In recent years, M-band orthonormal wavelet bases, due to their good characteristics, have attracted attention. The ability of 2-band wavelet packets to decompose high frequency channels can be employed to improve the properpty of wavelet for time-frequency localization, which makes more kinds of signals for analyzing by wavelet. Similar to the notations from the extention of 2-band wavelets to 2-band wavelet packets, the theory framework of M-band wavelet packets is developed, a generalization of the notations and properties of 2-band wavelet packets to that of M-band wavelet packets is made and the corresponding proofs are given.

Wavelet transform, especially the multiresolution representation, is a very effective tool for analyzing the information contents of a signal. Based on Mallat s multiresolution analysis, this paper discusses the theoretical analysis of M-band multiresolution signal decomposition, proposes a new algorithm for realizing the theory, studies the properties of an operator which approximates a signal at a given resolution, and applies the theory to images. The results show that, first, images can be decomposed and reconstructed by M-band multiresolution representation; second, the edges of image can be detected by M-band wavelet transform.

A novel method to estimate DOAs of coherent sources impinging on an uniform circular array(UCA) is presented in this paper. A virtual uniform linear array(VULA) is first derived by DFT preprocessing, transforming the UCA from element space to phase mode space which is special for circular arrays, and then the well-known spatial smoothing technique is applied to the VULA so that the decreased rank of covariance matrix due to coherence can be retrieved. Simulation results have strongly verified the effectiveness of the algorithm.

Estimation of model parameter for transient signal is very important in many aspects. This paper presents a new Markov ARMA model Q-slice estimation algorithm of transient signal based on bispectrum. Simulation results show that this new method has some special features, such as higher estimation precision, lower amount of calculation, higher fitting effect even in lower signal-to-noise ratio (SNR) situation.

The paper provides the high dimensional optimal problem from which the RELAX algorithm is derived, and compares the RELAX algorithm with the alternating projection algorithm from some aspects. Based on the discussion, an improved RELAX algorithm is proposed. The computer simulation confirms the theory is validity.

Some shortcomings of common fractal image coding methods are studied, then they are corrected with a new method. The new method is improved further in DCT domain. Coding results show the advantage of the new method.

Based on the Hopfield network a new method of binarization processing for blurring edge image is proposed. First, the binarization processing problem is turned into the optimization problem, then correspondent Hopfield network model and its parameters are designed. Experiments show that this method is of advantage for binarization processing the blurring edge image.

In this paper, by using the biorthogonal quadrature niters, the biorthogonal mul-tiresolution of finite dimension space equipped with inner product and the fast discrete wavelet transform (FDWT) are constructed. The dual transform method is proposed and the radar data storage is reduced by it. The method of choosing the wavelet coefficients, and the methods of correlation and nearest neighbor classification in wavelet domain based on the compressed data, are presented. The experimental results of the classification, using the high resolution range returns from six kinds of aircraft, show that the methods of transform, compression and recognition are efficient.

Sparse-array Synthetic Impulse and Aperture Radar (SIAR) can isotropically radiate by employing multiple frequencies (synthetic pulse) and multiple antennas (synthetic antenna). The range resolution only depends on bandwidth of transmitting signals. The distance grating lobes emerge in the case of increasing time-bandwidth product of transmitting signals. The performance of pulse compression is analyzed with the transmitting signals modulated by phase-coded sequences. The great ratio of pulse compression and range sidelobe suppression in SIAR are obtained, and its effective range and range resolution are raised.

In this paper, a new maximum likelihood (ML) classification algorithm is proposed to classify the multi-look polarimetric synthetic aperture radar (SAR) imagery. Experimental results with the NASA/JPL airborne L-band 4-look polarimetric SAR data demonstrate the effectiveness of the new algorithm. Furthermore, when using the algorithm in the classifications with subsets of the multi-look polarimetric SAR data, the polarization-channel optimization for the terrain type classification is implemented.

With a multiplicative speckle model, the unconditional density function of the mutli-look polarimetric synthetic aperture radar (SAR) data is derived, and the result is shown to be a generalized K-distribution. The multi-look intensity K-distribution is particularly compared with histograms of physical radar measurements of several typical terrain types and several empirical distributions. The relationship between the heterogeneity coefficient of the scene texture and the proper statistical distribution model is experimentally established. In addition, based on the results of the statistical analysis, an adaptive classification scheme is presented, and the improved classification shows the importance of the statistical analysis.

In this paper, the signal decomposition and synthesis technique based on FIR filters have been studied, the different data extension methods included in convolution procession have been investigated in detail, the problem involved in conventional symmetric extension method has been reported, the new symmetric extension method has been given. The experimental results show the new method performs better than the conventional methods.

Based on the generalized open-closing and close-opening operations in mathematical morphology, a class of generalized parallel complex morphological filters is constructed by using multiple structuring elements in this paper. These filters possess some important properties such as translation invariance, increasing, duality and idempotence and obey the threshold superposition rule. They can not only effeciently suppress noise in images but also preserve the geometrical features of images. The results of computer simulation show that the new filtering method is quite effective.

The requirement of electromagnetic compatibility (EMC) is a very important network performance index for organizing and planning mobile telecommunication network. The paper put emphasis on researching the ideal network topographical structure which could ensure the electromagnetic compatibility within a plane mobile telecommunication network, analyzing and researching how to consider the selection of the number n of frequency group in network cluster and the modules M of network trellis under the requiremen...

The performance of leaky bucket policer with priority management (PLB) in ATM networks is ananlyzed using extended fluid flow method. This PLB scheme introduces PBS selective discarding priority management into service policing. It guranties lower loss probability for high priority cells. The analysis and numerical experiments show that this scheme is useful. By the performance analysis, suitable PLB parameters for efficacious control may be obtained.

In this paper the bifurcation, chaos and synchronization of the modified Chua s circuit are experimentally studied. On the basis of this research the chaotic secure communication has been successfully demostrated.

In this paper an optimum design method for high gain microstrip antenna covered with dielectric layers is proposed. In the method, the moment method and optimization for the gain of a point source model are used to find the optimum thickness for every dielectric layer, so that the highest antenna gain can be acquired. Besides, the use of point source model much decreased the computation needed in the method. The calculated results also show that the high gain microstrip antennas covered with dielectric layers can still have rather high radiation efficiency.

The finite element method is combined with the mode-matching method and the multi-mode network theory to analyze the scattering and radiation characteristics of a class of planar dielectric waveguide discontinuities. Unlike the conventional method to treat the radiation as a "source-field" problem, in the present approach, the dispersion characteristics of dielectric guidedwave structures are calculated first, and then the radiation problem is transferred to the propagation problem of a series of surface-waves and space waves from the viewpoint of scattering, thus the analysis is tremendously simplified.

An efficient method to calculate the quasi-static capacitance of the via hole structure embedded in a multilayered dielectric media named Dimension Reduction Technique(DRT) is presented. Since the method takes full advantage of the characteristics of the stratified structure in integrated circuit, it can easily deal with the varied numbers of the dielectric layers as well as the varied structure parameters while only need little CPU time and memory space. The numerical results given in this paper are in good agreement with those of Ansoft’s software.

One of the crucial technologies for developing high resolution field emitter displays and vacuum microelectronic devices in microwave and millimeter-wave bands is the design and fabrication of field emitter arrays which can provide with focused electron beam. After having a brief discussion of two kinds of double-gated field emission arrays (DGFEA), the concentric focusing (in plane gates) and the aperture focusing (2-tier gates), presented in the paper are the design and simulation methods of the aperture focusing DGFEA, and simulation results about its emission characteristics and focusing performance. It is believed from the results that the DGFEA can produce a pretty parallel electron beam, its maximum emission current density may exceed 300 A/cm² and it meets the major requirements for developing microwave and millimeter-wave devices, and other strong beam devices.

The expansions of MVL functions in prime field based on Modulo operations are analysed and the concept of modulo relativity is proposed. Then the canonical expansions of any quarternary logic function are presented.

In this paper, the models of canonical piecewise-linear devices are constructed by the controlled-switch and controlled-source of multidimensions, which are supported by PSpice. So that, the general canonical piecewise-linear network can be analyzed with the PSpice. The models present an engineering application value. The examples are discussed in detail also.

Chakradhar, et al(1988, 1990) represent the combinational circuit as a Hopfield neural network and formulate the test generation problem as an optimization problem. In this paper, a heuristic genetic algorithms is proposed based on traditional GA and circuit topology information. The algorithm is used for energy minimization of combinaitonal circuit s neural networks.

To settle the contradiction of noise-filtering and image detail-preserving in image filtering, a combination median filter based on Bayes-decision is proposed. The main contributions are: (1) deriving the union conditional density function of a defined test vector; (2) giving an algorithm for estimating pre-probability; (3) introducing a new filter structure, which combining median and multistage median filters based on Bayes-decision.

In this paper, using the component of the impulse response of target, a time-domain bispectrum estimating algorithm based on complex target is presented, this algorithm can estimate the scattering centers of target. The experimtental data of the compression airplane model according to proportion was analyzed in term of the above algorithm, using the impulse radar experimental system which is developed by the Team of Anti-Stealth Impulse Radar, Department of Electrical Engineering, National University of Defense Technology. The validity of the algorithm was shown by the experimental results.