Table of Content

20 August 2004 Volume 31 Issue 4

  

Original Articles

Similarity between the distorted model and the prototype of the large radio telescope

QIU Yuan-ying;DUAN Bao-yan;SHENG Ying;HUANG Yi-hong

J4. 2004, 31(4):  493-496. 

Abstract ( 2666 )   Save

Related Articles | Metrics

The dynamic differential equations for suspended cabin-cable structure of the large radio telescope are given, with the similarity conditions between the distorted model and the prototype obtained from the equations. Furthermore some typical scale factors such as the frequency scale factor, wind velocity scale factor, an mass scale factor are described by the known geometrical scale factors utilizing the essential dynamic similarity conditions. Based on the similarity analysis and the vibration experiments on the model, the natural frequencies and amplitude of wind-induced vibration of the prototype are predicted. Comparing the predictions with the dynamic simulation made by the nonlinear finite element method, consistent results are demonstrated adequately.

Determination of second-order reflection points in the system of cylinder and cone using Format's principle

LIANG Chang-hong;TAN Yuan-yang;ZONG Wei-hua

J4. 2004, 31(4):  497-500. 

Abstract ( 1391 )   Save

Related Articles | Metrics

The ray tracing of the second and higher order reflection is an important and difficult problem. The tracing equations in earlier papers involve several parameters, which leads to taking more time to solve the equations using the numerical searching method. In this paper, the analytical formulations for every reflection point's Z coordinate are deduced by using Fermat's principle and law of reflection in the system of cylinder and cone. Some application examples are given to illustrate that the tracing-time is dramatically reduced and the operational precision is improved. The method presented in this paper can be expanded to solve the ray tracing of the second order reflection of some other model and used as a reference to solve the higher order reflection problem.

基于3DES的跳频序列族构造方法的VLSI实现

LI Zan1;CAI Jue-ping2;JIN Li-jun1;CHANG Yi-lin1

J4. 2004, 31(4):  501-504. 

Abstract ( 1998 )   Save

Related Articles | Metrics

This paper deals with the VLSI realization of the frequency hopping sequences generator based on the encrypted mechanism of 3DES block ciphers. The VLSI architecture design of the algorithm is efficiently implemented using the VHDL language adopting the method of the Finite State Machine. Characterized by stability, fast operation and flexible data input, the realized generator can satisfy the requirement of 2000hops/s with the system clock ranging within 1.5MHz~24MHz, which has been used in fast FH radios.

Pseudo-linear Kalman filter based passive location and tracking techniques by two infrared stations

XIN Yun-hong1;2;YANG Wan-hai1

J4. 2004, 31(4):  505-508. 

Abstract ( 1825 )   Save

Related Articles | Metrics

A PLKF(pseudo-linear Kalman filter) is presented for tracking a moving target using two stations of IRST systems. The initial value of the filer is attained from the pseudo-linear equations, which improves the tracking accuracy and speed of the filter. Results of the simulation of tracking a maneuvering target by PLKF and EKF(extended Kalman filter) illustrate that at the beginning of tracking, the tracking accuracy of PLKF is better than that of EKF; in the near range of the tracking, both the tracking accuracy of PLKF and EKF are better whether the target is maneuvering or not; in the far range of the tracking, the tracking accuracy of PLKF is better than that of EKF when the target maneuvers, but is somewhat worse when the target moves at a constant velocity. From the view of global range, the influence of the moving state of the target is very strong on the tracking accuracy of the EKF and is weak on that of PLKF.

Analysis of RLC interconnect tree delay based on "effective capacitance"

DONG Gang;YANG Yin-tang;LI Yue-jin

J4. 2004, 31(4):  509-512. 

Abstract ( 1453 )   Save

Related Articles | Metrics

Interconnect delay evaluation is always a crucial concern in the VLSI design. An interconnect line in a VLSI circuit is in general a tree pattern rather than a single line. An approach to analyzing RLC Interconnect tree delay based on "effective capacitance" is presented in this paper. This new method is compared with the equivalent Elmore delay, which shows that the relative error by the new method is less than that by the equivalent Elmore delay.

A blind image watermarking based on zerotree wavelet

ZHUANG Huai-yu;WU Cheng-ke;LI Yun-song

J4. 2004, 31(4):  513-517. 

Abstract ( 1619 )   Save

Related Articles | Metrics

A blind image watermarking procedure based on teh zerotree wavelet is presented. A two-value image as a symbol of copyright protection is used, which is first preprocessed with a random location matrix. The embedding watermark data are obtained adaptively by the host image, and embedded into the zerotree wavelet of the host image. Two keys are used in the watermarking procedure: random location matrix and spread spectrum watermark matrix. The first one ensures the secuirty of watermarking procedure and the second one guarantees its robustness. Under the condition of no original host image, the experimental result demonstrates the robustness of our blind watermarking procedure against serveral image degradations and watermarking attacks.

Two novel calculations of coordinates of finite field elements

WEI Bao-dian;LIU Jing-wei;WANG Xin-mei

J4. 2004, 31(4):  518-522. 

Abstract ( 1570 )   Save

Related Articles | Metrics

First, we design an approach to determining the corresponding trace function of a linear transformation over finite fields, on the basis of which, we have proposed an effective methodology to compute the algebraic representations of coordinates of finite field elements in the form of trace functions. Second, by calculating the dual basis of the standard basis, we develope another method to find the algebraic representations of coordinates of finite field elements, also in the form of trace functions. The necessary data for both methods are only n elements over the corresponding fields with time complexity each of O(n³). This is a great improvement in contrast to the existing data and time complexity of O(p^n-1) and O(np^2n-2) respectively. Finally, based on the algebraic representations of coordinates which could be computed easily with our new approaches, we have given a direct proof of the equivalence between any two coordinate functions of Rijndael S-box and we have depicted this equivalence with only one matrix of order eight over GF(2⁸). Compared to the existing ones, in which the equivalence was found by the search of the affine parameters and was illustrated with no less than 56 matrices of order eight over GF(2), our proof seems more straightforward and simpler.

A novel method of the interference suppression for a uniformly-spaced linear array

LI Ping;SHI Xiao-wei

J4. 2004, 31(4):  523-525. 

Abstract ( 1438 )   Save

Related Articles | Metrics

This paper addresses a novel method of the interference suppression for a uniformly-spaced linear array. Through some virtual translation on the array, received signals may be virtually transformed. And the signals are transformed correspondingly. Then a weight is obtained by a rule of Max SNR. This method overcomes the disadvantage of caliber reduction by the spatial averaging technique. The validity of the method is illuminated by computer simulation.

The application of the spread spectrum on the multi-carrier radar syste

BO Hua1;MA Fu-long2;JIAO Li-cheng3

J4. 2004, 31(4):  526-529. 

Abstract ( 1908 )   Save

Related Articles | Metrics

Multi-carrier spread spectrum radar is presented. In the multi-carrier spread spectrum radar system, phase difference is used to measure the object range and the Doppler shift is used to measure the velocity of the object. Multi-carrier spread spectrum radar has the advantages of higher performance of anti-narrow band interference and lower peak average ratio. Simulation results show the effectiveness of the proposed radar system.

Equal-power N-th amplitude sequeezing in the five state superposition multimode superposition state light field

XU Ding-guo;AN Yu-ying

J4. 2004, 31(4):  530-532. 

Abstract ( 1385 )   Save

Related Articles | Metrics

The properties of the Equal-power amplitude squeezing of the generalized magnetic field component in the five state superposition multimode superposition state light field |Ψ⁽⁵⁾>_q is studied in detail by utilizing the multimode squeezed state theory. The results show that the generalized magnetic field component of the state |Ψ⁽⁵⁾>_q appears as an equal-power amplitude squeezing effect under the special conditions.

A study of a new fuzzy clustering algorithm based on the kernel method

WU Zhong-dong1;GAO Xin-bo1;XIE Wei-xin2

J4. 2004, 31(4):  533-537. 

Abstract ( 1791 )   Save

Related Articles | Metrics

We present a fuzzy kernel C-means clustering algorithm(FKCM) which is a generalization of the conventional fuzzy C-means clustering algorithm(FCM). This new FKCM algrotihm integrates FCM with the Mercer kernel function and can cluster non-hyperspherical data structure, data with noise, mixed data structure of multi pattern prototypes, asymmetric data structure, etc. This generalization can obviously improve the performance of the fuzzy C-means clustering algorithm. It is pointed out that the FKCM algorithm with the first-order polynomial kernel function is equivalent to the FCM algorithm. The results of experiments on the artificial and real data show that the fuzzy kernel C-means clustering algorithm can effectively cluster on data with diversiform structures in contrast to the fuzzy C-means clustering algorithm.

Analytical model for the electron Hall mobility in th n-type 4H-SiC

WANG Ping1;YANG Yin-tang1;QU Han-zhang2

J4. 2004, 31(4):  538-542. 

Abstract ( 1684 )   Save

Related Articles | Metrics

Theoretical calculations have been made on the electron Hall mobility and Hall scattering factor in the nitrogen-doped 4H-SiC based on an simplified analytical low field transport model considering five main scattering mechanisms including neutral impurity scattering. The results show that at lower temperatures, with increasing doping concentration, the impact of neutral impurity scattering becomes more significant. However the electron Hall mobility is controlled by intervalley phonon deformation potential scattering when the temperature is higher. In addition, the electron Hall mobility depends strongly on the compensation ratio when the donor concentration is kept constant. The Hall scattering factor is not always equal to one, but has a strong dependence on temperature. Comparisons with experimental data confirm the present calculation over a wide range of temperatures and doping concentrations.

A simplified fast processing method for high squint airborne SAR

HUANG Yuan-bao;LI Zhen-fang;BAO Zheng

J4. 2004, 31(4):  543-546. 

Abstract ( 1556 )   Save

Related Articles | Metrics

The high squint SAR signal characterizes a large range walk and a small range curvature, which enables us to compress it in range and azimuth separately. But in case of the large azimuth scene size, complex algorithms must be employed to cope with the constraint of focus depth. In this paper we present a simplified fast sub-aperture method which incorporates reshaping in imaging and is able to group sub-images seamlessly. Finally simulation results and real data image are given to verify our method.

Study of the schema theorems for recombination

MING Liang1;WANG Yu-ping1;JIAO Yong-chang2

J4. 2004, 31(4):  547-550. 

Abstract ( 1949 )   Save

Related Articles | Metrics

Schema theorems are important theoretical results in genetic algorithms. However, the general schema theorems only discuss the probability of the survival of schemas, but do not discuss the probability of the construction of schemas. In this paper we discuss the schema theorems affected by both the survival and the construction of schemas. First, some basic concepts including the subschema, the complementary-schema, the survival and the construction of schema are introduced, and then the schema theorems affected by both the survival and the construction of schemas are proposed. They are generalizations and extensions of the existing schema theorems.

Immune monoclonal strategy based on the Cauthy mutation

LIU Ruo-chen1;2;DU Hai-feng1;JIAO Li-cheng1

J4. 2004, 31(4):  551-556. 

Abstract ( 2074 )   Save

Related Articles | Metrics

Based on the clonal selection theory, the main mechanisms of clone are analyzed. An improved evolutionary strategy algorithm-Immune Monoclonal Strategy algorithm based on the Cauthy Mutation(IMCSCM) is presented, in which the Gauss mutation in the Classical Evolutionary Strategies algorithms(CES) is replaced by the Cauthy one. Compared with CES and the Immune Monoclonal Strategy algorithm applying the Gauss Mutation(IMCSCM), IMCSCM is shown to be an evolutionary strategy capable of avoiding prematurity, increasing the converging speed and keeping the variety of solution in the simulations. Using the theories of Markov Chain, its convergence is proved.

Research on estimating the generalization performance of RBF-SVM

DONG Chun-xi;RAO Xian;YANG Shao-quan;XU Song-tao

J4. 2004, 31(4):  557-561. 

Abstract ( 2161 )   Save

Related Articles | Metrics

Using the sparseness of a Suport Vector Machine(SVM) solution, properties of the raidal basis function kernel and the inter-median parameters in training the SVM, a new algorithm for estimating the generalization performance is presented, which overcomes many disadvantages of the existing algorithm such as longer computation time and narrower application range without additional complex computing. Theoretic analysis and experiments prove that it is a universial method for estimating the generalization performance of an SVM and it can be applied to a wide range of problems of pattern recognition using SVM.

Hybrid analysis of scattering by microstrip antennas residing in a cavity on an electrically large body

TANG Song;PENG Hong-li;LIU Qi-zhong

J4. 2004, 31(4):  562-564. 

Abstract ( 1873 )   Save

Related Articles | Metrics

A new hybrid technique for calculating the scattering of an electrically large conducting body by microstrip antennas on its surface is presented. The high-frequency and the low-frequency methods are combined via the equivalence theorem. This technique overcomes the shortcomings of the existing methods and can apply to concave large conducting bodies by introducing SBR or IPO. Some numerical results have been obtained, demonstrating the accuracy, efficiency and capability of the technique.

A technology for protection from latch-up based on current rejetion by LDO

ZHANG Gong-wei1;2;JIANG Xuan-xiang2;TANG Xue-han2;HAO Yue1

J4. 2004, 31(4):  565-568. 

Abstract ( 1727 )   Save

Related Articles | Metrics

This paper discusses the similarities and differences between the latch-up induced by electrics and the latch-up induced by radiation for CMOS IC's. A technical scheme for Latch-up protection based on current rejection by a low-dropout linear voltage regulator(LDO) is developed, with the onboard experimental result and the experimental data of TDO on ground of the LDO circuit used on the satellite given. Finally, the problem of protection and recovery from latch-up on the onboard-computer has been solved.

Private key distribution scheme of ID-based encryption

LI Xin-guo1;2;GE Jian-hua1;ZHAO Chun-ming1

J4. 2004, 31(4):  569-573. 

Abstract ( 1522 )   Save

Related Articles | Metrics

To solve the problems of authenticating the user's identity and distributing the user's private keys in the identity-based encryption(IBE) system, a new scheme for both identity authentication and private key distribution is proposed as a reinforcement for IBE by smartly using the concept of public-key certificate. A trusted third party is employed to transfer trust, and an additional public key is used to set up the secure channel for transmitting the private key. Not only is the number of authentications needed decreased significantly, but the private key distribution problem which has never been dealt with in the original scheme and relating applications is addressed in the new scheme. Based on IBE, we also propose an encryption scheme which has one public key with two corresponding private keys. Both of the private keys can be used for decryption, one of which is escrowed to a trusted third party, the other for non-repudiation signature.

A data mining based design for the detection engine of the intrusion detection system

Lü Xi-xiang;YANG Bo;PEI Chang-xin;SU Xiao-long

J4. 2004, 31(4):  574-580. 

Abstract ( 2045 )   Save

Related Articles | Metrics

We discuss our research in developing the detection engine of the intrusion detection system. The key ideas are to combine the slide window into the data mining technique to design the base detection engine which is the essential share of the meta detection engine. In addition, Apriori, a kind of data mining algorithm, is improved to mine network data. The improved algorithm does not scan all items in database and only links the items in the same list, so the detection efficiency is improved greatly. Also, other key details in IDS are put forward.

A new generalized colour visual cryptography scheme

MA Wen-ping;REN Ya-an

J4. 2004, 31(4):  581-584. 

Abstract ( 1486 )   Save

Related Articles | Metrics

Visual cryptography was first introduced by Naor and Shamir and the scheme has the benefit of requiring no cryptographic computation on the part of decoders. The modified visual cryptography is a new scheme, which requires little cryptographic computation on the part of decoders. In this paper, a new generalized (k,n) gray visual cryptography scheme is proposed based on modified visual cryptography. This scheme provides a more efficient way to hide a gray image which has 256-colors in n different shared images. Also, for a computer system, two types of unconditional security generalized (k,n) color visual cryptography scheme are proposed.

Distributed certificate services for the scalable Ad Hoc network

DU Xin-jun;WANG Ying;GE Jian-hua

J4. 2004, 31(4):  585-588. 

Abstract ( 1955 )   Save

Related Articles | Metrics

Because of the inherent characteristics in the Ad oc wireless network, the centralized authentication services provided by CA in traditional networks is not suitable for this network. In this paper, we apply the threshold digital signature scheme based on the discrete logarithm to distribute the authentication services into the whole network, so that each individual node can potentially provide other nodes with certification services. The scheme not only localizes the authentication service but also adapts to the scalability. At the same time, this scheme overcomes the limit to other distributed certification services which need to trusted party in the system initialization phase.

The cone super subdifferential stability o multiobjective programming with a perturbed order

HOU Zhen-hai;LIU San-yang;ZHOU Yong

J4. 2004, 31(4):  589-592. 

Abstract ( 1249 )   Save

Related Articles | Metrics

This paper deals with the existence of the cone super-subdifferential of the point sets mapping in local convex topological vector spaces and the stability of the super subdifferential of cone super efficiency point sets and efficiency point sets and the weak efficiency of multiobjective programming with respect to perturbation of cone respectively.

A new part-lossless image compression algorithm based on regions of interest

LI Ping;XU Lu-ping;XIE Xue-mei

J4. 2004, 31(4):  593-596. 

Abstract ( 1775 )   Save

Related Articles | Metrics

A new algorithm with lossless regions of interest is present based on the Integer Wavelet Transform(IWT) and SPIHT scheme. It utilizes the good space-frequency local characteristic of the wavelet transform, and lifts the wavelet coefficients of regions of interest in an image, and by this means, these wavelet coefficients will be separated from wavelet coefficients in the background. An image can be encoded with an arbitrarily shaped ROI by this method, and an arbitrarily shaped ROI can be decoded without transmitting any information about the shape of ROI. Experimental results show that the algorithm can losslessly preserve regions of interest and obtain a higher compression ratio of the whole image.

Multi-pose face recognition based on orthogonal views

WU Mang;WANG Yan;LIU Zhi-jing;LI Xia-zhong

J4. 2004, 31(4):  597-601. 

Abstract ( 2064 )   Save

Related Articles | Metrics

To deal with the problem of the traditional multi-pose face recognition based on the multi-view that it must have many face photos, a multi-pose face recognition algorithm based on the two orthogonal views is proposed. It consists of two steps. In the first step, a 3D face model is built by the person's two orthogonal views and multi-pose face images are produced by the projections of the 3D face model. In the second step, multi-pose face recognition is performed based on the training set that consists of the two orthogonal views and the synthesized multi-pose images. The experimental results show that the performance of the algorithm discussed in the paper is by far superior to that of the method based on the front view.

MAC protocol of WLAN using smart adaptive array antennas

ZHAO Li-qiang;FAN Chang-xin

J4. 2004, 31(4):  602-607. 

Abstract ( 1971 )   Save

Related Articles | Metrics

Based on the characteristics of IEEE 802.11 WLAN, a medium access control protocol SADCF is proposed to support smart adaptive array antennas. A smart antenna has omnidirectional and directional operation modes. In SADCF, before sending any data frame, the station stimulates the corresponding station to transmit a training sequence by means of omnidirectional RTS/CTS handshake mechanism. Based on teh training sequence, the directional beam can be formed by the smart antenna. Then the station can transmit and receive data frames in directional mode. Other stations save the transmission time of the sender in omnidirectional mode by using the omnidirectional network allocation vector, and multiplex channel when it transmits in directional mode. SADCF is compatible with IEEE 802.11. Simulation results show that the SADCF protocol can suport smart antennas effectively and provide high throughput by multiplexing the channel, and results of the analysis are close to the simulation results.

Remote sensing image fusion based on the imaging principle

NA Yan;YANG Wan-hai;SHI Lin

J4. 2004, 31(4):  608-610. 

Abstract ( 1649 )   Save

Related Articles | Metrics

The multi-spectral and panchromatic image fusion is discussed. A new wavelet transform based fusion method is presented after analyzing the basic principles of remote sensing imaging and fusion target. The experiments show that the fusion method can efficiently fuse the multi-spectral and panchromatic image with a satisfactory result.

A traitor tracing scheme based on RSA

MA Hua1;CAO Zheng-wen2

J4. 2004, 31(4):  611-613. 

Abstract ( 1524 )   Save

Related Articles | Metrics

The traitor tracing scheme is a very useful tool for preventing piracy in the context of digital content broadcast. A traitor tracing procedure allows the data-suppliers to reveal the identities of the subscribers that were implicated in the construction of a pirate-device that illegally receives the digital content. In this paper, a traitor tracing scheme based on RS is proposed, which allows efficient deterministic traitor tracing that recovers all traitors.

A micropayment protocol based on mobile agents

LIU Yi;WANG Yu-min

J4. 2004, 31(4):  614-617. 

Abstract ( 1416 )   Save

Related Articles | Metrics

Mobile agents are believed to play an important role in future 3-commerce systems. However, their adoption in largely hampered by the new security issues they raise. Based on mobile agents, this paper presents a micropayment protocol. The results show that the protocol has great flexibility, security, non-repudiation and proper efficiency.

A smart antenna circular array implementation and its analysis with mutual coupling

GAO Feng1;SHAN Run-hong2;LIU Qi-zhong1;LU Yong-chao3;XIAO Liang-yong3

J4. 2004, 31(4):  618-621. 

Abstract ( 1985 )   Save

Related Articles | Metrics

This paper investigates an efficient procedure integrating the genetic algorithm with the method of moment for the optimal design of the smart antenna element. Then a novel coaxial collinear printed dipole antenna is presented, which is used as an element in the circular smart antenna array. The analysis of the circular smart antenna array considering the effects of mutual coupling is obtained. Finally the effectivness of the analysis is evaluated by experimental antenna arrays. The results are useful for the development of a smart antenna system in the 3rd generation mobile communications.

The linear complexity and the k-error linear complexity of pⁿ-periodic binary sequences

NIU Zhi-hua;DONG Qing-kuan;XIAO Guo-zhen

J4. 2004, 31(4):  622-625. 

Abstract ( 1806 )   Save

Related Articles | Metrics

Not only should cryptographically strong sequences have a large linear complexity, but also the change of a few terms should not cause a significant decrease in linear complexity. This requirement leads to the concept of the k-error linear complexity of periodic sequences. A relationship between the linear complexity and the k-error linear complexity of pⁿ-periodic sequences over GF(2) is studied, where p is an odd prime, and z is a primitive root modular p². A necessary and sufficient condition that the k-error linear complexity be strictly less than the linear complexity is shown. A sufficient condition expressed by the error polynomial E^N(x) that LC(S+E)<LC(S) and an upper bound of the minimum value k for which LC_k(S)<LC(S), i.e. minerror(S), are given.

Comparison and improvements of several methods on wavelet image denoising

LIN Zhen-xian1;2;SONG Guo-xiang1;XUE Wen3

J4. 2004, 31(4):  626-629. 

Abstract ( 1857 )   Save

Related Articles | Metrics

Wavelet image denoising has been well acknowledged as an important method for denoising in Image Processing. The principles of several wavelet denoising methods including the Mallat forced denoising, the wavelet transform modulus maximum method, the nonlinear wavelet threshold denoising method and the denoising based on the interrelation of wavelet domain are described, and this paper compares these methods, and gives two improved methods. The results show that they are practicable and effective for image denoising.

Multi-scale edge detection of the image based on the interval biorthogonal wavelet

YANG Wei1;2;SU Wan-li3;SONG Guo-xiang1

J4. 2004, 31(4):  630-633. 

Abstract ( 1735 )   Save

Related Articles | Metrics

Th ability of the interval biorthogonal wavelet to detect the multi-scale edge is theoretically analyzed. The wavelet transform of the image is performed by the biorthogonal wavelet with the vanishing moments N=N(~)=3 in the interval [0,1]. Then a multi-scale edge detection algorithm of the wavelet is proposed to obtain the local maxima along the direction given by the angle image. Simulation results show that teh biorthogonal wavelet in the interval is more suitable for the edge detection of images than both Daubechies' biorthogonal wavelet and orthogonal wavelet.

An adaptive vector median filter for noise removal in the color microscopic image of cells in marrow

SUN Wan-rong1;2;YU Bian-zhang1;ZHU Tian-qiao2

J4. 2004, 31(4):  634-638. 

Abstract ( 1375 )   Save

Related Articles | Metrics

An adaptive vector median filter for removing the noise in the color image is designed and tested with the color microscopic image of cells in marrow. In the algorithm, first the average distance between vector median X_VM and all other vectors in a 3×3 window is computed and then noisy pixels are removed with adaptive selection of window size based on the average distance. When the window size is 1×1, the vector pixels should be unchangeable. The experiment with the color microscopic image of cells in marrow waw carried out and the results show that the filter can achieve a high degree of noise suppression and keeps the image sharpness. The adaptive vector median filter also shows a significant improvement over the regular vector median filter. The algorithm is rather simple to implement.

A fast algorithm for extracting the support vector on the Mahalanobis distance

WANG Xi-li1;2;JIAO Li-cheng2

J4. 2004, 31(4):  639-643. 

Abstract ( 1935 )   Save

Related Articles | Metrics

A method for extracting data which most probably are suport vectors for SVM by the Mahalanobis distance from a vector to a class is presented. How to compute Mahalanobis distance in the input and feature space is described in detail. The algorithm is fast since there are efficient methods for finding eigenvalues and eigenvectors of a symmetric matrix or comptuing pseudoinversion involved in finding the Mahalanobis distance. The training time for SVM can be reduced when the training set is preprocessed in this way. Experimental results illustrate its effectiveness.

A problem on the UAV navigation system based on multi-model variable structure intelligent control

XI Qing-biao1;YUAN Dong-li2;YAN Jian-guo2

J4. 2004, 31(4):  644-647. 

Abstract ( 1835 )   Save

Related Articles | Metrics

Due to the limit of traditional UAV navigation control, this paper puts forward a kind of navigation control system that is based on teh strategy of multi-model variable structure intelligent control. f a discontinuous control structure is built up we can get good control quality by using a different control structure when the state of control system changes. At the same time not only is this system insensitive to the parameter perturbation but also this system can restrain wind disturbance. In a word, this system has strong robustness. Since it adopts the digital control method this system is easy to realize in engineering with high reliability.

The application of orthogonal demodulation in OFDM signal demodulation

TIAN Feng1;CHENG Ren2

J4. 2004, 31(4):  648-651. 

Abstract ( 1359 )   Save

Related Articles | Metrics

This paper analyzes the calculation for frequency offset relative to given functions. It is found that it has a smaller range of frequency offset and can be easily attested by Doppler frequency offset. Considering this, the paper introduces a calculation method of carrier demodulation based on the orthogonal algorithm, which the frequency offset can be omitted. The results of computation simulation indicates that the orthogonal demodulation is applicable to the AM modulated OFDM signal, in that not only can frequency offset be ignored in it, but it can be available in the channel with a low SNR.

Digital watermarking using a human auditory system

JIA Jun;WANG Shuo-zhong;ZHANG Xin-peng

J4. 2004, 31(4):  652-656. 

Abstract ( 1473 )   Save

Related Articles | Metrics

This paper introduces a new technique of audio watermarking based on the property of the human auditory system(HAS) and MP3 coding, in which magnitudes of the frequency components below hearing thresholds are modified for watermark embedding. Neither the host signal nor any synchronization information is needed on the extraction side. Experimental results show that the proposed scheme can achieve high imperceptibility and satisfactory robustness against various common attacks.