This paper proposes an adaptive identification algorithm for nonminimum phase ARMA models in additive colored Gaussian noise. The model input is assumed to be an i. i. d., non-Gaussian random process. The algorithm utilizes higher-order statistics of the observed signal alone. It estimates the AR and MA parameters successively in each iteration without computing the residual time series. The stochastic gradient method is used in parameter updating. Simulation resutls show the effectiveness of the algorithm.

A novel blind estimate of direction of arrival (DOA) and Doppler frequency with unknown array manifold is proposed, employing Doppler frequency difference between a successive pulses as rotational parameter. The effectiveness of the new method is confirmed by computer simulations. Compared with the existing two-dimensional frequency estimates, the computation load of the proposed method can be saved greatly.

The group correlation properties of binary sequences is studied. And a conclusion is drawn that not only the group correlation function of a binary sequence itself is ideal, but also that of some of its subsets, when code length N is even, is ideal. Finally a general formula of the group correlation function of a binary sequence set is derived.

The least squares(LS) minimization problem constitutes the core of many real-time signal processing problems. A square root free scaled Givens rotations algorithm and its systolic architecture for the optimal RLS residual evaluation are presented in this paper. Upper bounds of the dynamic range of processing cells and the internal parameters are analyzed. Thus the wordlength can be obtained to prevent overflow and to ensure correct operations. Simulation results confirm the theoretical conclusions and the stability of the algorithm.

A new image edge detection algorithm which utilizes the wavelet transform theory is proposed. A symmetric scaling function corresponding to (c₀,c₃) = (0.05,0.05) is used to perform the dyadic wavelet transform. The authors get a multiresolution image strategy and extract edges. Experimental results show that the algorithm is adaptive and capable for dealing with an image edge.

This paper presents a tone recognizer for Mandarin speech using a combination of wavelet transforming and hiddefi Markov modeling techniques. The evaluation of pitch periods is exactly performed by a pitch detector which is based on the singularity detection of signal and multiresolution analysis with wavelet transform. The hidden Markov models with partitioned Gaussian mixtures(PGM) are used for tone recognition. In implementing the Viterbi algorithm for HMM s, a recursive relation is derived to improve th...

Detection of primitives in images is often needed in pattern recognition and machine vision. This paper proposes a method for the search of circles in images using a genetic algorithm, as an example of detection of primitives. The experimental results show that compared to the conventional template matching and Hough transform, the proposed method takes much less computational time.

Neural network BP training algorithm based on gradient descend technique may lead to entrapment in local optimum so that the network inaccurately classifies input patterns. This paper presents a hybrid training algorithm GA-QP combined to genetic algorithm with BP algorithm. Experiments show that the hybrid algorithm outperforms BP algorithm. Satisfactory experimental results are obtained.

The automatic target detector for High Resolution Radar(HRR) is considered. The detection performance of HRR is very important for target recongnition and imaging. Due to the size of the target is larger than the range resolution cell of the radar, the backscatered power of the target is recorded as a function of time. So a mapping from the function into a point of the real line or numbrical value T 0 is done. If T₀ is greater than lambda λ₀ (threshold), then H₁ is accepted, otherwise H₀ is accepted. In this paper, the method of conversion is also considered.

The chaotic secure communication method proposed by D. R. Frey (1993) is studied. The system is composed of regressive nonlinear digital filters. Several practical aspects of the communication are analyzed, such as the correlation functions, the tolerance to noise, the sensitivity to parameter mismatches and the effect of finite word length, etc. Another chaotic system is introduced to increase the security of the communication. Simulation result shows that the method is secure and easy to realize.

Trellis structures of block codes are discussed. L-section trellis structures of some BCH codes are presented. A fast maximum likelihood decoding algorithm for BCH codes is proposed correspondingly, Decoding porblem of q-ary images of q^m-ary block codes is also discussed. The direct sum partition and the associated decoding algorithms are given for the images.

As the second part of author s serial research (1988), the cipher significant and structure properties of H-Boolean functions are investigated in further by the characteristic matrix. The best updated lower bounds are found for the enumeration of H-Boolean functions.

All digital implementation of receiver is a main topic on digital communication recently. The design of interpolation filter is one of the important problems for all digital implementation of receiver. In this paper, for full response linear modulation signal, a interpolation criterion has been proposed. An interpolation formula is presented on bandwidth-limited transmission signal. For example, using the raised consine roll off function as the system pulse response, the feasiblity and effectiveness on the interpolation formula has been certified by theoretical and numerical analysis. The computer simulation result on 16-QAM signal is given.

The problem of computing the free distance of TCM signal sequence has been discussed, a new algorithm--the matrix algorithm, is proposed, and the estimation problem of state transition number required for computing the free distance of TCM signal sequence has been theoretically solved. The matrix algorithm is derived from the Viterbi algorithm, and is an implementation of Viterbi algorithm in the form of matrix. Compared with other algorithms, the matrix algorithm gains two advantages: (1) The explicit solution, and its relatively less complexity. (2) more flexible ability to the signal space distance variation. As some examples, the results of some TCM signal sequence on additive Gaussian white noise(AWGN) channel and fading channels have been presented.

This paper points out that in micro-cell situation, that is the cell s radius is between 100m and 200m, the cell s basestation can transmit the bit-synchronous information, and all the mobile phones of the cell can be bit-synchronous on the up-link according to this information. The analysis and calculation present that bit-synchronous can greatly decrease the bit-error probability.

On the basis of considering downlink power flux density limit, shadowing effect, frequency reuse in different systems and adjoining spots of same system, adopting voice activity and polarization isolation techniques, the capacity of CDMA mobile satellite communication system at different downlink power flux density limit mode, channel spread bandwidth and mobile user s quality factor of receiving system (G/T) are analysed and calculated, and compared with the ones of TDMA and FDMA. Finally, the conclusions are given.

By analyzing the threshold-jumping of Schmitt circuits, this paper indicates that the core element realizing this function in binary TTL Schmitt circuits is the differential current switch with controllable threshold value. Based on the characteristic having two kinds sigal-detection threshold in thernary TTL circuits, a ternary TTL Schmitt circuits having twice reactions of threshold-jumping is designed. The simulation with PSPICE proves that the designed circuit has ideal function of Schmitt circuits.

A methodology to synthesise continous-time current-mode filters using CMOS OTAs is presented. Second-order CMOS-OTA current fllters are systematically generated from the simulation of signal flow graph for a general biquadratic function .The non-ideal effects in high frequency compensation frequency characteristic of OTA-C filters are analyzed. A simple high off frequency method is discussed. Finally, A 4th-order Butterworth lowpasss filter with 1MHz cut and simulated with PSPICE program.

Using the fluid model and making the plasma as isotropic medium, the microwave field in a cylindrical cavity is obtained. Coupled with fluid equations for the electron and ion motion, the microwave discharge courses are calculated. The results show that the density distribution of ion and electron is similar to electric field distribution. Also, a critical electron density is exist. Above this density, microwave field will damp rapidly, so the gas ionization is mainly on the surface.

This paper point out that there is no n> 13 Barker code concluded by Yang Guangzheng(1995) is not correct. So it is still a conjecture that there is no n> 13 Barker codes.

This paper presents an image segmentation approach which is based on fuzzy-neural-network hybrid system(FNNHS). This approach can use the empirical knowledge and the ability of neural networks which learn knowledge from the examples, to obtain the well performed fuzzy rules. Furthermore this fuzzy inference system is completed by neural network structure. The segmentation process consists of pre-segmentation based on region growing algorithm and region merging based on FNNHS. The experiments illustrate the power and efficiency of this method used for complicated image.

Multipath time delay estimation is constrained by the width of the signal correlation function when using correlation method. This paper obtains a high resolution tune delay estimation by introducing Burg and Marple algorithm of the maximum entropy power spectrum estimation to non-resovable multipath time delay estimation. The principle, the performances and the results of computer simulation are given.

Prom a polynomial ideal point of view, a general algebraic method for decoding Goppa codes is presented. It is shown that such a method can correct t = [(d - 1)/2] errors, where d is the true minimum distance of the given Goppa code.

Based on a one-way function, this paper proposed an authentication-doubled access control scheme for information protection system. The scheme is safer than the previously proposed schemes. In the scheme, the user s secret key is used not only for computing the corresponding access privilege to the intended file, but also for authenticating the requesting user not to illegitimately access the protected file. The scheme can perform the access control in dynamic environments, such as change access priviledges and insert/delete users or files. Beside, the scheme is simple to establish.

The constitution, technical features and working mode of the real-time onboard displaying system of Airborne SAR real-time digital imaging processor are studied systematically. And the special development of the display board under the real-time system is discussed. Further more the characteristics of the real-time operating system and its pSOS+ kernel that contitutes the software enviroment in the real-time dispaly system is presented. And also the multitasks scheduling management, communication and synchronization mechanisms applied in the real-time onboard diplay system are discussed.

This paper gives the calculating and testing method for the parameters of materials of low loss and high dielectric constant in the millimeter-wave-band. The measuring principle. and the testing system are introduced. The measuring results are given. This method is precise, practical and convenient.