This paper proposes a hybrid error control scheme suitable for wireless ATM systems, that is, voice telephone is protected by pure RS code, video and data are protected by truncated RS/Type II hybrid ARQ protocol. The performance of this hybrid error control scheme over bursty channels is analyzed and simulated. It is found that, by restricting numbers of retransmission, low transmission delay and cell loss rate can be obtained.

Wavelet analysis has being gradually applied to communications, especially, wavelet modulation receives much considerations for its bandwidth efficiency. This paper discusses how to design optimal orthogonal wavelet and scaling functions which are generated by FIR filters, from the view point of bandwidth efficiency, the designing method and results are given in details. By comparison, the bandwidth efficiency of wavelet modulation has improved significantly for the same system complexity after optimization, so relatively good spectral performance using simple FIR filters can be obtained.

A new dynamic code allocation CDMA (DCDMA) protocol,which can support voice and data services simultaneously,is investigated in this paper. In DCDMA protocol, a voice terminal will reserve a spreading code only in talkspurt. The base station assigns a code after successfully receiving a request packet through a special request channel at the beginning of each talkspurt. Unlike a voice terminal,a data terminal must contend a spreading code for each packet transmission. So the code utilization can be significantly improved in DCDMA system. An analytical model and computer simulation is performed for evaluating the performance of the DCDMA protocol. The results show that the DCDMA protocol is very effective in utilizing code resource, and the QOS requirement on voice calls is still satisfied while the number of data terminal is increased.

A key escrow cryptosystem can provide protection for user s privacy, while at the same time, allows for the wiretapping when lawfully authorized. In this paper, a key escrow system based on RSA algorithm for network environment is given. A trusted center in this system specifies n trustees for every user, and distributes the part recovery ability of the user 's secret key to n trustees by (t,n) threshold scheme. When the user broadcasts his encrypted key, only if at least t cooperated trustees can enforce the wiretapping to this user.

This paper is based on the concept system for D=2 SM array code proposed by the author in 1994. The coding rule and existence condition for D = 3 SM array code are given, and its structure property, construction step, design analysis and some use are also discussed.

A enumerative formula of Boolean functions with the first-order correlation-immunity, which are important in stream ciphers, is given with help of column-balance matrixes, inclusion-exclusioin principle in combinatorics and edge-induced subgraphs in graph theory.

In this paper, a new method based the fuzzy pattern recognition and fuzzy cluster techniques for the evaluation of communication equipment/system is proposed. Contrast with fuzzy comprehensive evaluation, the new method is effective for practical applications while it keeps the advantages of fuzzy comprehensive evaluation.

The detected features evolution in scale-space is an important problem in multiscale analysis. With the 2-D Gaussian function as the kernel, the detected edge of an acute corner in scale-space, that is an ellipse determined by the slope of the corner boundary and the analysis scale, is given out clearly in this paper. Based on this, the assimilative effect of the scale factor to the detected edges of adjacent corners is discussed and the inequality of the distinguishable resolution of the corners in scale-space with the scale factor is given. The results given in this paper can be treated as the general reference for the multiscale analysis.

A general difficulty of using singular value decomposition (SVD) to split signal and noise subspaces is in the right choice of effective rank. The commonly used method toward this end is to use a fixed threshold. Yet despite its apparent physical significance, the lack of adaptability has strongly limited the popularity of subspace approach in line enhancement. In order to surmount this drawback, a cluster method based algorithm for determining the effective rank is proposed in accordance with the a priori information obtained from the time-frequency distribution of the noisy sinusoids. Simulation results show that the methodology advocated is effective for solving a class of multiple line enhancement problems.

To meet the requirements with large-scale neural networks for real-world applications, an inverse way of constructing objective functions was proposed in this paper, which translates the task of constructing objective functions into the design of error signals. Followed this way, a set of objective functions has been given as examples to eliminate the false saturation in Mean Squared Error (MSE) and overspecialization in Cross Entropy (CE). The verification of its power was also made by the comparison with MSE and CE in the tasks of estimating the scaled likelihood for the Hidden Markov Models' states in the Hybrid HMM/ANN models, and showed consistent advantages with the theoretical expectations.

A new blind beamformer with improved Hopfield network is presented in this paper. To estimate the steering vector, the signal property of cyclostationary is used. To suppress interference, traditional LCMV beamforming is employed. To improve the robustness, diagonal loading technique is exploited. To turn into realization in real time, a neural network structure is given. Simulations demonstrate the excellent performance of the proposed approach in a wide variety of situations.

Based on the scale and the state of intensity distribution in some neighborhoods of edge points, this paper proposes a measure vector of importance of point that consists of 4 components and corresponds to every point. For considering the background of application, this paper first trains a BP neural network using some samples that have classified by manual work, and then extracts more important edge points in a new image using the trained neural network. Because the image needs not be smoothed by some function in this algorithm, the edge deflection will not happen as usual, the location of gotten edge is in the accurate position. The effectiveness of this algorithm has been testified by some experiments.

This paper analyses further global asympototic stability of a class of delayed cellular neural networks by means of Lyapunov functional method and new inequality a²b≤(2a³+b³)/3, (a,b≥0) analysis technique, and some new sufficient criteria are obtained. These criteria are of theoretical and applicable important significance in the design of globally stable networks.

This paper presents a high-resolution DOA tracking algorithm based on phased array with a single channel receiver. It is an improved algorithm of the method using specific weights to realize high-resolution direction finding with single channel receiver. When signals are uncorrelated, by using the specific weights in turn and refreshing the power measurements continuously, the time-changed DOA of signals can be traced. The computer simulation results show that the algorithm is effective.

After translational motion compensation, rotational motion compensation is needed in ISAR technique. This paper considers the rotational motion compensation in the case of that the target moves along a straight trajectory with constant speed. It is found out that the nonlinear factor γ is most important parameter in its rotational motion compensation. In order to estimate γ, image entropy minimization principle is described here. Based on this principle, the unknown γ can be estimated accurately. The effecttiveness of this method are shown by simulated and experimental results.

The echo signal of a high-range-resolution radar(HRR) contains much information about target. By using this information and feature extraction method fully and properly, an efficient target recognition can be executed. A novel approach for radar target recognition is proposed in this paper. This approach explores canonical analysis on a matrix formed by the image vectors of training targets in different aspect angles. One image vector contains both the amplitudes and relative phases of the range profile of a target. A subspace is obtained from this analysis. Projection of an image vector into this subspace forms subimage. The subimages of a training target in different aspect angles are averaged into library feature vector for this target. Using the subimage of an unknown target as feature vector and minimum distance rule for target recognition, experiments on simulated data are done.

Phase compensation is a critical step of Synthetic Aperture Radar(SAR) and Inverse Synthetic Aperture Radar(ISAR) imaging technique. This paper proposes a progressive approach algorithm based on image cost function to compensate the random phase error in SAR/ISAR imagery. Four kinds of image cost function are discussed. The comparison of the proposed function is performed by analyzing the simulation and experimental results. It is found out that Shannon entropy function is the best one. Based on it, the interference phase in unknown signal can be estimated accurately.

Multi-fractal methods are introduced to radar clutters analysis in this paper to find general characteristics of clutters. Clutters and echoes from targets are analyzed respectively. The results show that the radar clutters are a kind of generalized frctal signals in some degree, and echoes from targets do not match with the fractal model. This fact leads to a new way to detect targets.

This paper studies on the false color synthetic processing for polarimetric SAR data. A new method to get false color image of polarimetric SAR data with the unitary Stokes matrix is presented. Because this method discards the information of total power, it can be used to analyze the target with more power or less power and the target with the same shape in polarization signatures. The paper also presents the way to find three optimum polarization s combinations to enhance the contrast between two types of scatters with polarization signature and use the polarization synthesis to get false color image of polarimetric SAR data.

A method of signal extraction in time domain is proposed in this paper, by means of which, the time-domain incident wave and reflected wave on the reference plane of the feeder can be extracted from the total-field responses obtained from the time-domain simulation for the microwave components fed by the nondispersive transmission lines, furthermore, its frequency-domain parameters can also be acquired. By comparison with traditional methods, it can avoid to directly compute the time-domain incident wave and reflected wave using the finite-difference time-domain method, so computation time and storage can be reduced. The numerical simulations of rectangular microstrip antenna show that the computation precision of this method is the same as that of traditional methods.

The spectral techniques is mainly used in the analysis of digital circuits. This paper studies its application in the synthesis of digital circuits. In this paper, the spectral function is realized by using bidirectional current-mode CMOS circuits, and it provides a new technique method for the design of digital circuits.

A new smart antenna for CDMA system was proposed in the paper. The beam pattern with approximately equal side lobe is adopted. The weighting vectors are got not by measuring or rectifying in advance, they can be got by average method, and can be realized easily in engineering. The enhanced capability of this antenna is approximately equal to adaptive antenna, but need not iterative procedure, has fast response and good robustness.

In this paper, a new approach based on genetic algorithm (GA) is applied to adaptive array, in according to the limitation of hardware, the amplitude weights are not changed, but the phase weights are optimized globally with genetic algorithm. For reducing the perturbations to the far field pattern and placing multiple deep nulls real-time in the directions of interference, a genetic algorithm adjusts some of the least significant bits of the beam steering phase shifters. Simulation results confirm the theoretical work.

To study the influence of rough sea surface on the space field of HF ground wave radar on coast, a derivation of the reflection is introduced first, which is suitable for rough sea surface with Gaussian distribution of heights, and then the reflection is substituted into the radar equation, finally, the propagation loss over rough sea surface is calculated. The resulting predictions in this paper are more accurate and realistic.

This paper introduces the construction of the ultra-high speed waveform digitized system based on laser probe. The functions of different stages of the high-speed dynamic divider circuit chip are measured with the double-frequency phase sweeping technique. A detail analysis about the chip failure and the characteristics of laser probe measurement is given.

In DS/CDMA communication system, users share time and frequency resourses, it will cause Multiple Access Interference (MAI). The conventional detector follows a single-user detection strategy can not reject MAI. In this paper many important technologies to reject AIAI are described and analyzed.

This paper systematically summarizes the merits, defects and perspectives of the neural network approaches being able to solve the physical design tasks, analyzes the present research situation of applying neural networks to physical design, proposes some approaches for mapping physical design problems to neural networks, and presents some methods to raise running speed of the neural network based the algorithm, which is executed in sequential computer.