.

Applications of cumulants to array processing. III. Blind beamforming for coherent signals

Abstract: For pt.II see ibid., vol.43, no.7, p.1663-7, 1995. We provide a cumulant-based blind beamforming method for recovery of statistically independent narrowband source signals in the presence of coherent (or perfectly correlated) multipath propagation. Our method is based on the fact that for a blind beamformer, the presence of coherent multipaths is equivalent to the case of independent sources with a different steering matrix. Our approach is applicable to any array configuration having unknown response. Signal sources must have nonzero fourth-order cumulants. There is no need to estimate the directions of arrival. Our method maximizes signal-to-interference plus noise ratio (SINR). A comparable result does not exist using just second-order statistics.