This paper first gives a SCP abstract model, then SCP s overload detection and maximum processing capability are discussed quantitatively. Based upon dynamic adjustment, a new two-level SCP overload control algorithm is proposed. Theoretical analysis and simulation prove the algorithm s effectiveness and fairness.

This paper presents several protocols without receiver collisions for WDM networks with tunable transmitter/tunable receiver architecture. These protocols include the simple non-receiver -collision protocol (called TS), the optimized scheduling (called ETS) and the modified ETS (called METS). Computer simulation indicates that these algorithms achieve extremely good performance. They are the efficient media access protocols for WDM networks. Moreover, the METS algorithm shows more outstanding performance with the increasing of the tuning time of optical device. It is with great practical value for the WDM network with arbitrary transceiver tuning latencies.

A single sever queueing system with multiple discrete periodic train arrivals and definite service time, which cames from a worst traffic status in intermediate nodes of ATM networks and nodes where wired ATM networks access wireless ones under CBR traffic, is analyzed in the paper, and the formula for its queue length residual distribution is derived.

The queuing models of three typical implementation schemes of selfhealing function in VP switches for backup-VP selfhealing algorithm are given. Computer simulations are made, and detailed analyses of the effect of implementation schemes of selfhealing function and architecture parameters in VP switches on selfhealing performance are presented. Analytic results are helpful to choosing reasonable implementation scheme of selfhealing function, architecture parameters and the number of backup-VP in VP switches for the best selfhealing performance.

This article deals with the reverse channel of IS-95 based cellular CDMA system. The M-ary OQPSK(Offset Quadrature Phase Shift Keying) modulation process and the RAKE receiver are described. Bit error rate over Nakagami(m) multi-path fading channel is derived and the result is used to investigate the influence of different parameters to the overall system performance.

In this paper, the continuous-time systems of the random access channel are analyzed by using the average cycle method. Analytic formulae for mean values of a successful period and a colliding period and an idle period are derived. Also, the upper bounds on the throughput of the system with capture effect and packet conflict resolution are provided. Finally, the simulation results of the random access channel are given.

Chaotic system is extremely unstable. Can it be used for measuring signal? This paper proves that it can be either theoretically or experimentally. This measuring method is completely different from one of the linear system. Acconding to the analysis of chaotic theory, it has extremely high sensitivity and resolving power. If this principle can be used, the study of measurement and nonlinear applied science would be break through.

This paper presents a new fuzzy K-nearest neighbor codebook design algorithm of vector quantization, the algorithm can eliminate the effect of initial codebook selection on the quality of clustering, is not trapped in local minimum, has a good convergence rate, and can get the codebook with good performance. Simulation results show both the convergence rate and PSNR of our method are significantly improved than that of fuzzy vector quantization algorithm presented by N.B. Karayannis, et al in 1995.

In order to implement the symbol-logic-based neural network system, a model of logical neuron is denned in this paper. The logical neuron circuits and all kinds of logic -function circuits have been realized by using current-mode CMOS VLSI implementation. Finally, the results of circuits simulations are verified.

Bidirectional associativ memory models with axonal signal transmission delays are two-layer heteroassociative networks. In this paper, existence and uniqueness of periodic solutions are emphatically discussed for BAM with axonal signal transmission delays. A sufficient condition is obtained. Correctness for obtained results are verified by use of some examples. The obtained results have improtant leading significance in the design and application of oscillation neural circuits for BAM with axonal signal transmission delays.

In this paper, the global asymptotic stability of equilibrium point for asymmetric Hopfield neural networks with time-delay is discussed, and the sufficient conditions for global asymptotic stability of the equilibrium point are obtained. Finally, the applications of these results in the design and the parameter selection of neural networks are illustrated.

In this paper, a method, which uses the new hybrid Pi-Sigma neural network to realize the Sugeno-Takagi fuzzy system, is presented. And then it is used for modelling the data association computation problem. Experimental results show that this method is a new one for data association modelling.

This paper presents a data fusion method of multiple hypotheses testing with distributed detection system. The data fusion algorithm is expanded to more general case. This paper describes the data fusion algorithm, the decision rule of system, the calculation equation of system performance, and givs the simulation results of system performance.

The paper analyses the improvement of track loss in clutter with multisensor data fusion. By adetemination of the transition probability desity function for the fusion prediction error, one can study the mechanism of track loss analytically. For nearest- neighbor association algorithm, we study the fusion tracking performance parameters,such as mean time to lose fusion track and the fraction of lost fusion track initiation,respectively. A comparison of the results obtained with the case of a single sensor is presented. These results show that the fusion tracks of multisensor reduce the possibility of track loss and improve the tracking performance. The analysis is of great importance for further understanding the action of data fusion.

This paper presents fuzzy synthetic track correlation algorithms by using the synthetic analysis method in modern mathematics. In this paper, the computation and evaluation methods of the fuzzy synthetic measure of similarity between two state estimation vectors are discussed in detail, the recursive forms of three typical fuzzy synthetic functions are derived, the fuzzy synthetic track correlation criterion is described as well. Moreover, the algorithm is compared with two classical methods through simulation. The simulation results show that the performance of the fuzzy synthetic track correlation algorithm is much better than that of the classical methods in dense multitarget environments, more cross, split and maneuvering track situations. Under above situations, its correct correlation rate is improved about 40 percent over that of the classical methods.

A novel neural network model---Linear Interpolation Neural Network(LINN) has been presented, which is used for radar target identification. And the 1-D range profiles of targets are used as identification feature. It is well known that the 1-D range profile reflects the precise geometric structure feature of a target, but it varies with the pose of the target. The LINN utilizes just the variation information of the 1-D range profile with the pose to improve the identification performance of targets in any posture.

A nvoel method,time-domain variational iterative method(TVIM),for microwave imaging of two-dimensional lossless electromagnetic target buried in half space using time-domain scattering data is presented in this paper.It is an iterative method derived using volume equivalence principle,variation principle and Fourier transform.The direct problem is sovled by finite-difference time-domain(FDTD) method and the inverse problem is solved iteratively.Some typical results are given to show its performance,such as converging property,ability to invert arbitrary targets,tolerance of random noise,etc.Comparision with the former methods is also done.Those factors which make TVIM performs better are also investigated.

The scattering characteristics of Ⅱ- Ⅵ semiconductor materials with tensor conductivity resulting of the Hall-effect,filled in waveguide with gaps are analyzed with 3-D edge-element.Some useful curves are given and the procedure of determining mobility and carrier concentration of Ⅱ-Ⅵ semiconductor materials with these curves is described.Since this method starts from the functional variation directly,it avoids the difficulty,met in the other methods,of solving the eigenvalue problem for very thin lossy isotropic dielectric loaded waveguide.The experimental results confirm the effectiveness,reliability and accuracy of the approach.

This paper deals with the profile reconstruction of the Array induction Imager Tool(AIT) measurement using the Distorted Born Iterative Method(DBIM) and an efficient Numerical Mode-Matching(NMM) method for the conductivity distribution in a two-dimensional axisymmetric medium.In each iteration of DBIM,the NMM method is appplied to solve a forward solver.Meanwhile,the analytic expression of the integration in the nonlinear integral equation has been derived and then the inversion of the efficiency and accuracy is improved dramatically.Several examples have shown that this method is practicable for the interpetation of electrical logging data.

A new coding scheme for medical images is presented in this paper on the basis of discussion on the similarity between the radiological images of sequential slices of the same human organs or of the same positions of different human bodies. The experimental results show that the scheme is correct.

The automatic segmentation of ultrasonic heart image using self-creating and organizing neural network has been studied. This kind of unsupervised clustering method can search for the optimal number of output nodes automatically to get the number of textures in the image, and finish the automatic segmentation. Experimental results show that this method has significant benefits over self-organizing neural network method.

In this paper, the Cauchy function is taken as a new target function of neural network accordings to the robustness theorem of statistics. Under the same network parameter conditions the BP net is trained using both mean squresand Cauchy target function firstly, then the net is tested by data sets including small Gaussian noises and outliers separately. Simulation results indicate that the network has both faster convergence speed and better performance against outliers after learning with robust target function.

In this paper, a timing-driven placement algorithm using Kohonen self-organizing is proposed. The object function of the algorithm is to minimize the critical nets. In the algorithm, net-based similarity vector is presented. Compared with the cell-based similarity vector, net-based similarity vector not only can directly deal with the multi-terminal nets, but also can describe the information about timing. The experimental result shows that it is a effective method.

In this paper, theoretical analyses and simulations are carried out to show that there are signal cancellation phenomena in the DED method proposed by Kumar (1996) for wide-band array processing in unknown noise, and new method is suggested to perform direction of arrival (DOA) estimation in unknown noise. In the new method, a wide-band focusing technique without initial estimate of the DOAs is employed, and the frequency-direction ambiguity in wide-band processing is avoided. Monte Carlo tests are carried out for comparing the performance of the new method with that of the DED method.

The BDD (Binary Decision Diagram) is very important for representing synchronous circuits. After analyzing and reducing the BDD, the state traversing is proposed on the basis of collapsing of input, routes and states on STG. Finally, the verification for the non-reset circuits has been described.