Spiky Noise Research Articles

MRS experiments in a noisy area of a detrital aquifer in the south of Spain

The signal measured in a Magnetic Resonance Sounding (MRS) is of very low amplitude, of the order of tens of nanovolts and is very easily disturbed by the presence of electromagnetic noise of industrial or natural origin. To expand the capabilities of this new geophysical method, it is of great importance to develop field and analytical techniques to reduce the influence of noise in the final results of the MRS. Some field techniques are analyzed in this paper, for several MRS taken in a noisy area at a detrital aquifer in the south of Spain. Two kinds of noise are present, both of amplitude higher than 1500 nV: continuous random noise and spiky noise, with bursts of very high amplitude and occurrence depending on the site and on the time of the day. Trial records made at these places are of little significance, and only complete records taken with the most suitable size, shape and orientation of the antenna, and with a large stacking number, allow to detect the decay pattern of the signal coming from the water-rich layers. The appropriate design of these recording parameters have proven capable of diminishing the effect of random noise, but the results are not as good for spikes, which demand the use of different techniques. The knowledge of the noise pattern obtained in the MRS measurement could be used to design digital filters to attenuate noise influence, together with the aid of surgical muting. The interpretation of the field data has been made without any prior knowledge of the geological information. Despite the high noise level observed on this site, the inversion result achieved present a remarkable correlation with geological data for the water content distribution and are worse for the decay time determination, which seems to be more affected by the existence of noise. Phase and frequency seem to be good quality control parameters before inverting the data, regardless the noisy aspect of the recorded values.

Cycle-accurate energy measurement and characterization with a case study of the ARM7TDMI [microprocessors

Energy characterization is the basis for high-level energy reduction. Measurement-based characterization is accurate and independent of model availability and is thus suitable for commercial off-the-shelf (COTS) components, but conventional measurement equipment has serious limitations in this context. We introduce a new technique for the energy characterization of a microprocessor using a cycle-accurate energy measurement system based on charge transfer which is robust to spiky noise and is able to collect a range of energy consumption profiles in real time. It measures the energy variation of the CPU core by changing the instruction-level energy-sensitive factors such as opcodes (operations), instruction fetch addresses, register numbers, register values, data fetch addresses and immediate operand values at each pipeline stage. Using the ARM7TDMI RISC processor as a case study, we observe that the energy contributions of most instruction-level energy-sensitive factors are orthogonal to the operations. We are able to characterize the energy variation, preserving all the effects of the energy-sensitive factors for various software methods of energy reduction. We also demonstrate applications of our measurement and characterization techniques.

Cumulative probability noise analysis in geophysical spectral records

The technique of rank ordering of spectral energy density samples was applied to high frequency (HF) ocean backscatter radar spectra to examine the properties of signal and noise in the record. HF ocean backscatter spectra have characteristic Bragg scatter lines and bands of energy with random phasing, and it is known that spiky noise from sferics and anthropogenic sources contribute to the observed energy density. Rank ordering of the spectral density data revealed three main features: (i) first-order Bragg scattering; (ii) continuum of randomly phased energy; and (iii) the underlying noise floor of the system. It is shown that the technique of rank ordering of spectral amplitude samples provides better insights and an improved algorithm for differentiating between bands of randomly phased sea echoes and system noise. The analysis reveals a constant offset in the spectral energy density data from the Ocean Surface Current Radar (OSCR) HF radar system, which should be removed before algorithms are applied to extract significant wave height information. The rank ordering of spectral energy density samples is a useful tool for differentiating between system noise and wide-band incoherent energy often obtained in a wide range of geophysical remote sensing.

Robust image association by recurrent neural subnetworks

A neural network consisting of a gallery of independent subnetworks is developed for associative memory which stores and recalls gray scale images. Each original image is encoded by a unique stable state of one of neural recurrent subnetworks. Comparing to Amari-Hopfield associative memory, our solution has no spurious states, is less sensitive to noise, and its network complexity is significantly lower. Computer simulations confirm that associative recall in this system for images of natural scenes is very robust. Colored additive and multiplicative noise with standard deviation up to σ=2 can be removed perfectly from normalized image. The same observations are valid for spiky noise distributed on up to 70% of image area. Even if we remove up to 95% pixels from the original image in deterministic or random way, still the network performs the correct association.

Robust Time-Varying AR Parameter Estimation

A new method for estimating time-varying parameters of nonstationary AR signal models, based on robust recursive least squares with variable forgetting factors, is described. The robust estimator differs from the conventional one by the insertion of nonlinear residual transformation, which has to suppress the influence of the spiky noise or outliers, while the VFF's are adapted to the signal via the robustified generalized likelihood ratio detection scheme. The method has good adaptability in the nonstationary situations, and gives low bias and low variance at the stationary situations. The feasibility of the approach is demonstrated with both simulation and experimental data.

Optimum detection of fading signals in impulsive noise

This paper deals with the synthesis and the analysis of optimum receivers to detect one out of M equally likely, equi-energy, fading signals in impulsive noise, modelled as a compound Gaussian, possibly correlated process. We show that the conventional coherent and incoherent detectors are still optimum, independent of the noise as well as the fading probability density functions. The performance analysis has been carried on with reference to the general case of arbitrarily distributed disturbance: in order to simplify the analysis, asymptotical expressions have been developed for high signal-to-noise ratios as well as high signal space dimensionality. Interestingly enough, this allows separating the effect of the noise spikyness from that of the fading law. Results indicate that, for deep fading, the noise marginal distribution does not dramatically affect the error probability, nor is it influential on the limit operating characteristics corresponding to infinite signal space dimensions. For non-fluctuating signals, instead, the noise distribution plays a primary role: spiky noise usually produces performance impairment; moreover, the limit performance in impulsive disturbance may exhibit marked deviations from the well-known stepwise shape which is typical of Gaussian channels. >

Wave mode identification of electrostatic noise observed with ISEE 3 in the deep tail boundary layer

Characteristics of VLF electrostatic noise observed with ISEE 3 in the low‐latitude boundary layer of distant geomagnetic tail (∼210Re) have been examined in detail. Gray tone dynamic spectra with high time resolution of 16 frequency channels of electric field exhibit successive spiky bursts with duration less than the wave sampling period (0.5 s) and with a wide frequency extent. A set of noise parameters such as peak amplitude, its frequency, and bandwidth has been determined from a Gaussian distribution function derived for each spiky noise spectrum by using a least squares fitting routine. In order to find a possible energy source for the noise generation, available correlations between the noise parameters and plasma electron parameters have been examined for a 2‐hour data interval when ISEE 3 was in the low‐latitude boundary layer. The analysis has been conducted on the basis of a hypothesis that the wide frequency extent of the noise spectra is composed of Doppler effects of waves propagating nearly omnidirectionally within the plasma rest frame, which is moving with the electron bulk speed. On the basis of the hypothesis the wavelength of the observed waves has been determined from the width of the frequency extent and the measured electron bulk speed. Although no obvious free energy source has been found from the analysis, the obtained dispersion relation for the electrostatic noise has demonstrated the validity of the hypothesis. The obtained dispersion relation is consistent with theoretical ion acoustic dispersions for the measured electron parameters. The result shows that the wavelength ranges from 2 to 8 times the plasma Debye length. The dispersion also indicates that the ion temperature is in a range from 1 to 4 times the electron temperature.

Nonlinear filtering of delta modulated signals

Nonlinear filtering of delta modulated signals is introduced. The proposed technique suppresses spiky noise for which a simple model is developed. Look-up tables are employed to facilitate and speed up the implementation of the algorithm. Considerable similarity with conventional median filters is observed.

Acoustic detection of milling tool touch to a workpiece

The initial contact of a rotating milling tool to a workpiece is detected in the presence of spurious noise spikes by digitally processing samples of a preprocessed vibration signal concurrently through basic tool touch detection logic and milling tool touch detection enhancement logic, and generating a touch indication if either logic set is satisfied. The basic logic detects a continuously increasing signal level higher than a threshold set above background noise. The milling logic detects an above-threshold signal sample and sets up an acceptance window about the expected time of contact of the next cutter. Tool touch is declared when an above-threshold signal occurs during the window; the milling logic is reset by a noise spike in the interval before the window starts. The basic logic detects tool touch in a dense spiky noise situation.

Convergence of iterated median rules

Application for the median rule to a digital image smooths spiky noise without blurring edges. It is natural to iterate the median rule in order to reduce the noise still further. One then desires to know whether repeated application of the median rule leads to convergence of the image. Suppose we apply iteratively a median rule with a symmetric neighborhood. If the rules are applied to the sites sequentially, then we prove that the iterated images converge. If, instead, the rules are applied in parallel, then the images are proved either to converge or to get trapped in a periodic orbit of period 2.

Detector to discriminate between tool break acoustic signature and spiky noise

First Page

Image restoration by an iterative damped least-squares method with non-negativity constraint

In this paper we describe a new digital method for restoring linearly degraded images in the presence of noise. The restoration procedure is an iterative damped least-squares (DLS) algorithm that is based on the principle of damped least squares. This method acquires the advantage of tolerance to noise by incorporating additional constraints of non-negativity of the object and the adaptive technique. We also use a median filter in the proposed algorithm to suppress spiky noise. After discussing the convergence of the iterative DLS algorithm, we present some results of computer simulations that demonstrate the effectiveness of the proposed method.

Image restoration by an iterative regularized pseudoinverse method

In this paper we describe a digital method for restoring linearly degraded images in the presence of noise. The restoration procedure is an iterative regularized pseudoinverse (RPI) algorithm that is based on the principle of least squares. This method acquires the advantage of tolerance to noise by incorporating additional constraints of non-negativity of the object and adaptive regularization. We also use a median filter in the proposed algorithm to suppress spiky noise. We present some results of computer simulations that include both space-invariant and space-variant degradations. The effectiveness of the iterative RPI method is demonstrated through computer simulations.