Showing papers on "Noise measurement published in 1994"

1/f noise sources

Abstract: This survey deals with 1/f noise in homogeneous semiconductor samples. A distinction is made between mobility noise and number noise. It is shown that there always is mobility noise with an /spl alpha/ value with a magnitude in the order of 10/sup -4/. Damaging the crystal has a strong influence on /spl alpha/, /spl alpha/ may increase by orders of magnitude. Some theoretical models are briefly discussed none of them can explain all experimental results. The /spl alpha/ values of several semiconductors are given. These values can be used in calculations of 1/f noise in devices. >

A robust algorithm for word boundary detection in the presence of noise

Abstract: The authors address the problem of automatic word boundary detection in quiet and in the presence of noise. Attention has been given to automatic word boundary detection for both additive noise and noise-induced changes in the talker's speech production (Lombard reflex). After a comparison of several automatic word boundary detection algorithms in different noisy-Lombard conditions, they propose a new algorithm that is robust in the presence of noise. This new algorithm identifies islands of reliability (essentially the portion of speech contained between the first and the last vowel) using time and frequency-based features and then, after a noise classification, applies a noise adaptive procedure to refine the boundaries. It is shown that this new algorithm outperforms the commonly used algorithm developed by Lamel (1981) et al. and several other recently developed methods. They evaluated the average recognition error rate due to word boundary detection in an HMM-based recognition system across several signal-to-noise ratios and noise conditions. The recognition error rate decreased to about 20% compared to an average of approximately 50% obtained with a modified version of the Lamel et al. algorithm. >

A recursive multiple model approach to noise identification

Abstract: Correct knowledge of noise statistics is essential for an estimator or controller to have reliable performance. In practice, however, the noise statistics are unknown or not known perfectly and thus need to be identified. Previous work on noise identification is limited to stationary noise and noise with slowly varying statistics only. An approach is presented here that is valid for nonstationary noise with rapidly or slowly varying statistics as well as stationary noise. This approach is based on the estimation with multiple hybrid system models. As one of the most cost-effective estimation schemes for hybrid system, the interacting multiple model (IMM) algorithm is used in this approach. The IMM algorithm has two desirable properties: it is recursive and has fixed computational requirements per cycle. The proposed approach is evaluated via a number of representative examples by both Monte Carlo simulations and a nonsimulation technique of performance prediction developed by the authors recently. The application of the proposed approach to failure detection is also illustrated. >

Reduction of background noise for speech enhancement

Abstract: Properties of human audio perception are used to perform spectral and time masking to reduce perceived loudness of noise added to speech signals. A signal is divided into blocks (2), passed through notch filters (4) to remove noise components and then appended to part of the previous block (6). An FFT (8) is then performed on the resulting block and the spectral components are fed to noise estimator (20). Each frequency component is then analyzed to determine whether it is noise. The frequency component's gain function is determined and a spectral valley filler (38) is used to processed the gain function after which the function is used to modify magnitude components of the FFT (12). In inverse FFT (14) then maps the signal back to the time domain to give a frame of noise-reduced signal.

Noise-reduction method for noise-affected voice channels

Abstract: A method that can be used not only for elimination of noise, for example in automatic speech recognition, but also to improve the voice quality for people, for instance during use of the speaker function of a car phone. The noise reduction is executed with two or multiple channels in such a manner that the temporal and architectural acoustical signal properties of speech and interference are utilized step-by-step and systematically. According to the method a pivotable, acoustic directional lobe is produced for the individual voice channels by respective digital directional filters and a linear phase estimation to correct for a phase difference between the two channels. The noise in the individual voice channels is estimated during speaking pauses, and the temporally stationary noise sources are damped by means of spectral subtraction. The individual voice channels are subsequently added whereby the statistical disturbances of spectral subtraction are averaged. Finally, the composite signal resulting from the addition is subsequently processed with a modified coherence function to damp diffuse noise and echo components.

Morphological constrained feature enhancement with adaptive cepstral compensation (MCE-ACC) for speech recognition in noise and Lombard effect

Abstract: The use of present-day speech recognition techniques in many practical applications has demonstrated the need for improved algorithm formulation under varying acoustical environments. This paper describes a low-vocabulary speech recognition algorithm that provides robust performance in noisy environments with particular emphasis on characteristics due to the Lombard effect. A neutral and stressed-based source generator framework is established to achieve improved speech parameter characterization using a morphological constrained enhancement algorithm and stressed source compensation, which is unique for each source generator across a stressed speaking class. The algorithm uses a noise-adaptive boundary detector to obtain a sequence of source generator classes, which is used to direct noise parameter enhancement and stress compensation. This allows the parameter enhancement and stress compensation schemes to adapt to changing speech generator types. A phonetic consistency rule is also employed based on input source generator partitioning. Algorithm performance evaluation is demonstrated for noise-free and nine noisy Lombard speech conditions that include additive white Gaussian noise, slowly varying computer fan noise, and aircraft cockpit noise. System performance is compared with a traditional discrete-observation recognizer with no embellishments. Recognition rates are shown to increase from an average 36.7% for a baseline recognizer to 74.7% for the new algorithm (a 38% improvement). The new algorithm is also shown to be more consistent, as demonstrated by a decrease in standard deviation of recognition from 21.1 to 11.9 and a reduction in confusable word-pairs under noisy, Lombard-effect stressed speaking conditions. >

An SNR estimation algorithm using fourth-order moments

Abstract: An algorithm is presented that allows an estimation of the SNR just by the observation of the noisy signal. For the estimation, shape factors of the signal's and the noise's probability density functions are used. The algorithm is based on the second and fourth order moments of the observed noisy signal. >

Residual noise shaping technique for active noise control systems

Abstract: An active noise control system attenuates the overall sound field. However, in some applications, it is desirable to change the spectral contents of the residual noise. In this Letter, a residual noise spectrum shaping technique based on the filtered‐E least‐mean‐square algorithm has been developed. This technique can be applied to active noise control of one‐dimensional ducts and three‐dimensional enclosures, for both narrow‐band and broadband noises. Computer simulations demonstrate that this method not only attenuates the noise level, but also effectively reshapes the spectrum of the residual noise.

UN-MUSIC and UN-CLE: an application of generalized correlation analysis to the estimation of the direction of arrival of signals in unknown correlated noise

Abstract: A new approach is proposed for the consistent estimation of the directions of arrival (DOA) of signals in an unknown spatially-correlated noise environment. The signal and noise model used is based on the assumption that the data are received by two arrays well separated so that their noise outputs are uncorrelated. The generalized correlation decomposition of the cross-correlation matrix between the two arrays is then introduced. Of particular interest is the canonical correlation decomposition. The analysis of the generalized correlation leads to various interesting geometric and asymptotic properties of the eigenspace structure. Two algorithms, UN-MUSIC and UN-CLE, are developed to estimate the DOA of signals in unknown spatially correlated noise based on the utilization of these properties. Computer simulations show that these methods are superior in performance compared to conventional methods. Furthermore, it is demonstrated that the new methods are equally effective even when only one sensor array is employed. >

Noise reduction method and apparatus

Abstract: A method and apparatus is disclosed to reduce noise which overlaps detected signals in a common frequency band for detection of distributed signals from multiple measurement points. The method decomposes a input signal into a noise-free part called basic signal and a noise-overlapping part called residual signal. Noise contained in residual signal is recognized by comparing it with a comparison signal. The comparison signal corresponding to a measurement point is computed by a linear combination of all other points weighted with parameters which reflect amplitude relationship of distributed signals and are calculated during a learning phase at the beginning of measurement. With residual signal and corresponding comparison signal, a noise index is calculated and then used to extract signal components from the residual signal. The extracted signal components are added to basic signal as the output of the method. The method is effective in reducing signal-overlapping noise with little distortion of signal.

Analytical and experimental studies of thermal noise in MOSFET's

Abstract: An analysis of the channel thermal noise in MOSFET's based on the one-dimensional charge sheet model, is presented. The analytical expression is valid in the strong, moderate, and weak inversion regions. The body effect on the device parameters relevant to the thermal noise is discussed. A measurement technique as well as experimental results of P- and N-MOSFET's of a 1.2 /spl mu/m radiation hard CMOS process are presented. The calculated channel thermal noise coefficient /spl gamma/ as in i/sub d//sup 2///spl Delta/f=4kT /spl gamma/ g/sub do/, agrees well with experimental data for effective device channel length as short as 1.7 /spl mu/m. >

Single-sensor active noise cancellation

Abstract: Active noise cancellation is an approach to noise reduction in which a secondary noise source that destructively interferes with the unwanted noise is introduced. In general, active noise cancellation systems rely on multiple sensors to measure the unwanted noise field and the effect of the cancellation. This paper develops an approach that utilizes a single sensor. The noise field is modeled as a stochastic process, and a time-adaptive algorithm is used to adaptively estimate the parameters of the process. Based on these parameter estimates, a canceling signal is generated. In general, the transfer function characteristics from the canceling source to the error sensor need to be accounted for. If these can be accurately measured in advance and are invertible except for the propagation delay between the source and sensor, then the essential problem becomes one of predicting future values of the noise field. The algorithm developed is evaluated with both artificially generated noise and with recordings of aircraft noise. >

Noise reduction system for binaural hearing aid

Abstract: In this invention noise in a binaural hearing aid is reduced by analyzing the left and right digital audio signals to produce left and right signal frequency domain vectors and thereafter using digital signal encoding techniques to produce a noise reduction gain vector. The gain vector can then be multiplied against the left and right signal vectors to produce a noise reduced left and right signal vector. The cues used in the digital encoding techniques include directionality, short term amplitude deviation from long term average, and pitch. In addition, a multidimensional gain function based on directionality estimate and amplitude deviation estimate is used that is more effective in noise reduction than simply summing the noise reduction results of directionality alone and amplitude deviations alone. As further features of the invention, the noise reduction is scaled based on pitch-estimates and based on voice detection.

Detection for binary transmission in a mixture of Gaussian noise and impulsive noise modeled as an alpha-stable process

Abstract: The impact of nonGaussian impulsive noise combined with Gaussian noise on the performance of the binary transmission is analyzed The impulsive noise is modeled as an alpha-stable process The probability of error for optimum, linear and nonlinear receivers is derived The proposed nonlinear detectors show substantial improvements in performance compared to linear ones The obtained results will be useful in performance evaluation of digital communication links subject to Gaussian and impulsive noises >

Magnetic flux noise in copper oxide superconductors

Abstract: We report on the magnetic flux noise in thin films of YBa2Cu3O7-x (YBCO), Tl2Ca2Ba2Cu3Ox, and TlCa2Ba2Cu3Ox and in crystals of YBCO and Bi2Sr2CaCu2O8+x, measured with a Superconducting QUantum Interference Device (SQUID). We ascribe the noise to the motion of flux vortices. In the low magnetic fields in which the experiments are performed the average vortex spacing always exceeds the superconducting penetration depth. The spectral density of the noise usually scales as 1/f (f is frequency) from 1 Hz to 1 kHz and increases with temperature to a peak which is of the same magnitude in all samples, at the transition temperature. Furthermore, the noise power increases with the magnitude of the magnetic field in which the sample is cooled, with a power-law dependence over several decades, whereas a supercurrent well below the critical current density applied to YBCO films suppresses the noise power by an order of magnitude. Most of the measurements were made on YBCO films, and for this set of samples the noise decreases dramatically as the crystalline quality is improved. A model of thermally activated vortex motion is developed which explains the dependence of the noise on frequency, temperature, magnetic field, and current. The pinning potential is idealized as an ensemble of symmetrical double wells, each with a different activation energy separating the two states. From the noise measurements, this model yields the distribution of pinning energies, the vortex hopping distance, the number density of mobile vortices, and the restoring force on a vortex at a typical pinning site. The distribution of pinning energies in YBa2Cu3O7-x shows a broad peak below 0.1 eV. Over narrow temperature intervals, most samples exhibit random telegraph signals in which the flux switches between two discrete levels, with activation energies and hopping distances much greater than those deduced from the 1/f noise measurements.

On the limitations of cepstral features in noise

Abstract: This paper examines the effects of additive Gaussian noise on the short-term cepstral analysis of speech. We identify three distinct modifications to the long-term statistics of the cepstrum that cause a gross mismatch after the addition of noise, namely: a mean shift, a change of variance and a distribution distorted from normal, with distinct bimodal characteristics. We assess the importance of each of these, and demonstrate the limitations of simple cepstral mappings. We then consider noise-masking, through the addition of a constant offset to the linear spectral estimates, which provides a feature space far more stable to changes in noise statistics. This leads to performance equivalent to that achieved by explicit modelling. >

Continuous speech recognition in noise using spectral subtraction and HMM adaptation

Abstract: This paper describes a scheme fur robust speech recognition at very poor signal to noise ratios. It consists of a continuous spectral subtraction (CSS) scheme integrated into a parallel model combination (PMC) compensation framework. In this CSS-PMC scheme, a smoothed estimate of the long term spectrum is continuously calculated and subtracted from the signal. At the same time, the HMMs are compensated using PMC for the signal distortion caused by the CSS stage. The paper presents an evaluation of the CSS-PMC approach using the Noisex 92 database. The results show high recognition performance for very noisy environments. For example, for the Lynx helicopter noise, CSS-PMC gives 97% accuracy at 0 dB SNR. >

Bias of mean value and mean square value measurements based on quantized data

Abstract: This paper investigates the imperfect fulfilment of the validity conditions of the noise model quantization. The general expressions of the deviations of the moments from Sheppard's corrections are derived. Approximate upper and lower bounds of the bias are given for the measurement of first- and second-order moments of sinusoidal, uniformly distributed, and Gaussian signals. It is shown that because of the uncontrollable mean value at the input of the ADC (offset, drift), the worst-case values have to be investigated; it is illustrated how a simple-form envelope function of the errors can be used as an upper bound. Since the worst-case relative positions of the signal and the quantization characteristics are taken into account, the results are valid for both midtread and midrise quantizers, while in the literature results are given for a selected quantizer type only. >

An interacting multiple model approach for target tracking with glint noise

Abstract: The application of the interacting multiple model (IMM) approach to the problem of target tracking when the radar measurements are perturbed by glint noise is considered. The IMM is a very effective approach when the system has discrete uncertainties in the dynamic or measurement model as well as continuous uncertainties. It is shown that this method performs better than the nonlinear filtering algorithm known as the "score function" method. It is also shown that the IMM algorithm method performs robustly when the exact prior information on the glint noise is not available, while the other method requires a more accurate statistical knowledge of the glint phenomenon. >

Telephone set with background noise suppression function

Abstract: A telephone set is provided with a first microphone at the position close to a speaker and a signal including remaining noise is outputted from the first microphone. Background noise is outputted from a second microphone. A noise elimination circuit estimates background noise reaching the ear of a receiver based on the background noise signal, the signal including the remaining noise, and a received voice signal from a transmitter. Also, the noise elimination circuit subtracts the estimated background noise from the received voice signal to supply the resultant signal to the speaker.

Encoding rate selection in a variable rate vocoder

Abstract: It is a first objective of the present invention to provide a method by which to reduce the probability of coding low energy unvoiced speech as background noise. The present invention determines an encoding rate by examining subbands of the input signal, by this method unvoiced speech can be distinguished from background noise. A second objective of the present invention is to provide a means by which to set the threshold levels that takes into account signal energy as well as background noise energy. In the present invention, the background noise is not used to determine threshold values, rather the signal to noise ratio of an input signal is use to determine the threshold values. A third objective of the present invention is to provide a method for coding music passing through a variable rate vocoder. The present invention examines the periodicity of the input signal to distinguish music from background noise.

Fuzzy filtering for mixed noise removal during image processing

Abstract: Presents a new nonlinear fuzzy filter for image processing in a mixed noise environment, where both additive Gaussian noise and nonadditive impulsive noise may be present. Averaging filters can effectively remove the Gaussian noise and order statistics filters or median filters can effectively remove the impulsive noise. However, it is difficult to combine these filters to remove mixed noise in an image processing environment without blurring the image details or edges. Trying to distinguish between noise and edge information in the image is an inherently ambiguous problem and naturally leads to the development of a fuzzy filter. We use local statistics to train the membership function of a fuzzy filter for image processing to remove both Gaussian noise and impulsive noise while preserving edges. We show that such a fuzzy filter gives superior results when compared to averaging filters, median filters, and other fuzzy filters. We also demonstrate the robustness of this filtering technique. >

Transition noise analysis of thin film magnetic recording media

Abstract: A simple but accurate expression for general non-stationary noise correlation in the presence of a recorded transition is analyzed in terms of both noise voltage and spectral measurements. The parameters of this analysis are solely the cross track correlation width s, the transition shape and parameter a, the head-medium spacing d, and the replay gap length g. It is shown that although the noise varies continuously through the transition, a reasonable decomposition that accounts for a large percentage of the total noise is into conventional position and amplitude jitter of a fixed transition shape. The relative weights depend on the head-medium parameters; for current head-medium configurations and for longitudinal recording, position jitter dominates. A simple closed form expression for the noise power spectrum is given. Published experimental measurements of signal and noise spectra made with pseudo-random write data compare extremely well with this theoretical analysis, and lead to very good estimates for a and s. The analysis is general and applies for low-density recording with both inductive and magnetoresistive heads as well as all magnetization orientations. >

Tracking low elevation targets in the presence of multipath propagation

Abstract: A tracking algorithm based on a measurement model of the multipath propagation effects on the elevation angle measurements is presented. This algorithm does not depend on the specific signal processing and the prevailing environmental conditions. The results of simulations based on several scenarios are presented. The signal processing model used to generate the elevation measurements of a low-flying target with multipath effects is also described. This mode incorporates a spherical Earth and finite target range in the geometry of reflection. The algorithm clearly indicates when the measurements are degraded, in which case it automatically increases the variance of the altitude estimate. >

Switched-current memory circuits for high-precision applications

Abstract: We discuss circuit parameters that limit the precision of basic dynamic current-memory cells. In addition to analyzing current-copying errors caused by the finite output conductances of the current sources and by the clock-feedthrough (CFT) of the feedback switches, we analyze the noise performance of the basic memory cell. To reduce CFT and noise, we propose a novel circuit based on Miller capacitance-enhancement. Measurement results of memory cells integrated in a 1-/spl mu/m CMOS process confirm the theoretical findings; with our CFT and noise reduction technique based on Miller enhanced capacitance and dummy switches, we achieve a dynamic range of 11 b at clock frequencies greater than 100 kHz. >

Electrical noise and VLSI interconnect reliability

Abstract: This paper discusses the characteristics of noise sources in Al-based thin films and their relationships to VLSI reliability. Techniques of applying noise measurements in detecting existing defects/damages in the films, determining electromigration activation energy, and predicting the time to failure of VLSI interconnects are presented. The noise measurement technique can be applied to wafer-level reliability testing because it is much faster than the conventional MTF method and is nondestructive in nature. Some important considerations for wafer-level reliability testing via noise measurements are also presented in this paper. >

High-frequency photodiode characterization using a filtered intensity noise technique

Abstract: Optical filtering of amplified spontaneous emission improves measurement dynamic range for frequency response measurements of optoelectronic receivers. For high bandwidth receivers, a novel periodically filtered intensity noise technique is proposed. Response measurements using these techniques on a 1 GHz and 30 GHz receiver are demonstrated. >

Phase noise measurement system and method.

Abstract: Three independent signal sources are used to statistically derive the power spectral density of the phase noise content of signals from each of them. This is accomplished by mixing each of the signals two at a time (i.e., signal one with signal two, signal one with signal three, and signal two with signal three) and capturing the resultant difference signals, such as with a waveform recorder, for example. A servo electronics loop is used to remove the carrier and any long term signal drift from the resultant difference signals. Statistical analysis is then used to compute the composite power spectral densities of the the resultant difference signals, and to solve for the individual power spectral densities of the original signals. The present system and method uses the mathematical relationships between the three sources that have similar magnitudes of phase noise, to compute the power spectral density of the noise content of signals from each source. The present system and method requires a minimum of interconnect hardware and only three inexpensive waveform recorders. Furthermore, the size, weight, and cost of producing the present phase noise test system is relatively low.

Low residual noise speech enhancement utilizing time-frequency filtering

Abstract: Spectral subtraction is a well known technique for enhancing speech corrupted by additive wideband noise. In this technique, the "clean" signal is approximated by subtracting a noise estimate from the spectrum of the corrupted signal. A negative side effect is the residual "musical" noise that is produced when isolated spectral peaks exceed the noise estimate. In this paper, a low residual noise enhancement method is presented. This method is based on spectral subtraction but incorporates an algorithm developed to suppress "musical" noise without affecting speech. The algorithm is referred to as time-frequency filtering because spectral peaks due to noise are eliminated on the basis of duration, bandwidth, and proximity to other peaks. Results showing the effects of combining spectral subtraction and time-frequency filtering are given. >