Table of Content

01 August 2006, Volume 26 Issue 8

Previous Issue    Next Issue

Articles

On highspeed access network technology

ZENG Hua-shen Gao Yu

2006, 26(8):  1751-1755. 

Asbtract ( )   PDF (1306KB) ( )  

Related Articles | Metrics

It is the development of optical communications that has raised transmission capability of a single fiber up to 6 orders of magnitude, and provided a solid communication basis for high-speed backbone networks. However, the access network technology develops comparatively slowly and has been regarded to as the “Last Mile” or “First Mile” communication problem. Access network refers to that from the edge of backbone networks to individual costomer computer,and playsan important part in the communication subnet composed if Relay System or Intermediate System in OSI/RM.Therefore,access network may get the three lower layers in OSI/RM involved. The access network communication techniques with an emphasis on wire-based high-speed communication techniques were addressed, although a few new wireless techniques with potentials were also discussed. Finally,it concludes that high-speed access network communication techniques in the future will be characterized by diversity, higher speed, long transmission distance, and lower cost.

Overview on wireless mesh networking and mobile transition techniques in IEEE 802 series

FANG Xu-ming Cai-Xia QI Zheng XIANG

2006, 26(8):  1756-1761. 

Asbtract ( )   PDF (1273KB) ( )  

Related Articles | Metrics

When constructing large scale wireless area networks, the traditional network techniques become unaccommodated due to lack of flexible extension and fast transition support. The IEEE standard organizations are devoted to advancing the technology of mesh networking and fast transition or mobility support. The mesh networking and transition or mobility support techniques were discussed based on Wireless Local Area Network(WLAN), Wireless Personal Area Network(WPAN) and Wireless Municipal Area Network(WMAN). The development tendency of wireless area networks was introduced to benefit the research and application of the wireless multihop networks.

Overview on call admission control schemes in wireless communication networks

FANG Xu-ming

2006, 26(8):  1762-1767. 

Asbtract ( )   PDF (1317KB) ( )  

Related Articles | Metrics

Call Admission Control (CAC) plays an important part in radio resource management in wireless communication networks. It permits or denies the call arrival connection requests based on the certain criteria. The design of CAC has direct impact on the network performance and resource unitization. CAC schemes have been widely investigated during the last two decades, which greatly improves the network performances, such as call blocking and dropping probabilities, packet delay, loss rate, and throughput. Related background knowledge about CAC in wireless communication networks and its requirements and functions were introduced. Some recent research results on CAC schemes were outlined with great emphasis. The problems or drawbacks in current schemes were pointed out. Finally, ideas for further research were discussed.

Network and communications

Research on gateway-aided overhead-restricted cluster routing protocol for MANET

2006, 26(8):  1768-1771. 

Asbtract ( )   PDF (871KB) ( )  

Related Articles | Metrics

As for the current cluster-based routing protocols for MANET, some problems exist such as the overloaded cluster-head, and the flooded entire network even chain reaction while two clusters are merging. To solve these problems,a Gateway-aided Overhead-restricted cluster Routing Protocol(GORP) for MANET was proposed. In GORP, gateways share part of the work of clusterheads for maintaining routing and transmitting data. In this way,the burden of cluster-heads was reduced and cluster-heads were no longer the bottleneck of the network. When the merging of clusters occurs, which is caused by the moving of a cluster-head to the range of another cluster-head, the moving cluster-head should be set as an ordinary node, and the original still cluster-head as the cluster-head of the new cluster. Thus, the stability of cluster could be kept to the most extent. Simulation results show that GORP can work well for MANET with different nodal mobility, nodal density and network scale.Cluster-based routing security in MANET can also benefit from the stable structure of cluster.

Subnet-based data-centric routing algorithm for sensor networks

2006, 26(8):  1772-1775. 

Asbtract ( )   PDF (742KB) ( )  

Related Articles | Metrics

Two new concepts as subnet and energy-core were proposed, and an innovative ant algorithm was designed. Besides,a distributed data-centric hierarchical ant routing algorithm for sensor networks, which is based on subnet and named DDCHA was proposed. Compared with previous centralized algorithms such as AC and GITDC, the proposed algorithm is distributed and requires local location information only. Additionally, the positive feedback ability of ant algorithm and the construction of subnets and energy-core can improve the degree of data aggregation, and reduce the time complexity of the proposed algorithm at the same time.Both theoretical analysis and experiment results show that the algorithm is effective.

Battery-power-based routing algorithm for mobile Ad hoc networks

2006, 26(8):  1776-1778. 

Asbtract ( )   PDF (512KB) ( )  

Related Articles | Metrics

A method which use mobile Agent to solve the battery-power-routing for mobile Ad hoc networks was introduced. A few mobile Agents communicated with every node to collect the network connection information to build the global information matrix of nodes. By information matrix,data packets were routed and transferred according to the remaining battery power. This approach needs fewer mobile Agents but gets more global battery power information, therefore it can reduce the cost of maintaining the routing information on each node. The experiment results show that the battery power on each node can be balanced with this routing algorithm.

Multicast congestion control based on hop to hop limited

2006, 26(8):  1779-1881. 

Asbtract ( )   PDF (576KB) ( )  

Related Articles | Metrics

A new IP layer multicast congestion control algorithm based on hop to hop Radom Early Detection（RED）, which is named STSR, was put forward,and its effectiveness was verified by means of mathematics.It was proved that STSR multicast congestion control had greater sensitivity and multicast transmission link had larger throughput. Simulation by applying NS-2 show that the STSR algorithm might ensure the TCP friendly as well as the distributed control theory of coarser granularity. While raising the IP layer multicast reliability and scalability,STSR guarantees the efficiency and quality to its greatest extent.

Research on networks simulation technology of channel access protocol

2006, 26(8):  1782-1784. 

Asbtract ( )   PDF (546KB) ( )  

Related Articles | Metrics

By using OPNET simulation platform, the modeling and simulating technology for channel access protocols based on ALOHA and CSMA/CD were studied. Modeling hierarchy framework of ALOHA and CSMA/CD was constructed, and their network model,node model and process model were realized. By analyzing the simulation results of channel throughout and utilization, the rationality and effectiveness of the simulation model was evaluated,so as to prove the feasibility of this networks simulation technology.

Minimum power drain rate routing algorithm for Ad hoc networks

2006, 26(8):  1785-1787. 

Asbtract ( )   PDF (545KB) ( )  

Related Articles | Metrics

In Ad hoc networks, the energy and drain rate of mobile node can influence the networks lifetime and stability. An routing algorithm with high efficiency named Minimum Power Drain Rate(MPDR) was proposed. This routing algorithm takes both mobile node residual energy and power drain rate into account, so that it could expand networks lifetime and support QoS. Simulation results show that the proposed algorithm has the longest lifetime compared with other power-aware routing algorithms.

Application of distributed scheduling algorithm in VOQ switch

2006, 26(8):  1788-1790. 

Asbtract ( )   PDF (506KB) ( )  

Related Articles | Metrics

To improve the communication efficiency, a new communication-scheduling algorithm named DSA（Distributed Scheduling Algorithm）was put forward in VOQ switch by scheduling high-speed data packet in every output port independently. Compared with other traditional scheduling algorithms,which can only schedule the constant length message, the DSA can directly schedule the communication under variable message length. The simulation results show that the DSA scheduler outperforms iSLIP scheduling algorithm under all traffic.DSA can work efficiently in VOQ switch.

Analysis of adaptive algorithm in filtered multitone system

2006, 26(8):  1791-1792. 

Asbtract ( )   PDF (557KB) ( )  

Related Articles | Metrics

In this paper, an adaptive bit and power allocation algorithm that based on signal distortion ratio (SDR) is proposed in FMT system. Performance of adaptive system is analyzed in slow fading RayLeigh channel. The simulation result shows that the system outperforms fixed modulation FMT system in BER performance.

Low power routing algorithm based on dataaggregation tree for sensor networks

2006, 26(8):  1793-1795. 

Asbtract ( )   PDF (543KB) ( )  

Related Articles | Metrics

With regard to the limited resources of nodes in sensor networks,a new low power routing algorithm based on data-aggregation tree combining the concept of the smallest Steiner tree was proposed.This algorithm could improve the data transmission through fast establishment of a virtual backbone network.Theoretic analysis and simulation results show that the algorithm can be more energy-saving.

Replica optimisation strategy based on continuous double auction

2006, 26(8):  1796-1798. 

Asbtract ( )   PDF (580KB) ( )  

Related Articles | Metrics

A replication strategy based on a double auction mechanism was discussed to solve the problem that the replication strategy cannot be adapted to the high autonomy and dynamics of data.Firstly, a double auction model was described. Secondly, a continuous double auction protocol's implementation based on the distributed Continuous Double Auction Protocol(CDA) and Zero Intelligence Plus(ZIP) bid strategy were given. Finally,simulation with Optorsim was carried out to the replication strategy and model. The experiment results show that, compared with the one-side auction, the proposed model and strategy provide less operation time and higher network use rate, better Computing Elements(CE) and Storage Elements(SE) utilization ratio.

Analysis and simulations for ergodic capacity channel of MIMO systems

2006, 26(8):  1799-1801. 

Asbtract ( )   PDF (670KB) ( )  

Related Articles | Metrics

Under the assumption that the channel state information was available only at the receiver,the capacity of MIMO systems in Rayleigh fading channel was investigated on the basis of information theory. Three special expressions for the MIMO capacity over ergodic flat fading channel were derived.An asymptotic formula for MIMO systems with equal number of transmit and receive antennas in small SNR was also given. Simulation results show that this approximation is relatively accurate.

Information security

Multiple one-way hash random key pre-distribution protocol in sensor networks

2006, 26(8):  1802-1806. 

Asbtract ( )   PDF (948KB) ( )  

Related Articles | Metrics

Firstly,a basic multiple one-way hash key distribution protocol was developed based on Leighton Micali protocol. It ensures that all neighbors’ nodes can establish secure links, but has poor safety performance. Then a novel key distribution protocol integrating multiple one-way hash with random key pre-distribution was presented. Only one-way hash function is requirement for sensor node, which is very suitable for sensor networks.Compared with other protocols,this one has better performance,which need less computing overhead and can enhance the security.

Research and application of architecture based on trusted computing platform

2006, 26(8):  1807-1809. 

Asbtract ( )   PDF (849KB) ( )  

Related Articles | Metrics

Key components and functions of trusted computing platform were introduced first. Then, its characteristics and mechanism were described. After that, an overview of the actual research state was given. At last, the future applications and problems of the technology based on the trusted computing platform were analyzed, and a prospect about its development was made. The successful application of "A Trusted Computing Platform System" funded by 863 Projects indicates that future of the trusted computing platform will be promising.

Multi-service-oriented asymmetric public-key traitor tracing scheme with flexible revocation

2006, 26(8):  1810-1812. 

Asbtract ( )   PDF (515KB) ( )  

Related Articles | Metrics

Matsushita et al. proposed a traitor tracing scheme with flexible user revocation,which was a symmetric scheme, and did not provide multiple-service capability. An improved scheme was proposed by using Oblivious Polynomial Evaluation(OPE) and service parameter. In this improved scheme, such features of Matsushita's as flexible user revocation, black-box tracing and security remained,and other advantages such as providing multi-service and preventing traitors from repudiation(asymmetry) were added. The whole capabilities of the improved scheme were better than that of Matsushita's.

Application of point-to-point SMS encryption based on Java ME

2006, 26(8):  1813-1816. 

Asbtract ( )   PDF (1068KB) ( )  

Related Articles | Metrics

To keep SMS from interception in transmission, a solution of application of point-to-point SMS encryption based on Java ME was presented. This solution used encrypted SMS to deliver the new key. Considering the characteristics of SMS, effective measures were adopted to ensure that both sides could correctly communicate with each other through SMS.The new key SMS and its confirmation SMS were allowed to be sent again in case they were not received.Both sides would continue using the old key if the triplicate lifetime of the new key had elapsed and the new key was not updated. The means of solving special cases that occurred in the course of key updating were specified. Meanwhile, these measures were also analyzed in terms of security and reliability, and cases in need of attention were pointed out.

ID-based CES threshold signature

2006, 26(8):  1817-1820. 

Asbtract ( )   PDF (722KB) ( )  

Related Articles | Metrics

To improve the efficiency of multiparty interaction and reduce the problem of multiple signature and communication, an ID-based CES threshold signature based on the bilinear pairing was presented. Effective threshold signature recovery and signature extraction of the proposed signature scheme can be realized without more interaction with the signer. The proposed signature scheme ensures the efficiency and robustness of signature, and can defend existential forgery on adaptively chosen message and ID attack in the random oracle model.

Simple verifiable secret sharing scheme

2006, 26(8):  1821-1823. 

Asbtract ( )   PDF (562KB) ( )  

Related Articles | Metrics

Two effective Verifiable Secret Sharing (VSS) schemes including the Feldman's VSS scheme and the Pedersen's VSS scheme were analyzed. But as they are the threshold schemes, it is inefficient for them to be applied to the general access structure. Then, a new VSS scheme for the general access structure was proposed. The dealer created the shares randomly, and sent them to the participants by secret channel. Each authorized subset of the access structure had a public message, through which the validity of the shares could be verified respectively. This scheme has two types: one is computationally secure and the other is unconditionally secure. Their security is equal to that of the Feldman's scheme and the Pedersen's scheme respectively. But, it is more efficient to implement the new scheme on the same level of security.

Telephone encryption based on 3DES algorithm and its implementation on FPGA

2006, 26(8):  1824-1826. 

Asbtract ( )   PDF (737KB) ( )  

Related Articles | Metrics

With regard to communication security, the safekeeping of password and real-time processing in speech signal encryption based on triple Data Encryption Standard (3DES) were analyzed. In addition, the basic framework of speech encryption and the hardware system design in encryption telephone were presented. The functional module of system software was designed and the compiled program was loaded into the FPGA. The encryption telephone was validated in the public telephone net. This research indicates that the encryption functional module can be applied to point-to-point speech communication and other low speed data traffic.

Application of SSL in constructing virtual private network

XueJie Zhang

2006, 26(8):  1827-1830. 

Asbtract ( )   PDF (778KB) ( )  

Related Articles | Metrics

The principle of SSL protocol and VPN technology was analyzed, and the background of SSL VPN was introduced. Through the analysis of the simplest form of SSL VPN——Web reverse proxy, essence of SSL VPN was revealed, and an end-to-end security scheme of SSL VPN was proposed. At last, a comparative analysis between SSL VPN and IPSec VPN was made.

Packet marking scheme for IP traceback based on node sampling

2006, 26(8):  1831-1832. 

Asbtract ( )   PDF (556KB) ( )  

Related Articles | Metrics

Probabilistic Packet Marking(PPM) is an effective technique for IP traceback. However, the traditional PPM schemes are built upon unreal hypothesis, which brings out many limitations and affects their application. Based on reasonable hypothesis, Advanced Marking Scheme (AMS) was improved. With adjustable marking probability, counting distance by TTL value and node sampling, the performance of AMS got improved including weak convergence, inability of incremental deployment and being vulnerable to spoof marking attack.

Robust watermark algorithm based on the wavelet moment and neural network detection

2006, 26(8):  1833-1835. 

Asbtract ( )   PDF (1121KB) ( )  

Related Articles | Metrics

To enhance the ability of defending attacks such as rotation and cropping, an algorithm based on wavelet moment modulation and neural network detection was presented. This watermarking technique hided an imperceptible watermarking in a host image by modulating the wavelet moments of the host image, then a neural network was established to learn the relations between the host image's moment and the watermarking data, the trained artificial neural network could almost exactly restore the watermarking imbedded in the host image. The experiment results show that the algorithm can effectively resist signal processing attacks, such as filtering, compression, rotation and cropping, and the detection does not need the host image.

Certified E-mail protocol based on off-line TTP

2006, 26(8):  1836-1837. 

Asbtract ( )   PDF (514KB) ( )  

Related Articles | Metrics

Most protocols of electronic commerce are made upon Trusted Third Party (TTP).To use off-line TTP protocol for the certified E-mail protocol solves the TTP bottleneck problem in some degree. But this protocol has certain limitations. Unfairness to each side caused by these limitations and possible attack means were analyzed in detail. Meanwhile, corresponding solutions and suggestions for improvement were made.

Graphics and image processing

Morphological sampling based on generalized morphological transform

2006, 26(8):  1838-1841. 

Asbtract ( )   PDF (623KB) ( )  

Related Articles | Metrics

A generalized morphological transform based on the set cardinality was presented.This transform can reflect the intensity when the structural element acts on the image,and prevent excessive dilation or erosion. Some properties of the generalized morphological transform provide the basis for establishing a new morphological sampling scheme for both binary and gray image. The experimental results show that the new method has higher efficiency and better information preserve function of morphological transform compared with other sampling means.

Nonlinear inverse diffusion image enhancement algorithm

2006, 26(8):  1842-1844. 

Asbtract ( )   PDF (789KB) ( )  

Related Articles | Metrics

Traditional image enhancement algorithms could not eliminate noise and preserve textures at the same time. According to the principle of linear inverse diffusion, criteria on nonlinear inverse diffusion image enhancement algorithm were put forward and suitable diffusion coefficient was proposed. The experimental results show that this algorithm can enhance image effectively,eliminate the noise in the image and preserve the textures and details.

Visual perceptual edge linking based on two-level threshold

2006, 26(8):  1845-1847. 

Asbtract ( )   PDF (699KB) ( )  

Related Articles | Metrics

Edge is an important feature of image, but the edges obtained by the traditional edge detection usually are not continuous. Therefore, a two-level threshold edge linking based on visual perception with hysteresis was proposed,which employed the principal of human visual multi-level perception of edge linking. First, high and low thresholds were used to produce their threshold images respectively. Then, the difference between high and low threshold images was served to fuzzy edge image. After that, for each fuzzy edge point, several weighted visual perceptive factors were summed to judge whether it was a real edge or not. If it was a real one, then this fuzzy edge point would be added to high threshold image as an edge point. By this way, most essential edges were linked and closed edge boundary of object was generated. Simulation results show that the method can efficiently link gaps between edges. Compared with the traditional edge liking methods,the proposed method is fast and the effect of edge linking is favorable to meet the needs of edge linking demand in some applied cases.

Level set image segmentation based on simple Mumford-Shah model

2006, 26(8):  1848-1850. 

Asbtract ( )   PDF (1075KB) ( )  

Related Articles | Metrics

To overcome the deficiency of conventional methods in image segmentation influenced by image noise and edge contour, a level set approach to image segmentation was proposed on the basis of simplified Mumford-Shah model. Initial curve was embedded in the level set function,and image segmentation and curve evolution were accomplished when the function's solution was obtained. The experimental results show that it is a robust approach to image segmentation with noise, irrespective of edge continuous information and the curve initial position. Satisfactory effect has been achieved on image segmentation, especially homogeneous grayscale images.

Image steganalysis based on moments of wavelet subband histgrams in frequency domain using predictionerror image

2006, 26(8):  1851-1853. 

Asbtract ( )   PDF (637KB) ( )  

Related Articles | Metrics

A general steganalysis system was proposed. The test image and its prediction-error image were first decomposed by 3-level discrete wavelet transform, then 78 statistical moments of subband histgrams in frequency domain (using DFT) were calculated as features, and Support Vector Machines(SVM) was utilized as classifier. The experimental results based on CorelDraw image database demonstrate that our steganalysis system is superior to other prevailing ones.

Human face detection based on skin color model and BDF

2006, 26(8):  1854-1856. 

Asbtract ( )   PDF (1319KB) ( )  

Related Articles | Metrics

A frontal face detection system in color images was constructed. First,a skin color information database was established according to the characteristic that skin color is distributed in centralized region along Y axis in YCbCr space. According to the database, skin region could be detected. Second, the Bayesian statistical model using discriminating feature was constructed to detect multiple frontal faces in skin region. Finally, in order to accelerate the face searching, some heuristic searching criteria were defined. The experimental results show that the face detection system uses few of samples,but displays relatively better performance while testing images with complicated background and diverse sources.

Image quality assessment method based on contrast sensitivity

2006, 26(8):  1857-1859. 

Asbtract ( )   PDF (728KB) ( )  

Related Articles | Metrics

By analyzing the contrast sensitivity of HVS and extracting the involved factors that influence the image quality with the help of wavelet transform,a new image quality assessment method was presented. According to the contrast sensitivity of HVS characteristics,namely spatial frequency-characteristic curve, 2D multi-level wavelet decomposition was applied to the test image for obtaining wavelet coefficients. From these coefficients, brightness, clarity and the relevant indicator in sub-bands were obtained. Then an arithmetical mean, which came from the geometric mean of these three measures multiplying sub-band weighted coefficients, was used as final comprehensive assessment indicator. The experimental results show that the method can reflect people's subjective feelings in a better way compared with other traditional methods.

Fast intra-prediction mode decision algorithms in H.264

2006, 26(8):  1860-1862. 

Asbtract ( )   PDF (591KB) ( )  

Related Articles | Metrics

Intra-prediction coding and mode decision based on Rate-Distortion Optimization (RDO) technique were used in the H.264,which raised the compression ratio,but increase the computing complexity as well. Intra-prediction coding includes Intra_4×4 and Intra_16×16 prediction mode. Based on the analysis of the features of Intra_4×4 and Intra_16×16,a threshold was proposed to save computation. Then, the algorithms for intra-prediction mode decision was modified. The experimental results show that the proposed algorithm can increase the speed of intra coding significantly with negligible loss of PNSR and increase of bite-rate.

Enhanced advanced pyramid broadcasting with scalable capability

2006, 26(8):  1863-1865. 

Asbtract ( )   PDF (1097KB) ( )  

Related Articles | Metrics

Mirrored-Pyramid Broadcasting (MPB) can reduce the client buffer requirement of Pyramid Broadcasting (PB), and Enhanced Mirrored-Pyramid Broadcasting (EMPB) and Advanced Pyramid Broadcasting (APB) are the improved ones. But APB does not have the flexibility or best performance. Based on the analysis of these schemes, a new broadcasting scheme of SAPB was proposed. It can meet the various client buffer requirements, and use a scalable parameter to get the scalable capability and better performance.By adopting this scheme, VoD system will be more flexible and adaptive.

New approach for colortexture segmentation of natural images based on the distribution evenness of colors

2006, 26(8):  1866-1869. 

Asbtract ( )   PDF (1228KB) ( )  

Related Articles | Metrics

Many present color-texture segmentation methods are either too computing complicated or lack of smoothness in the final results. To solve these problems, a new approach for color-texture segmentation of natural images was proposed, which was based on the distribution evenness of colors and called ISBEC for short. Firstly, input image color was quantized. Then the resulted index map was used to perform simultaneous color segmentation and multi-scale texture analysis. Finally, these results were combined together, and over-segmented regions were merged. ISBEC can also be used to segment gray-scale images if they are converted to color ones in which three color components of each pixel are identical. The experimental results demonstrate the effectiveness of ISBEC for color-texture segmentation of natural images.

Ear recognition based on feature fusion and Bagging neural network

2006, 26(8):  1870-1872. 

Asbtract ( )   PDF (686KB) ( )  

Related Articles | Metrics

A new ear recognition method based on feature fusion was presented. Firstly,ear projection coefficient features were extracted by NMF methods.Then these features were combined with the ear rotation invariance Zernike moments features so as to get a new feature which had higher discriminating power. With this new feature，a Bagging neural network was built up by using bagging algorithm that improved classification accuracy and generalization of neural network. At last,the results of some comparative experiments show that this ear recognition method can get higher recognition rate.

Depth measurement of the object based on the corresponding points matching

2006, 26(8):  1873-1875. 

Asbtract ( )   PDF (663KB) ( )  

Related Articles | Metrics

A method for measuring the depth of an object was proposed based on the corresponding points matching. Firstly, the transform relation between physics image coordinate system and pixel coordinate system was presented. Secondly, the matching of the same object in two different images was achieved by use of the similarity calculated through the correlation algorithm of two images. Thirdly, from the viewpoint of the corner of the object, a method for matching the corresponding points was presented based on the epipolar restriction of corresponding points. By making use of the disparity of corresponding points, the depth information was computed based on the images that were obtained from the parallel dual cameras. Simulation results demonstrate the effectiveness of this algorithm.

Applied research of Gaussian maximum likelihood classification in hyperspectral classification

2006, 26(8):  1876-1878. 

Asbtract ( )   PDF (636KB) ( )  

Related Articles | Metrics

The relationship between Gaussian maximum likelihood classification error and Bhattacharyya distance was analyzed, and the addition property of Bhattacharyya distance was enumerated under uncorrelated features condition. Based on such analyses, a new feature selection algorithm was derived. This algorithm adopted the relative Bhattacharyya distance summation of each feature as the criterion function to select the features which contributed more to the reduction of classification error. These features then could be used for Gaussian maximum likelihood classification. Adopting AVIRIS data, the experimental results verify the effectiveness of this algorithm.

Normalized rgb color model based shadow detection

2006, 26(8):  1879-1881. 

Asbtract ( )   PDF (735KB) ( )  

Related Articles | Metrics

A shadow detection scheme based on normalized rgb color model was proposed.Firstly, the photometric invariance of rgb color model was studied and some of its important properties were deduced. Combining the cues of moving cast shadow on brightness and chromaticity successively, this proposed shadow detection algorithm detected candidate shadow regions in rgb color space; then, a post-processing by exploiting region-based geometry information was employed to exclude pseudo shadow segments. Results were presented for several video sequences representing a variety of illumination conditions and ground materials when the shadows were cast on different surface types. These results show that this approach is robust to widely different background and illuminations.

Robust adaptive spatio-temporal video denoising algorithm based on motion estimation

2006, 26(8):  1882-1883. 

Asbtract ( )   PDF (593KB) ( )  

Related Articles | Metrics

A robust adaptive spatio-temporal video denoising algorithm based on motion estimation was presented. To adpot the adaptive Wiener filter before the motion estimation promoted the performance of the block matching. The extra judgment based on small block and the adoption of Duncan filter promoted the robustness of the motion estimation. Experimental results demonstrate the effectiveness of this algorithm.

Database Technology

Mixed EM algorithm for large data sets

2006, 26(8):  1884-1887. 

Asbtract ( )   PDF (745KB) ( )  

Related Articles | Metrics

EM algorithm often needs great computational costs. And its computing is inefficient when the data sets are large.A mixed EM algorithm based on partial Esteps method was presented which can reduce the intensity of computation，make it adapted to the scale of data sets better and have the standard convergence guarantee of EM. It is verified that the mixed EM algorithm can reduce computational costs evidently through its application to large data sets.

New mercerkernel based Fuzzy clustering algorithm with attribute weights in feature space

2006, 26(8):  1888-1889. 

Asbtract ( )   PDF (454KB) ( )  

Related Articles | Metrics

Considering the imbalance between the attributes fully, a new weighted fuzzy kernelclustering algorithm named WFKCA was presented. WFKCA performs clustering in high feature space mapped by mercer kernels. Lots of examples demonstrate that WFKCA is a useful clustering tool.

Research on space allocation strategy of the volume holographic data storage

2006, 26(8):  1890-1893. 

Asbtract ( )   PDF (888KB) ( )  

Related Articles | Metrics

Through analyzing physical addressing mode of volume holographic data storage (VHDS), two-dimensional addressing strategy of logical block address was given, by which the data-block of the file was allocated. By researching the continuity of file data-block allocation on two-dimensional logical block address, the concepts of “sequence continuity” and “stochastic continuity” were presented. Based on allocation quadruple-tree which was adopted as framework of the two-dimensional logic area, the allocation algorithm for multiple block was designed. On the prototype system of VHDS, testing of the allocation strategies of data-block in the “file” level was made. The testing results prove that two-dimensional allocation strategy can make file data-blocks to be allocated more consecutive.

Over-sampling algorithm based on preliminary classification in imbalanced data sets learning

2006, 26(8):  1894-1897. 

Asbtract ( )   PDF (754KB) ( )  

Related Articles | Metrics

To significantly improve the classification performance of the minority class, an over-sampling algorithm based on preliminary classification was presented. Firstly, preliminary classification was made on the test data in order to save the useful information of the majority class as much as possible. Then the test data that were predicted to belong to minority class were reclassified to improve the classification performance of the minority class. Using the data sets provided by University of California, Irvine, the new algorithm was compared with synthetic minority over-sampling technique and under-sampling method. The experimental results show that the new algorithm performs better than the others in terms of the classification performance of the minority class and majority class.

Algorithm for mining calendar-based temporal association rules

2006, 26(8):  1898-1899. 

Asbtract ( )   PDF (512KB) ( )  

Related Articles | Metrics

An efficient algorithm for temporal association rules based on calendar patterns was presented. A user-given calendar schema was adopted to specify the interesting rime intervals as calendar patterns. Then database was scanned once to find all frequent 2-itemsets and their 1-star calendar patterns. Aggregation property and Apriori property were utilized to find all candidate patterns. Finally, calendar-based temporal association rules were obtained through scanning. The experimental results indicate that this proposed algorithm is feasible and efficient.

Horizontal union algorithm of concept lattice based on Congener concept

2006, 26(8):  1900-1903. 

Asbtract ( )   PDF (723KB) ( )  

Related Articles | Metrics

Congener concept was introduced into the lattice union algorithm, which was used to update each child node in less time according to the order relation of the concepts. It was further testified that to insert a small lattice into a large one would spend less time. The test results show that this algorithm makes better improvement in time complexity than the horizontal union algorithms of common universe concept lattice found in related literature.

Clustering algorithm of nominal data based on entropy

2006, 26(8):  1904-1905. 

Asbtract ( )   PDF (386KB) ( )  

Related Articles | Metrics

Through analyzing the characteristics of nominal data, clustering algorithm of nominal data based on entropy and hierarchical method was proposed. In this algorithm, similarity between objects was measured by using entropy and similarity threshold was calculated directly by using data. The experimental results show that this algorithm is feasible and effective.

Dynamic optimal control for aperiodic soft real-time system

2006, 26(8):  1906-1908. 

Asbtract ( )   PDF (583KB) ( )  

Related Articles | Metrics

The aperiodic soft real-time system with random task arrival and unpredictable task execution time was studied, and a dynamic optimal control method for DM scheduling algorithm based on the multiple priorities queuing system was proposed. The method can statistically guarantee the schedulability of soft real-time systems, and meanwhile it can guarantee the Quality of Service (QoS) of incoming tasks and improve the throughput of system. Finally, an example show that this method is capable of improving the actual utilization and decreasing the deadline miss ratio of system. Therefore, it is an efficient workload management method.

Component-based dynamic updating software system

2006, 26(8):  1909-1911. 

Asbtract ( )   PDF (687KB) ( )  

Related Articles | Metrics

Based on the analysis of the key issues that affect dynamic updating, a component-based dynamic updating mechanism which obtained a safe updating point by interposing interceptor dynamically was presented, and the hot-swapping of dynamic components was achieved. The main procedure included interposing interceptor, blocking invocation, transferring state and redirecting requests. This design makes the application has better feasibility, expansibility and operation performance.

Property patterns of linear temporal logics

2006, 26(8):  1912-1915. 

Asbtract ( )   PDF (542KB) ( )  

Related Articles | Metrics

In the applications of model checking, the software properties are usually depicted by Linear Temporal Logic (LTL) formulas. The descriptions of software properties based on LTL formulas have common patterns although different applications background have to depict different software properties. The common property patterns of LTL formulas was abstracted in two dimensions, which one was according to the properties depicted by the developers, for example, liveness and safeness, and the other was according to the scope of formulas. Property pattern is a manner that finds the common solution to the common problems in software specification. At last, the application of Property patterns of LTL formulas was given in SPIN tool.

MPI based parallel computation of matching cube

2006, 26(8):  1916-1918. 

Asbtract ( )   PDF (558KB) ( )  

Related Articles | Metrics

According to PCAM, a parallel algorithm was figured out to accelerate the computation of matching-cube for stereoscopic vision. The computation of matching-cube was divided by vertical coordinate into many sub-computations, which minimized the communication between computing nodes. By assigning these computation jobs of different disparity values to multiple computing nodes and gathering all these DSI to root node, a matching cube was obtained. A relatively accurate computational time complexity modal was built on CPU cycles to analyze the performance on different platforms, which was very important to real-time applications. As the data-ependence during computation was very low, a nearly linear speedup could be obtained on MPI cluster parallel platform.

Study on de-pipeline algorithm for software-pipeline of IA-64

2006, 26(8):  1919-1921. 

Asbtract ( )   PDF (679KB) ( )  

Related Articles | Metrics

Software pipelining is a loop scheduling technique which extracts instruction level parallelism by overlapping the execution of several consecutive iterations. But in reverse engineering, it causes some difficulties to reverse translation.A de-pipelining algorithm based on the IA-64 architecture was proposed. This algorithm reversely converted the optimized assembly code of a software-pipelined loop to a semantically equivalent sequential counterpart. The experimental results have verified the validity of the proposed de-pipelining,which provides the basis for processing software pipelining in the binary translation.

Research of the algorithms for task cooperation execution based on multi-agent system

曾锡山

2006, 26(8):  1922-1924. 

Asbtract ( )   PDF (739KB) ( )  

Related Articles | Metrics

To improve the capability of a system to solve problems and accomplish certain tasks, we could decompose and distribute the tasks that the system needs to complete by means of software, besides improving the hardware environment. A heuristic algorithm for the task decomposition and distribution was proposed, and analysis and evaluation of the algorithm was made as well. The results show that this algorithm can effectively advance the task execution efficiency, and the algorithm conforms to compound poisson distribution.

Artificial intelligence

Research of automatic translation based on Re-Engineering

2006, 26(8):  1925-1927. 

Asbtract ( )   PDF (496KB) ( )  

Related Articles | Metrics

By analyzing the operating principle of the translator Bogart which integrates Re-Engineering technique and comparing it with the traditional translator, an improved storage method was suggested. This method adopted a new language processing system which supported the definition of program language. The test results indicate that the method can significantly improve the efficiency of automatic translation.

Simulation study on the parameters optimization of radial basis function neural network based on QPSO algorithm

2006, 26(8):  1928-1931. 

Asbtract ( )   PDF (740KB) ( )  

Related Articles | Metrics

Coping with such limitations of Particle Swarm Optimization (PSO) algorithm as finite sampling space, being easy to run into local optima, a new Radial Basis Function Neural Network (RBF NN) training method based on Quantum-behaved Particle Swarm Optimization (QPSO) algorithm was proposed. A multidimensional vector composed of RBF NN parameters was regarded as a particle in this algorithm to evolve. Then, the feasible sampling space was searched for the global optima. The simulation results show that this learning algorithm has easier computation and more rapid convergence compared with other traditional learning algorithms. And due to the characteristic of the algorithm model, its global convergence ability is better than the one based on PSO.

Channel selection in single-trial evoked potentials estimation

2006, 26(8):  1932-1934. 

Asbtract ( )   PDF (587KB) ( )  

Related Articles | Metrics

Adopting the “Imitatingnaturalreading” elicited potentials as carriers between brain and computer, the evoked potentials were extracted by support vector machines. Taking the signals from the four EEG recordings as features and setting the time length of signal in a trial as 300ms, the classification rates were tested on average with three time intervals, namely, 100ms~400ms, 200ms~500ms,and 300ms~600ms,for every subject respectively, and got the best classification accuracy of 95.9%(subject M),94.3%(subject H),and 93.8%(subject T) in three subjects respectively. The results are good to simplifying the design of our BCI.I.

Strategy and algorithm for automated negotiations between multi-agent

2006, 26(8):  1935-1937. 

Asbtract ( )   PDF (607KB) ( )  

Related Articles | Metrics

To solve the problems of long processes and high cost for traditional negotiations, three kinds of formulized curve clusters were presented to describe strategies which were adopted by Agents in negotiations. The strategies were named anxious, economical and eclectic. An automated negotiating algorithm was also proposed, which was an integration of the strategies both in mathematics and computer programming. And then two examples were used to test negotiating strategy and proposed algorithm. Experiments show that the proposed negotiating strategy and algorithm are more conformable to human's intuitive thinking than Pillatt's. The automated negotiating strategy and algorithm may find many applications in e-Business transactions, such as B2B and B2C.nd B2C.

New method of hierarchical reinforcement learning

2006, 26(8):  1938-1939. 

Asbtract ( )   PDF (565KB) ( )  

Related Articles | Metrics

A novel method of hierarchical reinforcement learning which named OMQ was presented by integrating Options into MAXQ. In OMQ, MAXQ was used as the basic framework to design hierarchies experientially and learn online, and the Option was used to construct hierarchies automatically. The performance of OMQ was demonstrated in taxi domain and compared with Option and MAXQ. The simulation results show that the OMQ is more practical than Option and MAXQ in partially known environment.

Rapid convergence algorithms for weight values updating based on BP network

2006, 26(8):  1940-1942. 

Asbtract ( )   PDF (567KB) ( )  

Related Articles | Metrics

To solve the slow convergence of standard learning algorithm in BP network,two rapid convergence algorithms were suggested for weight values updating. One is rapid transmission algorithm based on gradient change rate. The other is flexible transmission algorithm based on gradient orientation. The two algorithms were simulated and compared in Game Style Training System for Mine Accident Rescuing. Here the algorithms would help game roles learn to estimate the danger degree according to ingredients of mine air, and then help trainees or biorobots take corresponding actions. The simulating results show that shorter convergence time is taken for the two algorithms than the standard algorithm.

Application of adaptive artificial immune algorithm to data mining

2006, 26(8):  1943-1946. 

Asbtract ( )   PDF (731KB) ( )  

Related Articles | Metrics

In the artificial immune algorithm for clustering analysis, its suppression and stimulate thresholds determine cluster precision and network population scale. Theses thresholds adopt fixed value that is decided according to the problem characteristic itself and user's experience. However, this modus operandi results in narrow application situation and is dependent heavily on problem characteristic itself. Therefore, an adaptive artificial immune algorithm for clustering was proposed. This algorithm could achieve final network structure well matching the crude data feature and relieve the dependence on problem characteristic itself, because its thresholds were obtained from the dynamic immune network structure and were adapted well to the entire network structure during the process of evolution. Experimental results demonstrate its effectiveness.

Hybridized optimization genetic algorithm for multicast routing problem

2006, 26(8):  1947-1949. 

Asbtract ( )   PDF (775KB) ( )  

Related Articles | Metrics

The mathematic model of dynamic multicast routing with Nodes delay and delay variation constraints was analyzed. Based on the model, an optimization algorithm called GP-GA was proposed by hybridizing Gene-Pool (GP) with traditional Genetic Algorithm (GA).This method made use of the gene-pool to save the solutions during the process so as to direct the remaining evolution. In the mean time,the crossover and mutation operator were improved to accelerate the convergence speed. Considering that the problem may be trapped by local optimization easily, the evolution strategy based-on "reserved and non-reserved" was also constructed to enhance the ability of finding optimal solution and decrease the probability of "premature" phenomena commendably. A great number of simulations demonstrate that the probability of GP-GA converging optimal solutions is higher than general GA, and the algorithm is also effective for being adjusted to the dynamic multicast routing.

Improvement of fuzzy-set-based ant colony clustering algorithm

2006, 26(8):  1950-1952. 

Asbtract ( )   PDF (529KB) ( )  

Related Articles | Metrics

LF algorithm was improved and a new method of ant colony clustering based on fuzzy set theory was put forward. Firstly, the average distance was defined. Then, fuzzy set theory was introduced into the concept of similarity, and the membership function of similar degree between a data object and its neighbor was defined. Finally, the pickup or drop of this data object was determined by the comparison between degree of membership and confidence level λ.The new method overcomes such shortcomings in LF algorithm as that dissimilar data object may not be picked up and similar data object may not be dropped, and simplifies LF algorithm.

Investigation of automatic generation of class-object test data based on genetic algorithm

2006, 26(8):  1953-1955. 

Asbtract ( )   PDF (609KB) ( )  

Related Articles | Metrics

In the automatic generation of test data, there is no sound solution to generating class-object test data in the practical applications. To overcome this shortcoming, automatic generation of class-object test data on Genetic Algorithm (GA) was investigated. On the basis of Extended Hamming Distance (EHD), a modified method was proposed for fitness function of generating class-object test data by using GA, and the feasibility of reducing the execution time was proved in theory. At last the method was applied to real system. The testing results show that the method significantly improves the efficiency and accuracy of generating class-object test data.

Improved evolutionary algorithm

2006, 26(8):  1956-1958. 

Asbtract ( )   PDF (687KB) ( )  

Related Articles | Metrics

A new evolutionary algorithm named GT-POPDM-PSO was proposed.It combined GT algorithm with PSO algorithm, adopting non-uniform mutation operator to improve local searching ability and stability. It is proved that this proposed algorithm outperforms the two algorithms proviously referenced and has better results for multi-modal functions in particular.

Improved real coded hybrid genetic algorithm

2006, 26(8):  1959-1962. 

Asbtract ( )   PDF (798KB) ( )  

Related Articles | Metrics

To solve such problems as premature convergence and slow evolving speed of the simple genetic algorithm during evolution, a comprehensive improved measure was put forward for the real coded genetic algorithm, including the netlike distribution of initial population creation, the best-keeping after operation of the genetic operators in each step, the improved dynamic crossover probability and dynamic self-adapting mutation probability. What's more, to replace the worst individual of current generation by the best one of the father generation was applied, and closed crossing avoidance as well. The numerical simulations show that the improved genetic algorithm is more effective in realizing the global optimization and promoting evolution efficiency, and has stronger adaptability in solving complex optimization problems.

Typical applications

Short guide to the display devices

Wei-zong QIAN

2006, 26(8):  1963-1967. 

Asbtract ( )   PDF (1196KB) ( )  

Related Articles | Metrics

In recent years, display devices have developed very rapidly and a great variety come to the market. Besides the computer terminal display, large-screen display supported by computer program has also been widely used. In order to provide an overall picture about the performance of various kinds of display devices, and help instruct to choose proper device in certain system, the basic structures of several commonly-used display devices were roughly introduced in this paper, as well as their operating principles, characteristics and general development situations.

ε－SVR mathematical model and algorithm for sale prediction problems

2006, 26(8):  1968-1971. 

Asbtract ( )   PDF (726KB) ( )  

Related Articles | Metrics

A mathematical model（LR_ε-SVR,NLR_ε-SVR） and an algorithm (ε-SVR_SP) were proposed on the basis of Support Vector Machine (SVM) regression, which could predict tobacco sale and sale tendency. Historical data and current context data were taken as the input, and such results as sale tendency in the future and the forecasting tobacco sales were outputted. By using kernel-based learning algorithm, the dimension of feature space in training process has been reduced, thus, the calculating complexity becomes lower and a small amount of training data will be sufficient. Simulated experiments, practical data set tests and comparative experiments with neural network show the advantages of the proposed approach in computation precision and efficiency.

Effective method for dam recognition from visible images

2006, 26(8):  1972-1974. 

Asbtract ( )   PDF (720KB) ( )  

Related Articles | Metrics

An effective method for dam recognition was proposed.First,the original image was sampled according to the resolution of source image. The large upstream water area of dam was obtained by using the gradient histogram and the region growth, and the large water area was expressed through the Freeman code. The downstream water was pursued and the possible dam region was divided preliminary. Then one group of features of dam was extracted. The algorithm has the characteristic of high recognition rate, good extension and so on. Experimental results show that this proposed method is effective for dam recognition in the visible images.

Application of improved multi-scale watershed algorithm in medical image segmentation

2006, 26(8):  1975-1976. 

Asbtract ( )   PDF (768KB) ( )  

Related Articles | Metrics

With regard to the over-segmentation of watershed algorithm and the characteristics of medical image, a new algorithm that can effectively enhance the edge of depth image and eliminate local minima was presented. Firstly, morphological gradient image was got by muti-scale filter, which can extract image edge effectively. Then, a new filter method based on image edge was proposed, which can eliminate local minima. Finally, procedures and results of the algorithm were introduced. Experimental results demonstrate that watershed transformation with the proposed algorithm produces very good segmentations, even without region merging.

Analog circuit faults diagnosis based on clustering preprocess and SVM

2006, 26(8):  1977-1979. 

Asbtract ( )   PDF (665KB) ( )  

Related Articles | Metrics

Support Vector Machines (SVM) is invented from binary classification problem and it is difficult to be used directly in analog circuit faults diagnosis because of multi-fault modes and overlapped recognition areas. A diagnosis method was proposed by using dynamic clustering method as a preprocessor of SVM. Firstly, Fuzzy C-Means (FCM) algorithm was used to generate father and mother clustering gather, then their children were generated by binary tree structure, and every clustering would train a SVM network to be stored. The simulation experimental results show that accuracy of this diagnosis can be higher than 99%.

New effective method for spam filtering

2006, 26(8):  1980-1982. 

Asbtract ( )   PDF (587KB) ( )  

Related Articles | Metrics

A new effective method for spam filtering according to the principle of granularity was presented. First, this method divided spam class and legit class in train corpus into four small classes, and four center vectors were obtained. In the view of the principle of granularity, smaller granularity is used to describe knowledge in train corpus. When filtering,the new E-mail was compared with four center vectors respectively to decide which class it belonged to. This method was tested on spam corpus and compared with KNN. The results show that the new method has some advantages including high accuracy, high speed of filtering and so on.

Implementation of Web information extraction system based on similar pages

2006, 26(8):  1983-1986. 

Asbtract ( )   PDF (734KB) ( )  

Related Articles | Metrics

The core algorithm of RoadRunner was analyzed. After analyzing the deficiencies of RoadRunner, a Web information extraction system based on similar pages was designed and implemented. The system architecture was introduced, then the key techniques, such as the method for getting similar Web pages, reliably dealing with Web nosy blocks and automatically deducing rules for extracting data items were presented.

Implementation of P2P resource management technology based on JXTA

2006, 26(8):  1987-1989. 

Asbtract ( )   PDF (719KB) ( )  

Related Articles | Metrics

A method of resource management for P2P was put forward. It adopted communication methods of JXTA platform and implemented the system of resource management and data accessing based on Distributed Hash Table (DHT). By the communication message format designs, the system realized the cooperative searches among nodes.

FPGA implementation analysis of Fuzzy CMAC and its control application

2006, 26(8):  1990-1992. 

Asbtract ( )   PDF (589KB) ( )  

Related Articles | Metrics

An FPGA implementation structure of a FCMAC（Fuzzy CMAC） was proposed. Compared with other neural networks implemented by FPGA, it contains the learning algorithm that can be employed to realize the online learning. First, the model of FCMAC was given and the relevant hardware modules were analyzed. Second, these hardware modules were implemented with VHDL. At last, the implemented FCMAC was used for control, and the corresponding controller was tested. The experimental results indicate that this method for hardware implementation of FCMAC is feasible, and the controller has high computing speed, high precision and better ability of antiinterfere. It is a new and efficient method to implement IP control module or intelligent control based on a single chip.

New text categorization method based on the frequency of topic words

2006, 26(8):  1993-1995. 

Asbtract ( )   PDF (593KB) ( )  

Related Articles | Metrics

The word frequency matrix currently used in text categorization is characterized with high dimensionality and excessive sparsity. These two features caused some difficulties to computing. To solve this problem, according to the search engine users' selections, a new text categorization method based upon the feature of topic words frequency was proposed. This approach was designed to filter new concept topic words by statistical method, and then the FCM clustering algorism was applied to the documents, using the frequency of topic words rather than the frequency of single word as the feature.This method performs well in the experiment. Furthermore, this method was compared in many aspects with a text categorization method based on keyword clusters, and some useful conclusions about implementation and application were reached.

Application of data mining in customer relation management

2006, 26(8):  1996-1997. 

Asbtract ( )   PDF (637KB) ( )  

Related Articles | Metrics

The similarity formula of clustering and assessment criteria were determined by using standardized customer data. The automatic clustering of customers was also realized by using the layer-agglomeration approach and K-average approach. On the basis of balancing algorithm efficiency and accuracy of clustering, improved formula of clustering distance and K average algorithm approach were presented. Satisfactory results were achieved.

Digital watermarking technique based on 2-D Barcode

2006, 26(8):  1998-2000. 

Asbtract ( )   PDF (677KB) ( )  

Related Articles | Metrics

After analyzing the digital watermarking techniques for binary images, the conclusion that a successful information hiding technique must be based on the special structure of the hidden media was made. Aiming at a specific binary images——PDF417 2D Barcode, a new information hiding method based on boundary shift was brought up. Being combined with RS-code, this method can be widely used in 2D Barcode digital watermarking techniques.

Application of simulated annealing algorithm in multi-objective airline crew rostering system

2006, 26(8):  2001-2004. 

Asbtract ( )   PDF (723KB) ( )  

Related Articles | Metrics

At present, the crew rostering system is based on the manual way to accomplish, and there are many restricted terms that need to be considered. Thereby it is necessary to make use of computer technology to make the crew rostering automatic and reasonable. The crew rostering system was roughly introduced, and the method about how to build models for crew rostering system was described in detail. The multi-objective of crew rostering system was achieved by simulated annealing algorithm. The actual data of airline company were used to testify its reasonableness and effectiveness.