Dai Yisong

Adaptive denoising based on wavelet thresholding method

The paper presents a novel adaptive denoising method based on a wavelet thresholding method. First, it presents a new thresholding function that has a continuous derivative while the derivative of a standard thresholding function is not continuous. The new thresholding function makes it possible to construct an adaptive algorithm based on the wavelet thresholding method. Second, by using the new thresholding function, the paper presents an adaptive method based on SURE (Steins unbiased risk estimate) risk. Finally, several numerical analysis examples are given; the results show that the proposed method is very effective in finding the optimal solution in the mean square error (MSE) sense. It also indicates that this method gives better MSE performance than other wavelet thresholding methods.

An improved approach and experimental results of a low-frequency noise measurement technique used for reliability estimation of diode lasers

Abstract In this paper, we give the experimental results of low-frequency noise measurement and reliability forecasting after we have studied the low-frequency noise and electric ageing of 23 GaAs/GaAlAs proton-bombarded heterostructure LDs and 21 V-groove InP/InGaAsP lasers. We discuss the measuring circuits, measuring parameters, and suggest an improvement.

Cross-spectral moment and SVD methods estimating the parameters of close sinusoids in colored noise

The cross-spectral estimation methods are efficient in estimating the parameters of sinusoidal signals embedded in colored noise. But up to now, only FFT and cross-periodogram methods are used in this field. The modern auto-spectral estimation method is introduced into cross-spectral estimation in this paper, meanwhile the cross-correlation based Yule-Walker equation is proposed theoretically and the moment and singular-value decomposition (SVD)) algorithms for cross-spectral estimation have been developed. Finally, a numerical example is given for comparing the presented methods with the well-known Cadzows SVD method.

Low-frequency low-noise circuits design using an En-In model

In view of the limitations of a Rn-Gn model in the low frequency range and the defects of an En-In model in common use now, this paper builds a complete En-In model according to the theory of random harmonic. The parameters for the low-noise design such as the equivalent input noisy voltage Ens, the optimum source impedance Zsopt and the minimum noise figure Fmin can be calculated accurately by using this En-In model because it considers the coherence between the noise sources fully. Moreover, this paper points out that it will cause the maximum 30% miscalculation when neglecting the effects of the correlation coefficient γ. Using the series-series circuits as an example, this paper discusses the methods for the En-In noise analysis of electronic circuits preliminarily and demonstrates its correctness through the comparison between the simulated and measured results of the minimum noise figure Fmin of a single current series negative feedback circuit.

A novel continuous-time current-mode biquad filter employing CCIIs

A novel configuration of a CCII-based continuous-time current-mode biquad filter is proposed in this paper. The filter is only made of minim CCII elements and grounded resistors and capacitors; the topology is really simple and the performance is really good, so it is suited for full integration. A computer simulation is carried out and the results show that the proposed circuits are effective.

An improvement of the recursive closed-form algorithm for non-minimum phase fir system identification

An improved algorithm which is based on the recursive closed-form algorithm for non-minimum phase FIR system identification via higher order statistics is presented. In order to increase the parametric estimation accuracy, the improved algorithm uses the optimal iterative method to seek the nonlinear least-square solution. Finally, the simulation examples are also given.

Measurement of low-frequency noise spectrum and computation of its parameters

In this paper, a low-frequency noise spectrum testing system is presented, its measuring range is 0.25 Hz approximately 100 kHz, and accuracy is higher than 4%. Moreover, the white noise level and corner frequency are computed by applying the weighed least square method.<<ETX>>