Journal of Computer Applications

Time-varying half-edge graph：a novel topology model for attributed relationships of Internet resources

2007, 27(3): 512-513.

Asbtract ( )

PDF (751KB) ( )

Related Articles | Metrics

The half-edge concept is lead in to describe the attributes of Internet resources and to establish an unifying description frame for the characteristic of various resources in network environment. Expanding the normal representation of resources relationships, time-varying half-edge graph topology model describes the attributed relationships of Internet resources. According to scale-free networks theory, an actual algorithm of time-varying half-edge graph is finally presented. Time-varying half-edge graph demonstrates the dynamic topology of attributed relationships and can be well expanded. The scale-free characteristic of real network may be reappeared.

Three routing algorithms for publish/subscribe system in mobile Ad Hoc networks

2007, 27(3): 514-515.

Asbtract ( )

PDF (778KB) ( )

Related Articles | Metrics

Routing algorithms is a key issue in largescale publish/subscribe system under Ad Hoc network. The protocol of publish/subscribe system was introduced. Three contentbased routing algorithms of publish/subscribe system in mobile Ad Hoc networks were proposed. Their performances were evaluated by network simulations. The results show that RAFT-CBR provides both faulttolerance and reliability, and it can guarantee 100% information delivery at the expense of transmission delay; CBR, with low delay, is more suitable for distributed publish/subscribe protocol in wired networks at the expense of reliability; FT-CBR, with low delay, extends CBR to provide faulttolerance.

A localization system based on DV-Hop localization algorithm and RSSI ranging technique

2007, 27(3): 516-518.

Asbtract ( )

PDF (939KB) ( )

Related Articles | Metrics

As an assistant localization method, received signal strength indicator (RSSI) ranging module was added into the DVHop localization algorithm to resolve its problems existing in experimental environment. In order to implement the localization system, first, the RSSI model was built up; then, this model was applied to control DVHop localization process from anchor nodes and nonanchor nodes respectively. It is proved that improved localization system improves system stability and localization precision with a little sacrifice on computing complexity. Therefore, it can be applied into wireless sensor networks.

An adaptive Retransmission Scheme over CDMA2000 Wireless Link Layer

2007, 27(3): 519-521.

Asbtract ( )

PDF (768KB) ( )

Related Articles | Metrics

When TCP application is used in 3G wireless packet data service, its data transmission performance is badly influenced by the frame error rate (FER) of physical layer. The link layer retransmission technology is an important measure to improve the TCP data transmission performance. The effect of the link layer retransmission schemes of CDMA2000 on TCP performance was mainly studied. An adaptive retransmission scheme over CDMA2000 wireless link layer was put forward. The simulation of this algorithm using NS2 shows that this scheme can greatly improve the TCP throughput.

Multi-path constraintbased load balancing methods for MPLS traffic engineering

2007, 27(3): 522-524.

Asbtract ( )

PDF (550KB) ( )

Related Articles | Metrics

Two multi-path constraintbased load balancing methods were proposed for Internet traffic engineering of MultiProtocol Label Switching (MPLS), importing the idea of load balancing at the beginning of path setup. In a normal constraintbased shortest path first (CSPF) routing algorithm, there is a high probability that it can not find a feasible path through networks for a large bandwidth constraint. The proposed methods can divide the bandwidth constraints into two or more subconstraints and find a constrained path for each subconstraint, if there is no single path satisfying the whole constraints. Simulations show that they enhance the success probability of path setup and the availability of network resources.

Blind scheme for channel estimation with high accuracy

2007, 27(3): 525-527.

Asbtract ( )

PDF (545KB) ( )

Related Articles | Metrics

A blind least mean square (LMS) scheme was proposed. The new scheme could trace the channel characteristic adaptively through timevariance gain design. And the decision information function in receiver was also adopted to correct the gain. Simulation shows that the new scheme has better performances with high accuracy and fast convergence than that of the LMS method, in terms of different time delay spread, time and signal noise rate (SNR).

A chord-based optimization method for distributed spatial data storage

2007, 27(3): 528-530.

Asbtract ( )

PDF (615KB) ( )

Related Articles | Metrics

Concerning the locality characteristic and positional correlation of spatial data accessing, a peertopeer (P2P)based method for distributed spatial data storage was proposed, which extended Chord. The concept of grouping nodes was introduced. The size of group could be adjusted according to the load of nodes by exploiting it, so the load balancing properties of the system were retained. Both the theoretic analysis and the simulation experiments show that it has higher efficiency in spatial range query than Chord.

Flow characteristics analysis based on P2P application

2007, 27(3): 531-533.

Asbtract ( )

PDF (852KB) ( )

Related Articles | Metrics

To understand the characteristics of peer-to-peer (P2P) traffic in network, a P2P file sharing application traffic was analyzed. The analytical results show that traffic of Maze has the following characteristics: the traffic is burst; the locality is less evident than that of WWW and FTP; the flow size follows the distribution different from WWW; flow duration time is longer than that of WWW and FTP distinctly; and the proportion of big flows in Maze traffic is higher than that of WWW and FTP.

Robust acess control rotocol of P2P network

2007, 27(3): 538-540.

Asbtract ( )

PDF (809KB) ( )

Related Articles | Metrics

Centralized authority isn’t appropriate in peer-to-peer systems, because nodes in peer groups are equal inherently. This Video denoising algorithms based on discrete wavelet transform (DWT) do not allow real-time processing due to their huge computational complexities. To overcome the shortcoming, a new real-ime algorithm for video denoising based on lifting scheme was proposed. First, each frame of the video was decomposed into multi-level wavelet to obtain the scaling coefficients and the wavelet coefficients by using lifting scheme. Then, each level of wavelet coefficients was filtered by soft-threshold algorithm. After the inverse wavelet transform, temporal filter was applied on the filtered video to enhance the denoising effect further. Experiments show that this algorithm has better quality in terms of real-time processing and denoising effectiveness.

Resource scheduling algorithm for computing grid based on utility optimization

2007, 27(3): 541-542.

Asbtract ( )

PDF (562KB) ( )

Related Articles | Metrics

Concerning the time and budget restriction in resource scheduling, a utility function model was proposed by analyzing the relationship between the time and budget of task and the user utility function. Then, based on the utility function model, a utility optimization based heuristic resource scheduling algorithm was put forward for task farming application model. Simulation results show that our proposed algorithm provides user with more utilities than costtime optimization algorithm.

Design and implementation of embedded VPN gateway system

2007, 27(3): 543-545.

Asbtract ( )

PDF (575KB) ( )

Related Articles | Metrics

After analyzing the principles of IPSec data processing, a proposal that the control plane tasks in IPSec was processed by host processor, as well as the data plane tasks by slave processor was put forward. Based on S3C2510 network processor and μCLinux operating system, the embedded IPSec VPN gateway system was also implemented.

3D Reconstruction of Knee Joint based on 2D Contour Series

2007, 27(3): 546-549.

Asbtract ( )

PDF (888KB) ( )

Related Articles | Metrics

A tetrahedral mesh generating method based on 2D contour series is proposed to reconstruct 3D medical geometric models of genual tissues. Firstly, contours in the chosen slices are extracted and branch-matched. Secondly, planar domains bounded by these contours in slices are triangulated into 2D triangular meshes. Finally, 3D tetrahedral mesh is generated by forming a set of tetrahedrons between every two adjacent slices based on the corresponding triangular meshes. The proposed method was applied to 3D model reconstruction for the knee joint virtual surgery simulation system, generating several 3D meshes, take, for example, the model of the genual thighbone is a mesh with 494 nodes and 2046 tetrahedral element, and the model of the genual fat tissue is a tetrahedral mesh with 2854 nodes and 14011 tetrahedrons. There 3D models are successfully used to simulate the knee joint surgery, which proves the feasibility and effectiveness of the proposed 3D model reconstruction method.

Design of configuration management for UPnPbased digital home gateway

2007, 27(3): 546-549.

Asbtract ( )

PDF (888KB) ( )

Related Articles | Metrics

The operational and manageable application mode of digital home gateway was proposed. This kind of home gateway breaks the limit to the simplified business of existing home gateway. It not only provides interoperability of digital home network devices, but also extends broadband services directly to home network devices. It implements automatic configuration servers remote configuration management for home network device, facilitates interpenetration and intercommunication of telecom network and home network. Automatic configuration and terminal management technology for multiple services in or out of home network, is combined with UPnP protocol framework, and provides reusable software design pattern. Home gateway realizes "zero configuration" of network connection and "transparence" of data transmission. Furthermore, it can provide operational and manageable service capability for telecom networks. Therefore, it conforms to the development trend of digital home.

Minimal attribute reduction algorithm based on particle swarm optimization with immunity

2007, 27(3): 550-552.

Asbtract ( )

PDF (710KB) ( )

Related Articles | Metrics

The problem of finding a minimum attribute reduction of a decision table was converted into a 01 combinatorial optimization problem. A proper fitness function of a particle for this problem was defined. A binary particle swarm optimization algorithm with immunity (BPSOI) was proposed to deal with the problem. Experimental results based on UCI repositories show that the proposed algorithm can quickly converge to a better solution in less runtime and within fewer generations. Its effectiveness and feasibility are also verified in the comparison with several other algorithms.

Algorithm for Sammon's nonlinear mapping based on fuzzy kernel learning vector quantization

2007, 27(3): 553-555.

Asbtract ( )

PDF (591KB) ( )

Related Articles | Metrics

An new algorithm for Sammon's nonlinear kernel mapping based on reliable and stable fuzzy kernel learning vector quantization was presented. The data space was mapped to high dimension feature space with Mercer kernel function, and fuzzy kernel learning vector quantization (FKLVQ) was done on the feature space to obtain the effective and stable clustering weight vectors. Finally Sammons nonlinear kernel mapping only for the data points and the clusters was executed on the output space and the feature space, thus reducing computational complexity and preserving the distance resemblance between the clusters and the data points from the data space to the output space. Simulation results demonstrate the reliability and stability of the proposed algorithm.

Dynamic mining decision rules based on partial granulation

2007, 27(3): 556-558.

Asbtract ( )

PDF (512KB) ( )

Related Articles | Metrics

Concerning decision table, the concept of partial granulation was introduced, and an algorithm of dynamic mining of decision rules based on partial granulation was proposed. By making use of the advantage of characterizing granularity of partial granulation, different from the single equivalence relation in the classical rough set theory, this algorithm can study one decision table from different angles and different hierarchies. The case study shows that this algorithm is effective.

An improved density-based outlier mining algorithm

2007, 27(3): 559-561.

Asbtract ( )

PDF (1186KB) ( )

Related Articles | Metrics

Based on the characteristic that outliers are not included in the appointed neighborhood of inliers, an improved algorithm for outlier mining was proposed. Data was judged whether it was included in the appointed neighborhood of inliers firstly. If the answer was negative, the number of data that was included in the appointed neighborhood was counted. Experimental results show that the improved algorithm is effective and efficient in outlier mining and it is faster than the original algorithm.

Mining fuzzy temporal association rules

2007, 27(3): 561-564.

Asbtract ( )

PDF (717KB) ( )

Related Articles | Metrics

With the help of fuzzy calendar algebra and fuzzy operators, it is easy to describe desired temporal requirements. Time intervals may have different weights according to their membership functions. Integrated with the idea of progression and increment, an algorithm combined with incremental mining and progressive counting called BFCTAR was proposed to mine fuzzy temporal association rules based on fuzzy calendar algebra. Theoretic analysis and experimental results indicate that this algorithm is efficient and feasible.

Pattern discovering of Web user access pattern based on maximal frequent path method

Cheng Lv Chu-yuan Wei Han-tao Zhang

2007, 27(3): 565-569.

Asbtract ( )

PDF (800KB) ( )

Related Articles | Metrics

As far as the issue on Web user access pattern is concerned, adopting Maximal Frequent Path (MFP) can mine more universal patterns. A new user access pattern tree named WUAP-tree was devised. Furthermore, a new algorithm named WUAP-mine was proposed for mining user access patterns, which was based on E-OEM model for page topological structure and users' browse path. The algorithm utilized WUAP-tree that could neither generate candidate sets nor use recursive ways. It could mine frequent Web users access patterns by scanning transaction database and output-depth-first traversing WUAPtree only once. The algorithm is very easy to query Web user access patterns from WUAPtree directly. At last, theoretical analysis and experimental results prove its effectiveness and efficiency.

Minimum error estimation using wavelet for time series similarity search

2007, 27(3): 570-573.

Asbtract ( )

PDF (916KB) ( )

Related Articles | Metrics

Wavelet transform is used as a dimensionality reduction method to permit efficient similarity search over highdimensional time series data. The traditional algorithms use the first k wavelet coefficients as an approximation of the original time series. But sometimes choosing the first k coefficients is not the best when approximating the original time series. Perhaps choosing other k wavelet coefficients is better than choosing the first k. The theorem was given to better approximate the original time series set. The main idea of the theorem was to choose the k columns of the wavelet sequences set which had the maximum square sum. The experiments show that it can better reduce the relative error compared with original algorithms.

Heterogeneous data integration method based on Excel and DTS

2007, 27(3): 574-576.

Asbtract ( )

PDF (884KB) ( )

Related Articles | Metrics

To integrate the data in heterogeneous databases, based on the design mode of SOAs three heterogeneous layers, the architecture of integrating application databases system for medium enterprises and small companies was proposed. This structure adopted two kinds of data transformation service technologies, Excel and DTS.

Implemention of an intelligent meta-search engine based on fuzzy integral algorithms

2007, 27(3): 577-579.

Asbtract ( )

PDF (600KB) ( )

Related Articles | Metrics

An intelligence metasearch engine based on fuzzy integral algorithms was implemented. The search agent was adopted because of the high parallelism and intelligence of the system. The scheduling of search sources was optimized by integrating the decision-tree algorithms and genetic algorithms, and then a new method of fuzzy integral was presented to merge the results. By the experiments on Web and results comparison with other search engines and other meta-searches in the same conditions, the meta-search engine presented in this paper has a better performance.

A virus immunization model based on community in large scale networks

2007, 27(3): 580-583.

Asbtract ( )

PDF (825KB) ( )

Related Articles | Metrics

With the development of computers and Internet, more and more people use E-mail. Viruses of E-mail have caused large damages. Traditional intentional immunization based on degree does not take the positions of infected nodes into account, but just protects the nodes which have high degree. The concept of community was introduced into the research of virus and immunization, and an immunization model based on community was proposed. According to the different infection stages of viruses, this model immunized infected communities or health communities. It slowed down the virus spread rate and kept the virus from spreading to more communities. Intentional immunization based on degree can not keep the virus in certain communities, and as a result the infected nodes diffuse in almost all communities. Immunization based on community can keep the virus in certain communities. These two models are different in the number of nodes that need to be protected, and in community infection rate. So, immunization based on community is different from that based on degree in terms of principle, and it is novel and effective.

Design and realization of identity-based mobile payment system

2007, 27(3): 584-586.

Asbtract ( )

PDF (551KB) ( )

Related Articles | Metrics

According to the bilinear property of the Weil pairing defined on elliptic curves and Eulers criterion, an identitybased signcryption scheme was proposed. The scheme could obtain the quicker velocity of encryption and decryption, distinguish the right message from the wrong message, and resist the attack of continuous cryptograph sending. This scheme could reduce the storing and managing cost of the public key. The signature size was only about a quarter of the GuillouQuisquater signature. Based on the signcryption algorithm and the properties of the mobile, a new mobile payment system was proposed. Its security and efficiency performances were also analyzed.

An authentication model of CDN system based on P2P - DPKI

2007, 27(3): 587-589.

Asbtract ( )

PDF (782KB) ( )

Related Articles | Metrics

Corresponding security mechanism must be offered in peer-to-peer (P2P) content distribution network (CDN), while keeping highly efficient Video-on-Demand (VOD) service. Traditional mechanisms based on C/S architecture do not agree with the distribution characteristic of P2P system. An authentication model named Distributed Public Key Infrastructure (DPKI) was proposed to guarantee the security of authentication as well as the performance of the system.

A peer-to-peer incentive mechanism based on PKI

2007, 27(3): 590-592.

Asbtract ( )

PDF (492KB) ( )

Related Articles | Metrics

To solve free rider and homogenization problems in peertopeer (P2P) network, an incentive mechanism based on public key infrastructure (PKI) and constructed P2P network was proposed. The mechanism not only encourages node to provide downloading function, but also benefits resources publisher. It restrains free riding behavior effectively, and alleviates homogenization.

Analysis and improvement of multi-secret sharing schemebased on general access structure

2007, 27(3): 593-594.

Asbtract ( )

PDF (542KB) ( )

Related Articles | Metrics

Peng proposed an efficient multi-secret sharing scheme based on general access structure in 2006. In his scheme, not only the shares held by the participants are reusable, but also the shares distributed by the dealer and the shadow shares provided by participants are verifiable. The security of Peng's scheme was analyzed and two attacking methods were proposed. It demonstrates that both of the shareauthenticating and shadowsecretauthenticating methods are insecure. Using our attacks, a dishonest dealer can distribute fake shares to participants, and malicious participants can easily forge fake shadow shares, so the honest participants will be cheated. At last, improvement was made to avoid those attacks.

A single sign-on scheme supporting double authentication method

2007, 27(3): 595-596.

Asbtract ( )

PDF (560KB) ( )

Related Articles | Metrics

The requirement of supporting multiple authentication in single signon scheme often complicates the realization of authentication protocol and crossrealm authentication. A flexible single signon scheme that supported certificate, password and their combination authentication was proposed. In this scheme, the protocol template and the temporary certificate ticket were adopted to avoid the repeat design of authentication protocols and simplify the trust models of crossrealm authentication.

An enhanced role-based access control model of database

2007, 27(3): 597-600.

Asbtract ( )

PDF (920KB) ( )

Related Articles | Metrics

The traditional role based access control model (RBAC) can not avoid some security flaws, because the components and the constraints of the model are too simple. Based on the traditional RBAC, a enhanced RBAC model (ERBAC) was brought forward, which expanded the components and the constraints. Besides, this new model features the separated administrative permissions, complete constraints and better utility.

A computational trust model based on MAS in E-commerce

2007, 27(3): 601-604.

Asbtract ( )

PDF (720KB) ( )

Related Articles | Metrics

Some trust problems come into being due to the dynamic and uncertain characteristics of E-commerce. Therefore, a computational trust model based on multi-Agent (MAS) was proposed to deal with the trust relationship in E-commerce. Compared with the present available trust model, this proposed one can deal with the fraud information independently. Meanwhile, aging factor was introduced to evaluate the trust relationship between agents. Simulation experiments demonstrate that the introduction of aging factor and trust precision can help evaluate the trust relationship more efficiently. What's more, the ontological property of trust was also considered in this model.

Audio zero-digital watermarking algorithm based on energy difference of wavelet coefficients

2007, 27(3): 605-607.

Asbtract ( )

PDF (543KB) ( )

Related Articles | Metrics

An audio zero-digital watermarking algorithm based on energy difference of wavelet coefficients was proposed. Original audio signal was decomposed by wavelet transform. And the energy differences of the approximation coefficients and detail coefficients were collected and classified into two classes according to the relationship between the energy differences and the thresholds. In this way, the watermarks were embedded and extracted. The restoration of the linear scaling was used to resist the influence of the linear scaling on time axis. The experimental results show that this algorithm is robust to many operations and attacks on digital audio signal, such as low-pass filtering, re-sampling, re-quantifying, echo jamming and MPEG encoding.

Research on MIPv6 secure firewall system

2007, 27(3): 608-609.

Asbtract ( )

PDF (617KB) ( )

Related Articles | Metrics

Concerning the problems of the firewall in mobile IPv6 (MIPv6) network, a new MIPv6's secure firewall system was proposed. A hybrid authentication method integrating digital signatures and certificates was introduced. It can authenticate unauthorized packets which come from the outside of the firewall effectively, and solve the secure problems of MIPv6 network commendably.

Secure key management scheme based on cluster architecture in mobile Ad Hoc networks

2007, 27(3): 611-613.

Asbtract ( )

PDF (588KB) ( )

Related Articles | Metrics

Mobile Ad Hoc Network (MANET) is a new wireless network in certain domain. With its unique characteristics, people pay more attention to its security coupling with its adaptable application scope. Combining the system private key and the cluster private key, a completely distributed secure key management scheme based on cluster architecture was proposed. It enhanced the authenticity, confidentiality, availability and robustness.

A proxy signatures scheme based on (t,n) threshold

2007, 27(3): 614-615.

Asbtract ( )

PDF (562KB) ( )

Related Articles | Metrics

A new proxy signature scheme based on (t,n) threshold was proposed by constructing the Lagrange polynomial interpolation. The scheme could strictly restrict the times of use of a warrant. Compared with the traditional onetime proxy signature, this scheme can effectively control the times of the signatures that the proxy signer can make, and reduce the time and lower the traffic.

Research on secure authentication based on session initial protocol

2007, 27(3): 616-618.

Asbtract ( )

PDF (848KB) ( )

Related Articles | Metrics

At present, most of the session initial protocol (SIP) authentication schemes only provide client-to-server authentication. And HTTP digest authentication is one of them. The weakness of this scheme in server spoofing and password guessing attacks was analyzed, and a more secure strategy was proposed which avoided these shortcomings. The experimental result illustrates that this proposed algorithm is effective.

A novel method to improve the performance of intrusion detection systems

2007, 27(3): 619-620.

Asbtract ( )

PDF (206KB) ( )

Related Articles | Metrics

The knowledge base is so huge that intrusion detection systems based on knowledge use much time to search the results. Hence a lot of packets are lost in the environment of high-speed network. In order to solve the problem, the feed back mechanism of the network traffic was introduced. Random projection algorithm was used to flexibly adjust the scale of the knowledge base during the pattern matching procedure. The real-time performance of intrusion detection systems was enhanced. The rate of lost packets and the false warning rate of the systems were reduced.

Information hiding scheme based on PE file resources section

2007, 27(3): 621-623.

Asbtract ( )

PDF (661KB) ( )

Related Articles | Metrics

According to the characteristics of the PE file resources section format, four kinds of redundancies in the resources section were summarized, and a new scheme of information hiding was proposed. The scheme synthesized the advantages about four kinds of redundancies. The information was dispersed into the expanded redundant space between the resources with the method of shift redundant space. This scheme enhanced the information imperceptibility and the resources section redundancies utilization rate, compared with other schemes of information hiding based on PE file.

Application of file system filter driver in network secure terminal

2007, 27(3): 624-626.

Asbtract ( )

PDF (620KB) ( )

Related Articles | Metrics

Design and implementation of a system named Network Terminal Device Information Anti-Leakage System (NTDIALS) was designed and implemented. To prevent the leakage of confidential information in network terminal, NTDIALS realized the realtime monitoring and dynamic encrypting and decrypting of information at the layer of file system filter driver, which was both safe in encryption and convenient in access control.

Negative-selection algorithm of mutation search in self space to train detector

2007, 27(3): 627-629.

Asbtract ( )

PDF (603KB) ( )

Related Articles | Metrics

To overcome the defects of the present negativeselection algorithm in fault detection, based on self-space, abnormity detection algorithm was improved. It trained detector according to the mutation mechanism of genetic algorithm and the principle of vaccine inoculation and clone selection. The simulation shows that under the precondition that the improved algorithm can generally cover the self-space, its computing complexity dramatically declines while its fault detecting efficiency greatly improves.

Differential evolution algorithm with position-order encoding for solving traveling salesman problem

2007, 27(3): 630-632.

Asbtract ( )

PDF (733KB) ( )

Related Articles | Metrics

First, the differential evolution algorithm with differential operator and selection operator was introduced. Then, a new and universal encode method, positionorder encoding method, was proposed. Based on the approach, a discrete differential evolution algorithm, position-order encoding based differential evolution algorithm (PODE), was proposed. The numerical results about two TSP instances in TSPLIB show that PODE has excellent convergence and stability.

Research on dynamic order insert crossover operator for travel salesman problem

2007, 27(3): 633-634.

Asbtract ( )

PDF (853KB) ( )

Related Articles | Metrics

According to the characteristics of travel salesman problem (TSP), a new crossover operator, dynamic order insert crossover (DOIC) operator, was designed, which combined order insert crossover and used the greedy selection strategy in the cross of the genetic algorithm. This operator can make use of the local information effectively and inherit excellent gene from the parents. It has been proved effective through the simulation on some examples.

Research on reliability prediction model based on orthogonal layer-clustering algorithm

2007, 27(3): 635-637.

Asbtract ( )

PDF (715KB) ( )

Related Articles | Metrics

Because of the disturbance of the test data on the prediction of software reliability growth model (SRGM), and the very demanding precondition that test work load must be distributed at regular intervals, orthogonal layer-clustering (OLC) algorithm was proposed. All the data that cannot meet the precondition got normalized. And J-M pediction model based on OLC algorithm was used to evaluate the reliability of the software. Experimental results prove that the J-M prediction model based on OLC algorithm is superior to the traditional J-M model.

Application of ant colony algorithm in parameters optimization of fuzzy Petri nets

2007, 27(3): 638-641.

Asbtract ( )

PDF (709KB) ( )

Related Articles | Metrics

How to determine the various parameters of fuzzy production rules is significant to the built up of fuzzy Petri net (FPN), which has not been solved yet. Maximumminimum ant system (MMAS) of ant colony algorithm (ACA) was originally introduced into the procedure of exploring parameters of FPN. An optimization algorithm based on the techniques of multithreading was proposed. Realization of the algorithm did not depend on experiential data and no strict requirements for primary input were needed. Simulation shows that the trained parameters gained from above MMAS are highly accurate and the resultant FPN model possesses strong generalizing and selfadjusting capabilities.

Negotiation strategy based on similarity and issue dynamic property

Li-Ming WANG Dong-Sheng NIU

2007, 27(3): 642-645.

Asbtract ( )

PDF (1037KB) ( )

Related Articles | Metrics

To solve the problems of the existing proposal generating strategy mechanism in dynamic issues area, a new strategy mechanism was presented. In the strategy mechanism, the design of the META strategy was simplified by a new integration of existing strategy mechanism; and the existing trade-off algorithm was improved to be adapted to the issue dynamic changes in negotiation effectively. The experiments show that, facing issue dynamics problems, the proposed strategy mechanism guarantees the negotiation time as well as the joint gain, so it shows an excellent adaptability to issue dynamics in negotiation.

Construction of a multilingual language Ontology framework

2007, 27(3): 646-649.

Asbtract ( )

PDF (1138KB) ( )

Related Articles | Metrics

To study the language Ontology, a language Ontology oriented to multilingual information processing was put forward. Not only its design philosophy, definition and reasoning mechanism, but also its structure and construction were described. The layered treestructure was introduced to construct the Ontology including noun, verb, adverb and adjective. In that Ontology, the semantics of phrases and the languages were designed as class nodes while the phrases were designed as leaf nodes. This Ontology can express the semantics of multilingual phrases, and it also can connect them with the semantics and transfer a phrase from one language to another language. Additionally, the reasoning mechanism and the formula to calculate the similarity degree between the phrases were provided to infer the semantics of a phrase.

Conflict detection algorithm in collaborative design

2007, 27(3): 650-652.

Asbtract ( )

PDF (595KB) ( )

Related Articles | Metrics

Conflict is the essence of collaborative design. The causes of conflicts include different thinking angles, judging criteria and specialty knowledge of different designers in collaborative design environment. A consistency model of operation sequence and conflict detection algorithm based on geometry level were proposed. Accordingly, a prototype conflict detection system in collaborative design was implemented.

An approach to extract answer in question answering based on FCA

2007, 27(3): 653-655.

Asbtract ( )

PDF (896KB) ( )

Related Articles | Metrics

As the number of available processing options continues to grow and our question answering (QA) systems attempt to answer more complicated questions, it is essential to have a principled method for dynamically generating different QA strategies. A question answering method was proposed using formal concept analysis (FCA) to extract answers. In the process of extraction, we firstly searched the frequently asked questions (FAQs). If the answers in the FAQs cannot satisfy the users requirement, we continued to retrieve the relevant documents from Web by search engine. Then, the concept lattice was built by making use of the first N documents. Thereby, the concept matching was utilized to extract answer from the lattices. Different strategies were used to match and select answers to different questions.

SLA-based implementation mechanism of pricing knowledge support system on next generation network

2007, 27(3): 656-658.

Asbtract ( )

PDF (1006KB) ( )

Related Articles | Metrics

A pricing knowledge support system (KSS) architecture on next generation network (NGN) was proposed, by collecting and analyzing the process data of service level agreement (SLA) decision, SLA detection and SLA evaluation. It obtained network key performance indexes (KPIs), market KPIs and policy KPIs of ISP. By analyzing these KPIs, each key quality index (KQI) was got, and by evaluating these KQIs, the ISP made profit margin decision and forecasted the price impact on operation environment. An XMLbased SOA implementing architecture of it was developed. Simulation result on the SLAbased NGOSS primitive platform verifies its merits in the comparison with ITDSS/TelPrice which won IST award.

Improved projection based M-estimators for fundamental matrix in stereovision

2007, 27(3): 659-662.

Asbtract ( )

PDF (1186KB) ( )

Related Articles | Metrics

In stereovision and image motion analysis, it is important to eliminate the mismatched feature points to get accurate fundamental matrix estimation. In order to overcome the limitations of kernel estimation and shortcomings of the method for determining inliers/outliers threshold in ErrorInVariables (EIV) model, the projection based Mestimators (pbM) was improved. First, a new mode search method named adaptive kernel density estimation was proposed. Then, a different strategy was given to determine the inliers. Finally, both simulation and experiments on real data validate the effectiveness of the improved pbM algorithm.

Realization of hidden line removal in 3D dynamic geometry

2007, 27(3): 663-665.

Asbtract ( )

PDF (770KB) ( )

Related Articles | Metrics

As OpenGL cannot realize dynamic hidden line removal in a dashed way, the realtime hidden line removal is a big problem while designing the three dimension dynamic geometric software. Based on the OpenGL system and the present hidden surface removal technology - depth buffer algorithm, the overlap of two lines and the rotation angle of the viewpoint about the local coordinate system were computed to achieve the dynamic hidden line removal in a dashed way. Furthermore, when a line is hidden by a surface, and a curve is hidden by plane or surface, this method can also be applied.

A real-time video denoising method based on lifting scheme

2007, 27(3): 666-668.

Asbtract ( )

PDF (935KB) ( )

Related Articles | Metrics

Video denoising algorithms based on discrete wavelet transform (DWT) do not allow real-time processing due to their huge computational complexities. To overcome the shortcoming, a new real-time algorithm for video denoising based on lifting scheme was proposed. First, each frame of the video was decomposed into multilevel wavelet to obtain the scaling coefficients and the wavelet coefficients by using lifting scheme. Then, each level of wavelet coefficients was filtered by softthreshold algorithm. After the inverse wavelet transform, temporal filter was applied on the filtered video to enhance the denoising effect further. Experiments show that this algorithm has better quality in terms of realtime processing and denoising effectiveness.

Improved inverse compositional algorithm and operators comparison

2007, 27(3): 669-672.

Asbtract ( )

PDF (174KB) ( )

Related Articles | Metrics

Traditional inverse compositional image alignment algorithm (ICIA) compares intensity values between a template image and an input image. However, it is easily affected by the variations of the lighting conditions, which leads to poor convergence or even divergence. Local orientation of an image tends to be less sensitive to lighting conditions than intensity values. Therefore an inverse compositional gradient (ICG) algorithm was proposed based on the local orientation. Experimental results demonstrate that this algorithm could overcome the influence of the lighting variations efficiently. In addition, the evaluation of the local orientation and image gradients is essentially equivalent. Hence several gradient operators were used to compute the local orientation, and the influences of the operators on the ICG algorithms under different lighting conditions were analyzed and compared.

Combined unsupervised image segmentation using watershed and hierarchical clustering with MRF

2007, 27(3): 673-676.

Asbtract ( )

PDF (1194KB) ( )

Related Articles | Metrics

A new combined multistage method for image segmentation was proposed. In the first stage, an oversegmented image was got using immersion watershed segmentation, and then primitive segmented results were provided for following merging. In the second stage, region based hierarchical clustering was used with a new heuristic computational model to merge adjacent primitive regions spatially. The model was derived from Bayesian method and Markov random field, and contained two interactive components. From experiments with three different kinds of images, the proposed method shows itself as a very effective way in unsupervised image segmentation. The hierarchy model merges regions in a way same as human perception and finishes within several seconds even for complex aerial image.

Image registration algorithm based on motion vectors estimation

2007, 27(3): 677-679.

Asbtract ( )

PDF (876KB) ( )

Related Articles | Metrics

The goal of image registration is to align or match different images of the same scene and remove the potential geometric distortion existed. An algorithm of image registration based on motion vectors estimation was proposed. The algorithm takes advantage of complex wavelet transform (CWT) over the conventional wavelet transform, such as shiftinvariance and better directional selectivity, to estimate motion vectors. The experimental results demonstrate that our proposed algorithm has very good performance in removing geometric distortion.

Noise image edge detection based on lifting scheme and floating threshold

2007, 27(3): 680-682.

Asbtract ( )

PDF (658KB) ( )

Related Articles | Metrics

Traditional algorithms of image edge detection were analyzed. In order to better detect the image edge information, firstly, nonlinear median filter technology of characteristic direction was applied to filter out the Gauss noise; secondly , (3,1) biorthogonal spline wavelet was used as the primitive wavelet filter, adopting lifting scheme to build the most superior biorthogonal wavelet filter, and applying the superior biorthogonal wavelet to the multiresolution wavelet edge detection; at the same time, floating threshold was defined and applied to the image edge detection, and the image edge position was located accurately. The experimental results show that the algorithm for the image edge detection has higher examination speed and better efficiency.

Morphable model with shapeconstrained 3D face components

2007, 27(3): 686-689.

Asbtract ( )

PDF (816KB) ( )

Related Articles | Metrics

Traditional 3D face morphable model constructs the space of the shape parameters and texture parameters by studying those from lots of 3D face samples. But the model can not describe the whole morphable space and can not reconstruct 3D shape well from 2D face image. To solve the problem, a face was constructed by some components, such as eyes, mouth, etc. By this method, the morphable model can construct the parameters space better. But, this kind of component model has more parameters and the overlaps between those components are difficult to deal with. In this paper, we used the shape model as a whole and the texture model as components. In this way, we can get a balance between performance and complication.

Online palmprint recognition using line edge map

2007, 27(3): 690-692.

Asbtract ( )

PDF (697KB) ( )

Related Articles | Metrics

How to make use of palmprint to do the personal identification is the study focus of this paper. Line feature, among all the features of palmprint, is a very important one. However, considering the palmprint lines are very irregular and hardly accurate depicted with mathematics, this paper proposed an approach that these palmprint lines were approximately represented by some line segments, and put forward employing the line segment Hausdorff distance method to match two sets of line segments, and meanwhile improved this method. The experiment shows our approach can achieve very effective personal identification.

Efficient local asymmetric region detection in medical images

2007, 27(3): 693-695.

Asbtract ( )

PDF (199KB) ( )

Related Articles | Metrics

A new method for efficient local asymmetric regions detection in medical images was proposed by removing symmetric regions. First, the 3D image was aligned to its reflex using symmetry based rigid image registration algorithm. Then based on the registered image, a modified adaptive erosion operator was used to remove the approximately symmetric region. These two steps were iteratively repeated using a multiresolution scheme to extract the final asymmetric region. The experiment to detect cholesteatoma in human head CT shows that the proposed algorithm detects asymmetric regions correctly in 80% cases.

Application of PCA in dimension reduction of image Zernike moments feature

2007, 27(3): 696-698.

Asbtract ( )

PDF (724KB) ( )

Related Articles | Metrics

Higher dimension of image feature is the critical problem and the dimension reduction is the most important phase in image processing. It was pointed out that the dimension of Zernike moments feature vector was generally high after briefly introducing the basic concept of the Zernike moments and the image Zernike moments shape feature vector. Based on the principal components analysis, it was shown that the principal components analysis (PCA) could be applied in dimension reduction of image Zernike moments feature. Meanwhile, the process of the dimension reduction based on PCA was put forward. The experimental results demonstrate the feasibility of the application.

Parallel processing algorithm for synthetic aperture rada based on pipeline and its implementation

2007, 27(3): 699-702.

Asbtract ( )

PDF (763KB) ( )

Related Articles | Metrics

According to the characteristics of RangeDoppler processing algorithm, a pipelinebased parallel processing algorithm for Synthetic Aperture Radar (SAR) data was proposed. The algorithm coded in C/MPI was successfully implemented in a 32-node IBM PC cluster. Results show that this pipelinebased algorithm is more suitable for SAR parallel data processing with higher parallel efficiency compared with other general parallel algorithms.

An improved realtime transmission method for remote screen images

2007, 27(3): 703-705.

Asbtract ( )

PDF (688KB) ( )

Related Articles | Metrics

An improved real-time transmission method for remote screen image was proposed. In this method, color filter，grid，xor and block compress between frames were combined, and screen capturing and compressing were improved. The experimental results show that this improved method decreases the data size, shortens the dealing time of screen, and increases the transmission speed.

Design of μC/OS-II multitask period based on rate monotonic scheduling algorithm

2007, 27(3): 706-708.

Asbtract ( )

Related Articles | Metrics

When multitask is scheduled, the embedded realtime operation system μC/OS-II always runs the task that has highest priority in the ready table. This strategy may lead to that the task cannot be scheduled if the period of task is designed improperly. Depending on the rate monotonic (RM) scheduling algorithm, each task could be scheduled by the analysis of RM algorithm in the design of tasks period.

Intellective PXI bus extended card design based on ARM microcontroller

2007, 27(3): 709-711.

Asbtract ( )

Related Articles | Metrics

In the development of measure & control system, the usage of intellective bus extended card is always necessary to satisfy the real-time requirement of sophisticated applications. The designing idea and development plan of bus extended card based on ARM microcontroller were introduced, with reference to the research of bus extending technology and embedded computing technology. The periphery circuit connecting method of onboard microcontroller and data access technology through DPRAM between microcontroller and mother board CPU were put forward.

Design of BSC function test system based on ACE reactor

2007, 27(3): 712-714.

Asbtract ( )

Related Articles | Metrics

The global system for mobile (GSM) communication projects require imminently to make function test on base station controller (BSC) system. But BSC platform encounters more difficulties due to the variety and heterogeneity of hardware and software. In order to reduce the cost of software and simplify the configurations of test system, software was used to simulate the logic of hardware by rich components of Adaptive Communication Environment (ACE) and Reactor design pattern. Two stubs were introduced to simulate the physical NetHawk Server cards. Based on IP communication, a new test environment was built under Linux operating system. The experiment proves the reliability and robustness of the new test system, and the costly hardware resource shortage is not a problem in the system as well.

Identification of tea taste signals based on fuzzy rule generation algorithm

2007, 27(3): 715-717.

Asbtract ( )

PDF (746KB) ( )

Related Articles | Metrics

After the deep study of weighted fuzzy neural networks and particle swarm optimization algorithm, a pruning algorithm was proposed to optimize the fuzzy rules automatically. The fuzzy neural network was constructed on the basis of the effective rules obtained through fuzzy rules searching. Numerical simulations for taste identification of tea show the effectiveness of the proposed algorithm.

Study on flood spreading model of Hongze Lake based on cellular automata

2007, 27(3): 718-720.

Asbtract ( )

PDF (689KB) ( )

Related Articles | Metrics

Basic principle and structure of Cellular Automata (CA) was analyzed, and flood spreading model of Hongze Lake was established by making use of CA rule. Based on grid spatial data structure, adopting the two-dimensional CA model, the cellular space and cellular states were established, and the rule for flood spreading modeling was made. The update of the cellular sate acted according to the rule of 9 neighbors. The computed data by choosing typical index was tested. The results show that two-dimensional CA model can model the flood spreading process more simply and faster.

An approach to optimize cancer chemotherapy using population based incremental learning algorithm

2007, 27(3): 721-723.

Asbtract ( )

PDF (573KB) ( )

Related Articles | Metrics

A methodology for using Estimation of Distribution Algorithm (EDA) to optimize cancer chemotherapy was proposed. The performance of population based incremental learning (PBIL) algorithm in optimizing cancer chemotherapy was studied. The experimental results show that PBIL algorithm outperforms genetic algorithms in both the speed of finding a feasible treatment schedule and the quality of the final solution.

Application of interactive evolutionary computation in virtual actor's expression modeling

2007, 27(3): 724-726.

Asbtract ( )

PDF (543KB) ( )

Related Articles | Metrics

An Interactive Evolutionary Computation (IEC) method was illustrated for training the facial expressions of the virtual actors in three dimensional computer graphical movies according to the effect that the human directors expect. Fuzzy neural network was used to accommodate directors expert knowledge so that appropriate virtual actors facial expression could be produced. In order to speed up the convergence of the learning algorithm, genetic algorithm was incorporated into the fuzzy neural network with IEC. Experimental results indicate the success of our approach here.

Particle swarm optimization based on uniform design and its application in flight control systems

2007, 27(3): 727-729.

Asbtract ( )

PDF (191KB) ( )

Related Articles | Metrics

Two problems need to be solved: how to select the object function and how to determine the initial swarm and parameters of the particles swarm optimization (PSO), when PSO is applied in the optimization of flight control systems. Concerning these problems, an optimizing strategy based on reference model for the flight control system (FCS), and a method based on the uniform design to select the initial swarm and parameters of PSO were given. The simulation shows that the optimizing strategy can solve the optimal design of the FCS effectively, the initial swarm got by the selecting method distributes uniformly, and the PSO with these initial swarms has more rapid convergence velocity.

Multi-velocity particle swarm optimization and its application in soft sensor

2007, 27(3): 730-732.

Asbtract ( )

PDF (695KB) ( )

Related Articles | Metrics

Multi-velocity particle swarm optimization algorithm (MVPSO) is an improved PSO algorithm, which can overcome the shortcomings of easily getting in the local extremum. Besides, this improved algorithm is characterized by clear definition, simple operation and easy realization. Both MVPSO and PSO were used to resolve several typical test function optimization problems. The results indicate that MVPSO has higher efficiency, better performance and more advantages in many aspects than PSO. MVPSO was applied to train artificial neural network to construct a practical soft sensor of product concentration of penicillin fermentation process. The experimental results indicate that this method is feasible and effective in soft sensor application.

Proxy cache replacement algorithms for streaming media based on smallest cache utility

2007, 27(3): 733-736.

Asbtract ( )

PDF (714KB) ( )

Related Articles | Metrics

The Smallest Cache UtilityK (SCU-K) algorithm was proposed. It comprehensively considered the latest K times of visits in order to make the cached size dynamically adapt to the change of the media popularity, bytes benefit and the size of cached parts. In such cases, the probability of the files prefix part being replaced was reduced and such problem of the LRU and LFU algorithm as the streaming media file being continuously replaced and finally being released was avoided. The experiment proves that the SCUK algorithm has better performance in increasing cache space utilization rate and byte hit ratio, and reducing startup delay than LRU, LFU and LRU-2.

Research and implementation of media player based on digital television middleware

2007, 27(3): 737-739.

Asbtract ( )

PDF (731KB) ( )

Related Articles | Metrics

Based on the hardware platform of digital television (DTV), the architecture and functions of the middleware model were analyzed with reference to the DVBMHP standard. With the study of JavaTV and JMF, the core part of the middleware system called media player was implemented. A media player performance test was launched on the DTV platform and good broadcasting effect was got.

Representative excerpts extraction of music based on similarity and novelty

2007, 27(3): 740-742.

Asbtract ( )

PDF (561KB) ( )

Related Articles | Metrics

The method for automatically producing representative excerpts of music based on similarity and novelty was proposed. The segment similarity was maximized to the entire work to find the most similar excerpt, and the novelty of the segment was maximized to find the most meaningful excerpt. The two measures were combined to find the most representative excerpt. At the same time, mathematical morphological filtering was used to preprocess audiosignal to eliminate incidental elements. Experimental results demonstrate that the method can find the most representative excerpts, just using very few assumptions about the source music.

A rapid prediction mode decision algorithm for AVS-M intra video coding

2007, 27(3): 743-745.

Asbtract ( )

PDF (452KB) ( )

Related Articles | Metrics

AVS-M is a mobile video standard owned by China. In luminance intra prediction of AVS-M, there are nine modes and two methods for the optimized mode selection-the SAD method and the RDO method, and the choice of optimized mode is very complex. A rapid prediction mode decision algorithm for intra encoding was proposed based on the sorted SAD result. Simulation results show that compared with RDO method, the rapid algorithm can save 30%~40% encoding time with a little performance degradation.

Rate control by fuzzy measurement for H.264/AVC

2007, 27(3): 746-748.

Asbtract ( )

PDF (510KB) ( )

Related Articles | Metrics

The usual rate control algorithms of H.264/AVC always adopt mean absolute differences (MAD), which is simple but does not comply with the human visual system, as the major measurement of the coding error. A class of image distortion metric based on fuzzy integral was referred to and applied to the rate controlling. The result shows that this algorithm achieves both higher PSNR and higher subjective visual quality in either the CBR case or the VBR case.

Improvement research based on H.264 TM5 rate control algorithm

2007, 27(3): 749-751.

Asbtract ( )

PDF (515KB) ( )

Related Articles | Metrics

Aiming at the shortcoming of H.264 TM5 rate control algorithm on sequence smoothness keeping and buffer controlling, the picture complexity and buffer situation were considered simultaneously, and an improved frame layer rate control algorithm based on H.264 TM5 was proposed. The frames of higher picture complexity were quantized with bigger parameter, while frames of lower picture complexity were quantized with smaller parameter. The buffer situation was also monitored to prevent the buffer from overflowing. The simulation shows that the improved algorithm can not only get smoother picture sequence, but also control the buffer better.

Research on product ontology construction in Ecommerce environment

2007, 27(3): 752-755.

Asbtract ( )

PDF (844KB) ( )

Related Articles | Metrics

Product ontologies in Ecommerce establish a shared and common understanding of product domain by building hierarchies of product classes enriched by attributes. In standard product ontologies, such as UNSPSC （Universal Standard Products and Services Classification Code）, the main drawback is the lack of attributes attached to the concepts. As table is an important means for organizing products information in Ecommerce Webstation, a semiautomatic approach was proposed to extract property conceptual ontology from tables. The approach used UNSPSC as embryonic kernel ontology. Table recognition, table normalization, table structure detection, unit product ontology and global product ontology building were researched. The semiautomatic approach can avoid the nonconsistence due to the absence of description standard in standard product ontology and realize the automatic extension and complement of the kernel ontology.

An automatic generation algorithm of MPI communication code

2007, 27(3): 759-761.

Asbtract ( )

PDF (561KB) ( )

Related Articles | Metrics

Automatic parallel code generation by information from computation and data decomposition is an important optimization for achieving high performance on largescale parallel machines. According to the theorem of Saman P.Amarasinghe and Lam, an algorithm was proposed that generated parallelized code automatically in the recognition tool MPT (Message Passing Interface). The performance of this algorithm was analyzed in the end.

A heuristic activity-on-arrow network autogeneration algorithm

2007, 27(3): 762-764.

Asbtract ( )

PDF (541KB) ( )

Related Articles | Metrics

A heuristic activity-on-arrow network autogeneration algorithm was proposed. It generated the optimized activity-on-arrow network for simplifying without a second scan. The algorithm can not only generate the correct and optimal activity-on-arrow network, but also has high efficiency.

Design and Implementation of a Case-based Reasoning Human-machine Conversation System

2007, 27(3): 765-768.

Asbtract ( )

PDF (834KB) ( )

Related Articles | Metrics

Adopting the Case-based Reasoning method and Combining Human-machine Conversation, Speech Recognition (SR) and Text to Speech (TTS) technologies, this paper researches the design and implementation of a Human-machine Conversation System which allow English learners to rehearse social conversations in an engaging environment without the need for a human partner to be present. This paper describes the system’s speech function, dialogue management and data access. Experiments prove that the system well meets the user’s demand of improving his or her hearing and speaking English.

Reuse technology and application of engineering account bill software

2007, 27(3): 769-771.

Asbtract ( )

PDF (619KB) ( )

Related Articles | Metrics

Take the engineering account bill system as an example, the application of software reuse technology in different development stages of software designing was studied. The problems of software field reuse and software layer reuse were discussed. Horizontal reuse of software product field and vertical layer reuse in the process of software development were achieved. Thereby, reusable feature, software quality and developing efficiency of software products were improved.

Table of Content