In recent years, the wavelet transform has been widely used in many fields. The fast tower-type algorithm is an efficient tool. Its importance is just like that of FFT in the Fourier analysis. Therefore the hardware implementation of the DWT is an important problem in its applications. With the introduction of the parallel systolic FIR filter structure to the design of the wavelet decomposition filter, an implementation structure of wavelet decomposition filter is presented in this paper. As being parallel and systolic, the structure can increase data throughput and reduce power consumption of the system as well as enhance operation speed.

Binary Phase Coded Waveform(BPCW) is a widely used waveform in radar pulse compression, but the Sidelobe Suppression Filter(SSF) designing algorithms for BPCW are always limited to short codes such as Barker code, and the ranks of the SSF are always high. A new kind of SSF based on time domain synthesis is presented. It can be used for various length codes, including the m-sequence and so on. Some simulation results of time domain synthesized SSF for the Barker code and m-sequence are given, with the performance comparison of the new filter with some known filters.

Authors, in this paper, propose and analyze a new kind of detail-preserving adaptive filter algorithm according to relations between local temporal signal and its order statistics. Theoretical analysis and simulation results show that the adaptive filters determined by similarity measurement between temporal signal and monotonic order statistics which belongs to the root signal set of standard median filter have good performances in the aspects of removal of complicated distributed noises as well as preservation of signal edges and details.

This paper proposes an adaptive multiwavelet prefilter. The prefilter deals with the multi-wavelet initialization problem resulting from the non-lowpass property of multiscaling functions. Theory analysis and simulation results show that the approximation accuracy of the prefiltering algorithm is satisfactory.

In this paper, a set of filters is used to extract features of gray scale s distribution from an image. Different features detect different kinds of edges. In the end, those results are integrated to get the ultimate image edge. The method presented in this paper turns edge detection into a problem of pattern recognition. So the method can fit to a complex edge model. Experimental results prove that the presented method is feasible and effective.

Two-scale IR imaging model of sea surface based on rigid object is built in this paper. The scattering properties of sea surface are discussed from the aspect of active imaging. The emulational results based on this model show that it is a good model to explain the effect of wind on IR active imaging of sea surface.

To counter the disadvantage of polynomial method for image registration, this paper studies the image registration method applying orthogonal polynomials and weighted orthogonal polynomials. The results show that the method of orthogonal polynomials is more flexible; and the method of weighted orthogonal polynomials has good local characters. Finally, the experimental results of these methods are compared and testified.

Based on dynamic programming, a time-varying wavelet packets decomposition algorithm is proposed. Using this algorithm, the problems, that the single tree algorithm cannot adapt to nonstationary signals and the double tree algorithm has strict binary restriction, are solved. Simulation results varify that this algorithm is more flexible than the others in time-varying signal processing.

Signals representation in dyadic wavelet domain is very redundant. Compared with wavelet series reconstruction, signals dyadic wavelet reconstruction dependency on the individual coefficients in transform domain will be decreased. Therefore, with the same error decision probability, the better reconstruction can be expected. Based on this idea, this paper extends the existing wavelet-based denoising approaches to the dyadic wavelet-based denoising. Numerical experiments show that the dyadic wavelet-based denoising can significantly improve the signal-to-noise rate (SNR).

For bandpass signal under digital quadrature sampling, there is spectrum shifting. Although the sampling frequency satisfies the sampling theorem, it may cause spectrum aliasing. The digital quadrature sampling theorem for bandpass signal is presented and the optimum design for the sampling system is also discussed in this paper.

2-D projective collineation provides an ideal mathematical tool for describing planar element correspondence in multiple views. This paper extends the solution methods for 2-D projective transformation matrix from 4 points corresponding to variety combinations of points and lines corresponding, such as 4 points, 4 lines, 1 point and 3 lines, 3 lines and 1 point, and 2 points and 2 lines, etc. Based on the relationship between two views indicated as 2-D projective collineation, prediction problems about points or lines from two views or three views can be solved effectively. Without any estimating parameters involved, the prediction results’ accuracy is improved. Freely combining points and lines on describing relationship between views makes the application field wider. The methods proposed here have been verified by experimental results.

Digital watermarking is a new research topic about copyright protection in recent years. Watermarking is a technique for labeling digital picture by hidding secret information in the images. Sophisticated watermark embedding is a potential method to discourage unauthorized copying or attacking the images. This paper presents a technique of watermark embedding and extracting based on discrete wavelet transform of blocks in the image. Each pixel of a watermark is embedded in the wavelet coefficient of the middle and low frequency of a block in the images. Disputation about copyright can be settled down by using the watermark embedding technique presented in this paper. Both theoretical analysis and experimental results show that this technique is reasonable and satisfactory.

In this paper a method of joint source/channel coding is presented. For a source with separate symbols, codes having desired minimum Hamming distance dmin are selected within range [0,1). Then, under the case having channel error, the decoder utilizes the dmin, the corresponding relationship between code and symbol, and the statistical characteristic of source to recover the original sequence of symbols without the use of channel coding. The introduced scheme is suitable for bandwidth-limited binary symmetric channel (BSC) with fairly high error rates.

This paper describes the principles of Unique Word(UW) detection and the design methods for a UW-detector at lower SNR. An engineering solution for a general purpose UW-detector based on maximum likelihood criteria is proposed. The simulation results show a 3-4dB improvement over the often employed hard-decision/threshold detection when the length of UW is 16 and the SNR is low. The developed UW-detectors have been used in auther s recent communication systems.

One disadvantage with subspace-based DOA estimate method is that it is difficult to incorporate prior knowledge of the signal correlation into the eigen-decomposition. Herein, an estimator that combines ideas from subspace and four-order cumulant matching method is proposed, which makes it possible to incorporate the prior knowledge of signal correlation and sparsity of the DOA estimation into the estimator. Theoretical analysis and numerical results illustrate the robust and high resolution character of the estimator.

In this paper, several methods of evidence assignment and combination of optimum interval with Dempster-Shafer theory are proposed to identify communication signals. Under low SNR circumstance, the identifiability is highly promoted. Moreover, the design of classifier is simplified. The availabilities of these methods are verified by computer simulation.

In this paper, the function approximation of Gelenbe Neural Network (GNN) is discussed and it is proved that GNN can approximate any G-type polynomial by using constructional method. Number of units used by this method can be reduced to 2/(n+1) of that used by previous methods for a same G-type polynomial, where n is the degree of G-type polynomial.

This article presents a neural network model for the optimization of arbitrary convex functions with linear constraints. The equilibrium point of the energy function constructed is the optima] solution of the original problem. The problems, which would arise in conventional neural network optimization methods, are overcome. The neural model is globally stable and can converge to the optimal point. The computer simulation results verify the effectiveness of the method.

Automatic word segmentation for the Chinese language is a fundamental and difficult problem in the field of computer Chinese language information processing. This paper presents a new method for segmenting the input Chinese language text sentence into words, which consists of a character-based N-gram model and an efficient Viterbi search algorithm. In addition, two performance evaluation ration targets, i.e. Recall and Precision for word segmentation algorithm are discussed, The effectiveness has been confirmed by evaluation experiments using the closed texts and open texts corpus.

This paper presents the basic concept of Extended Wavelet Transform (EWT) algorithm based on Extended WVD (EWVD) transform theory after analyzing the spectrum characteristics of airborne SAR moving-target backscatters, and then takes a comparison and analysis on airborne SAR moving-target Doppler parameters estimation by EWVD and EWT methods respectively. The flexibility and advantage of airborne SAR moving-target Doppler parameters estimation based on EWT method are indicated and the simulating experimental results demonstrate the accuracy of our conclusion.

Information model is one of the key technologies of TMN (Telecommunication management network). This paper discusses the advantages of auto-translating information model from GDMO description to programming language description, analyzes the construction of GDMO templates, brings out an algorithm for auto-extracting key contents of information models, and puts forward three proposals for improving GDMO templates.

According to the layering and partitioning model of transport network, the relationship of the layer network (different type and same type layer networks) management systems in transport network is studied. The application of the research result for ATM transport network management is studied in detail.

In this paper, a multi-station cyclic service system with priority, limited capacity is discussed. Including: analysis of the construction and performance characteristics of the system, the characteristic of the queue state transition, the algorithm of the queue state transition probability, the unconventional algorithm of queue state probability and other algorithms of the performance characteristic parameters of the system and queue. Some new concepts such as the state conditional transition and the state transition combination and some methods such as the inverse-solution of replacing the distribution parameter with the lumped one, etc. are presented in this paper.

As a winning approach in PSTN, IN will play an important role for IP telephony service provision. Current research efforts focus on interworking of IN and Internet via gateway. The problem of this solution is that it inherits the drawbacks of conventional IN. The main contribution of this article is to redefine the IN conceptual model and presents a distributed architecture to make it fit IP telephony platform. The possibility to enhance gatekeeper to be SSP is highlighted. To show that this solution is practical, an example of service implementation is given. Compared with traditional architecture, MA-based service architecture has many advantages in distributed control, service customization and flexibility. Besides, this approach can combine legacy systems and get a smooth evolution.

An improvement upon public key certification system, based on the work done by Bruno Crispo(1996) and F. Bergadano(1997), is proposed in this paper, which can solve the security problem of the existing public key certification system. Not only the attack from outside of system but also the fraud of system inner can be resisted by the improved scheme.

A subspace-based time delay sequential estimation algorithm named MUSIC-SEA is presented for asynchronous bandlimited DS-CDMA systems in multipath fading channels. The sequential estimation idea is introduced, and the knowledge of the estimated time delays is involved in the subsequent estimation procedures. The whole procedure can be described as follows: the noise subspace is estimated first, and then coarse estimation and fine search of time delay are implemented for every time delay sequentially. The iterative steps are presented, the performance of the algorithm is analyzed, and simulations are carried out. The results show that the algorithm can mitigate the interference of multipath and provide time delay estimation with high probabilities and high accuracy.

An algorithm for time delay estimation in asynchronous band-limited DS-CDMA systems is presented. The algorithm is based on subspace concept. The noise subspace is estimated first, and then coarse estimation and fine search of time delay are implemented sequentially. The performance of the algorithm is analyzed, and simulations are carried out. The results show that the algorithm is near-far resistant and robust to the number of users partly, the estimation accuracy is improved greatly, and the required computation is reduced.

The roles of hot electrons and holes in dielectric breakdown of the thin gate oxide have been quantitatively investigated in the paper by separately controlling the amount of injected hot electrons and holes with Substrate Hot Holes(SHH) injection method. The results shows that the cooperation of hot electrons and holes is essential for the Time Dependent Dielectric Breakdown(TDDB) in thin gate oxides and thus a new physical model is presented.

Frequency Modulated Continus Wave(FM-CW) is widely used in pulse compression radars and the matched filtering technique is its main processing method. But for large time-band production systems and arbitrary FM-CW, how to design a matched filter is a difficult task. This paper proposes a general matched filtering processing method, which avoids the difficulty of designing matched filter and has practical applications.

This paper presents morphological neural networks of opening and closing operation for pratical use, and the algorithm to design optimal parameters of a morphological filter, Experimental results show that this method is good in practice and easy to extend. It has better filtering properties than that of the conventional morphological ones. The task of creating a morphological filter can be divided into two basic problems, selecting a morphological operation and Structuring Element (SE). The set of morpholo...

Weighted method of data fusion is a simple and effective method, and the weighted coefficient determines the effect of fusion. In this paper, an entropy function is provided to measure fusion information, which relates the uncertainty of data, the consensus among sensors and the performance parameters of every sensor. Then a weighted fusion method based on entropy function is developed. Considering the practical IR/mmW multisensor system , a parley temporal-spatial group decision-making fusion algorithm can...

Singularity often carries the most important information of signal, so it is valuable to study the singularity of signal. Because of the limitation of classical approaches (such as Fourier transform) in processing the singularity of signal, this paper applies multifractal theory to analyse the singularity of signal, and presents an algorithm and some fundamental results which indicate that multifactal theory is really a useful tool for analysing signal singularity.

A general and efficient technique is presented for deriving the closed-form spatial Green s functions of 3-D Poisson equations in multilayered media. Based on the spectral-domain equivalent transmission line model of the layered Green s function, an efficient Krylov subs pace order-reduction technique is used to get, the rational approximation of the Green s function in the spectral-domain, then the spatial-domain Green s function is acquired directly via the residual theorem. The derived Green s function c...

In this paper, the effect of radome on antenna radiation with the spectral-domain approach is analyzed. Radiated pattern of antenna with radome is calculated, and the conclusion is got that the experimental results agree with the calculated. The quantitative analysis of the effect of radome on antenna radiated pattern is given and provide valuable basis for the design of antenna and radome.

The Diamond-Like Carbon (DLC) film and Nitrogen-doped DLC(N-DLC) film are deposited on molybdenum coated ceramic substrate by using the pulsed laser deposition technique. The microstructure and morphology of the films are examined using X-ray diffraction, Raman spectroscopy and scanning electron microscopy. Field emission measurements are carried out by using the diode structure. The field emission measurements indicate that nitrogen doping can lower the turn-on field and increase the current density.