The localization of multiple current dipoles is a difficult problem for the inverse study of MEG. A method of localization of multiple current dipoles based on the minimum-dipole number solutions is investigated. The revised simulated annealing algorithm as the optimization strategy is used in the simulation experiment. The result shows that the method is efficient to the localization of multiple current dipoles.

A fast algorithm is proposed for tracking low observable small targets in a digital image sequence. Wavelet transform is used for image filterting to improve SNR. Fixed sample size hypothesis testing is applied to generate tentative tracks which avoids target tracks missing. Target trajectories are confirmed by truncated sequential probability ratio test. The simulation results show that it can detect and track small targets efficiently with SNR<2dB.

A simple and efficient method to design adaptive prefilters based on minimizing the high pass energy of the signal multiwavelet decomposition is presented in this paper. This method can be generalized to 2-dimensional images by processing rows and columns respectively. In this way adaptive prefiltering can be performed for images so that the reconstructed image quality is greatly improved. The experimental result shows that the adaptive prefilter given in this paper is superior to the other prefilters available.

A model for the decomposition of polarimetric SAR data is developed in this paper. The model is assumed to consist of four backscatters: Bragg scatter, double-bounce scatter, odd-bounce scatter and volume scatter. This scattering model can successfully describe the mechanism of man-made and nature targets to scatter the incident electromagnetic wave. Using NASA/Jet Propulsion Laboratory (JPL) AirSAR data, the model has been tested, the result shows that, the model can describe the mechanism of polarimetric backscatter from buildings and vegetated areas quite well.

A new method for computing high-frequency RCS of objects with complex configuration, called. Curved Surface Pixel (CSP) method, is presented in this paper. The distinction between CSP method and GRECO method is described. Several different approaches of the pixel methods are analyzed. Many computation examples for spheroid, ellipsoid, cone, missile mode and aircraft mode show that the CSP method has advanced computation accuracy.

Multisensor data fusion using Derapster-Shafer evidential reasoning is based on information s independence, but it is not true in many practical situations. This paper describes a multisensor data fusion method based on Dempster-Shafer evidential reasoning, and also gives a generalized Dempster-Shafer theory of evidence which is efficient in dealing with dependent, information. A test is tried out for object assortment based on the fusion of information from force and thermal sensors of the robot.

In this paper, the structure and algorithm of Max-Min fuzzy neural network (MMNN) are studied in detail. In order to get rid of some intrinsic localization of the method and boost up the capability of the MMNN, a series of steps are presented and the improved project (IMMNN) is gained. With a view to making the capability even much better and compressing the time of the convergence, the op-IMMNN is put forward in which the parameters of IMMNN are optimized by genetic algorithm combined with simulated annealing. In the simula- tion, the result of op-IMMNN is superior over the conventional MMNN’s. Finally, a satisfactory result is also obtained when op-IMMNN is regarded as a classifier to distinguish the types of the ships according to their actual radiated noise. Comparing with the neural network based on the back propagation algorithm, the advantages of the op-IMMNN are fully put up.

There are two main types of approaches in the research of fault tolerance of Multilayer Perceptrons(MLP): improvement in the learning algorithm and component redundancy after training. A dynamic redundancy BP algorithm is presented. In the training steps of the conventional adaptive BP algorithm with a momentum term, the most important weights are replicated based on their "significance". Applying the algorithm the fault tolerance of a network can be improved effectively. Compared with fault injection while training--a typical improved learning algorithm, although this dynamic redundancy algorithm gives no prominence in fault tolerance, the training time can be greatly reduced.

An approach is introduced to voice detection that differs from normal approach via energy, correlation and zerocrossing criteria. By measuring the gross shape of the short-term speech spectrum using spectral entropy to detect voice segment, it is shown that the spectral entropy can be used effectively even in heavy background noise. The simulation results show that the approach via spectral entropy has good performance for anti-noice.

Constellation design has great influence on the performance of satellite positioning system, the constellation should provide multi-coverage and satisfy certain geometry. First, this paper presents coverage property of satellites and DOP factors, based upon these, researches for regional satellite positioning system and relevant estimates of positioning performance are provided. Finally, the application of and some questions about regional positioning constellation are explored, and conclusion is also given.

Ethernet becomes a very popular technology because of its simple, mature, low cost, and good performance. But its capture effect will cause large delay jitter. How to solve this problem has become a new project. In this paper , a new adaptive arbitration algorithm is presented, in order to handle the capture effect and to improve the ethernet performance.

In this paper, A barrier function descent direction method for implementing CDMA multi-user detection (BMD) is proposed. As compared to conventional detectors and neural network optimal algorithms, the method derives a new descent direction from the barrier function other than the gradient direction. Numerical results show that the method is always able to find the global minimization point. It has many advantages featuring near-far effect resistance, able to suppressing multiple access interference and channel noise, etc. The method paves a new way for realizing optimal multi-user detection in CDMA system.

In this paper, a heuristic approach for constrained via minimization for VLSI three-layer routing on the basis of segment-crossing graph model is presented. Two types of algorithm-global sense and local sense are combined into the new algorithm. In addition, many physical constraints are taken into consideration. According to the results, this algorithm is verified to be very efficient and encouraging.

In this paper, the pilot-channel-assisted QPSK in direct-sequence code-division multiple-access (DS-CDMA) system is investigated. The performance loss due to imperfect channel estimation is calculated. Then the data-to-pilot channel power ratio is optimized analytically and the probability is presented. Finally, simulation results are given.

In this paper,an algorithm of wavelet transform without multiplication is introduced. It can implement 2D separable wavelet transform rapidly. The speed of transform is greatly improved comparing to the classical float algorithm. It is also easy to be implemented in hardware. This algorithm will have broad application potentials in image compression.

Current-mode multifunction biquardratic niters based on current controlled conveyor (CCCII) with two grounded capacitors is proposed. The circuit offers the following advantageous features: realization of different biquadratic filter signals from the same configuration employment of only two grounded capacitors which are ideal for integration, orthogonal control of ω₀ and Q, and low sensitivities.

The classical Kirchhoff formula in the iterative physical optics (IPO) approach, which was brought forward by Obelleiro(1995), has been replaced by a Kirchhoff formula with line integral form. Based on the new method, an improved IPO approach has been presented. The numerical precision of computation results obtained by the improved IPO approach is almost the same as that of the results computed by IPO approach in Obelleiro s paper(1995). However, the new approach reduces the computation time greatly in analyzing the electroniag- netic scattering of large open-ended concave cavities such as a trihedral corner reflector and an aircraft cockpit, etc.