A multisensor distributed extended Kalman filtering algorithm is presented for nonlinear systems, in which the dynamic equations of the systems and the equations of sensor s measurements are linearized in the global estimates and global predictions respectively, and the suboptimal global estimates based on all available information can be reconstructed from the estimates computed by local sensors based solely on their own local information and transmitted to the data fusion center. An analysis of the properties of the algorithm presented here shows that the global estimate has higher precision than the local one and smaller linearization error than the existing method. Finally, an application of the algorithm to radar/IR, tracking of a maneuvering target is illustrated. Simulation results show the effectiveness of the algorithm.

Fractional Brownian motion, continuous everywhere and differentiable nowhere, offers a convenient modeling for irregular nonstationary stochastic processes with long-term dependencies and power law behavior of spectrum over wide ranges of frequencies. It shows high correlation at coarse scale and varies slightly at fine scale, which is suitable for and successful in describing and modeling natural scenes. On the other hand, man-made objects can be constructively well described by using a set of regular simple shape primitives such as line, cylinder, etc. and are free of fractal. Based on the difference, a method to discriminate man-made objects from natural scenes is provided. Experimental results verify its efficiency.

A subspace-based tracking algorithm is proposed for high-resolution DOA estimation, using the generalized singular value decomposition (GSVD) of the sample data matrices. A GSVD updating procedure is presented. With this procedure, a new approximate decomposition can be computed from previous one, with finite operations at each iteration. Combined with exponetial weighting technique, this algorithm can solve DOA estimation tracking problems efficiently.

A more efficient permutation algorithm which has less computer operation and better structure is presented here for radix-2 FFT (FHT). It can fasten the FFT and FHT efficiently when N becomes large.

P. M. Djuric, et al.(1992) researched on the segmentation of nonstationary stochastic process into piecewise stationary stochastic process by Bayesian theory, and gave a dynamic equation about the number of segments, their boundaries and AR model orders for each segment, but did not give details of solution for the equation. Because the solution for the equation is very complex, this paper investigates the solution, derives some recursive relations, simplifies the problem, saves computation time and goes further into the segmentation of nonstationary stochastic process into piecewise stationary stochastic process.

A feedback mode with M×N cells in cellular neural networks (CNN) is equivalent to a network which possesses 2M×N cells, a neighborhood of mirror-like strucutre, space-variant templates and without feedback links as well as without input templates. The stability of the feedback mode and transformations of CNN with the neighborhood of mirror-like structure are discussed.

Airborne squint mode SAR has the advantages of better flexibility and larger field in applications than the classic sidelooking mode SAR. But the imaging processing of squint mode SAR is a strict two-dimensional phenomenon. This paper compares the 2-D FFT method, fast polynomial transform(FPT) method and two 1-D FFT method for the imaging processing of squint mode SAR.

In a CDMA wireless conmunication network using the conventional signal detection, the capacity of the system is limited and the performance is degraded by the multi-access interference (MAI). The multiuser signal detection, which make use of cross-correlation information between two spread spectrum codes, can reduce or eliminate the MAI in a Gaussian channel so as to mitigate the near-far effects and increase the system capacity. This paper extends the multiuser detector to the flat Rayleigh fading CDMA environment and discusses the bit error rate of typical multiuser detection algorithms in such a environment by combining analysis and simulation. It is concluded that the multiuser detection is superior to conventional detections in the flat Rayleigh fading channels.

A synchronous CDMA communication system in a tone jamming environment is considered. The bit error rate of an anti-multiple-access interference receiver using the single-user detector is given. The receiver (A-T receiver) is of good performance of anti-interference when there are many active users in the system.

The security relationship between p-schemes and MR(p)-schemes is analysed and their security classes are given. In addition, some multisignature schemes based on MR(p)-schemes and some password authentication schemes based on unproved ElGamal signature schemes are proposed. The results may be helpful for new understanding of the security, performance and relationship of two kinds of ElGamal type signature schemes.

Based on exact penalty function, a new neural-networks for solving the L₁-norm optimization problem is proposed. In comparison with Kennedy and Chua s networks(1988), it has better properties. Based on Bandler s fault location method(1982), a new nonlinearly constrained L₁ norm problem is developed. It can be solved with less computing time through only one optimization processing. The proposed neural networks can be used to solve the analog diagnosis L1 problem. The validity of the neural networks and the fault location L₁ method are illustrated by extensive computer simulations.

Effective description for Binary Decision Diagram(BDD) has proven useful in many applications as data structure for Boolean function. Based on the BDD of standard gate, recursive appending from primary input to output for BDD construction is proposed. Several benchmark circuits is shown to verify this method.

A method of simplifying three-valued T-gate combinational networks, known as a method of splitting the truth table, is given in this paper. By using the method, the three-valued T-gate combinational networks can be simplified to minimization or near minimization, the used operating time is less, and the programme can easily be programmed and operated on a computer.

The numerical solution of Green s function for the potential in 2-D arbitrary inhomo-geneous media with axial symmetry has been given by use of efficient half-analytical, half-numerical hybrid method. Then the logging responses of various kinds of the DC resistivity log with axisym-metric excitation have been obtained by using surface integral equation method to match the boundary conditions on the electrodes of the logging sonde. Comparing the results with that obtained by other methods, one can see good p...

As the scatterer s size is comparable with the wavelength, a rigorous solution of T-matrix is used to calculate scattering from the scatterer. When scatterers are non-uniformly clustered, coherency of collective scattering from scatterers must be taken into account. Numerical simulations of polarized scattering from random clusters of spatially-oriented and non-spherical particles are developed by multiple scattering formulation of T-matrix method. Numerical results present that the polarized bistatic and back-scattering are functionally dependent on clustering and other physical paramaters.

A numerical procedure to solve the non-linear continuity equation of the photo-injection carrier in a semiconductor rectangular waveguide is derived. The 3-dimension distribution of the photo-induced carrier in a semiconductor waveguide was presented, which is available for the study of the optically controlled dielectric waveguide.

In an X-ray preionized XeF(B→X) discharge-excited laser driven by a prepulse-main pulse circuit with magnetic isolation and a pulse compressor the dependances of the laser output and conversion efficiency on gas composition, total gas pressure, circuit parameters and delay time were experimentally investigated. A considerable improvement in the laser performance was achieved. An output energy of 345 mJ with an overall efficiency of 3% was obtained from an active volume of 0.11L at a total gas pressure of 2.5×10⁵ Pa in a NF₃/Xe/Ne mixture.

The application of cellular neural networks is determined by its templates. In accordance with the application of digital cellular neural network the author proposed before, a method for the design of its templates, i. e., inequalities construcion method, is given and a supervised learning algorithm is proposed based on the relaxation method. The learning algorithm provides theoretical basis for the design of DCNN. Simulation results on examples such as connected component detection shows the effectiveness and feasibility of the algorithm.

The lower bound of the soft weight of the error pattern which is beyond the error-correction capbility is calculated. The result is used to prove the equivalence of decoding performance of Chase2 algorithm and maximum likelihood decoding algorithm when the received sequence is subject to the generalized threshold in Ref.2. A variable threshold based on the confidence of received sequence is proposed to improve the speed of the soft decision decoding of linear blocks without performance loss.

This paper first presents a systematic procedure for designing double-edge-triggered D-flip-flops, instead of intuitive design, next, the double-edge-triggered D-flip-flops in the types of I²L and TTL are designed.

With higher-order mode (TE₀₂₁ mode) of the Dielectric Resonator(DR), a parallel feedback-type Dielectric Resonator Oscillator(DRO) operating at 10.7GHz is reported. Because of the high loaded quality factor of DR, the oscillator shows excellent phase noise performance (about 88dBc/Hz at 20kHz offset).

A method for the design of continuous-time current-mode higher order filters is presented. It is based on the recursive derivation of the integrator s time constants step by step from the denominator of the current transfer function. Only a small number of current mirrors is needed using this approach. The computer simulations for a sixth-order filter reveal that the method proposed is suitable to the design of high-frequency current-mode filters.

A avalanche diode circuit was developed to generate picosecond high voltage pulse. The structure of the circuit was described. A pulse with 2.8kV amplitude, 140ps width was generated with this circuit.

When the operating frequency gets very high, the machining error may make that the resonant frequency of the cavity in gyrotron oscillator deviates the design value, or can not meet the engineering requirement. The approximate formulas considering the effects of the symmetrical metal-ring in the resonant cavity of gyrotron oscillator are derived. The resonant frequency dependence for the gyrotron oscillator is obtained and discussed. The cold measecrements show that the theoretical results are reasonable.

In recent years, the radar network systems for modern military application, such as anti-stealth tactical and strategic surveillance systems, have been investigated with great interest, and multisensor data fusion has been a hot issue. Consequently, the critical need is to understand it in concept and application and provide a common frame of reference for detection data fusion algorithms. To fill that need, in this paper, the relation between the multisensor detection problem and classical detection problem is pointed out. Discrimination information is used to analyze and compare, in principle, the various representative data fusion algorithms for detection that appeared in some open literatures, and the other possible algorithms are proposed as well. The effects of data fusion on surveillance systems are discussed with some numerical results, and it is shown that data fusion technique is of great importance to anti-stealth systems.

This paper proposes a double screen technique for display realistic images of 3-dimensional multiobject on a 2-dimensional display screen rapidly. This technique is applied to process CT data of multiobject, such as liver, tumor and vessels, the results obtained are satisfying.