Started in January,1974(Monthly)
Supervised and Sponsored by Chongqing Southwest Information Co., Ltd.
ISSN 1002-137X
CN 50-1075/TP
CODEN JKIEBK
Editors
Current Issue
Volume 38 Issue 1, 16 November 2018
  
State-of-the-Art,Challenges and Perspectives of Agent-oriented Software Engineering
MAO Xin-jun
Computer Science. 2011, 38 (1): 1-7. 
Abstract PDF(672KB) ( 399 )   
RelatedCitation | Metrics
Agent oriented software engineering is a novel software paradigm that is considered as an important approach to supporting the development of complex software systems based on agent technology. Many attentions had been paid by researchers in academic literature and practitioners in industry literature. In the past ten years,great progresses of Agent oriented software engineering has been made. However, there are still a great number of problems that should be solved before it moves to industry application and its potentials should be exploited extensively. After investigating the background and characteristics of agent oriented software engineering, the paper overviewed the statcof-thcart of the researches on agent-oriented software engineering, identified and analyzed a number of key challenges of such technology, outlooked and discussed the future directions from technique, process and tool viewpoints.
Wireless Sensor Network Management Technology
ZHAO Zhong-hua,HUANUFU Wei,SUN Li-min,DU Teng-fei
Computer Science. 2011, 38 (1): 8-14. 
Abstract PDF(724KB) ( 502 )   
RelatedCitation | Metrics
Wireless sensor networks are resource-constrained and application-related. Wireless sensor networks are different from other traditional computer networks, so the traditional network management is no longer applied to wireless sensor networks and wireless sensor network management is faced with many challenges. This paper briefly described the technology background of the wireless sensor network management; gave the corresponding management characteristies in wireless sensor networks with the characteristics of wireless sensor network itself;then put forward a common framework of wireless sensor network management and discussed the contents of the various management and research progress in detail;finally,we discussed the public challenges facing the wireless sensor network management and pointed out the future research directions.
Research of Learning-based Planning Techniques
CHEN Ai-xiang,JIANG Yun-fei,HU Gui-wu,CHAI Xiao-long,BIAN Rui
Computer Science. 2011, 38 (1): 15-19. 
Abstract PDF(567KB) ( 340 )   
RelatedCitation | Metrics
After nearly 10 years of effort, the mordern smart planner, whether its efficiency or processing capacity, all have been greatly enhanced. Subject to the limitations of current planning theory, to further enhance the efficiency of existing planning techniques under current framework has become more difficult Existing planners have not to learn ability, most of them, can not learn from previous experience, useful knowledge to solve. In this paper, we first review the development of planning techniques to learn, and then focused on learning techniques used on the best learning-based planner among international planning competition, concluded the main problems and challenges in current learning technology research of planning.
System Calls Based Intrusion Detection:A Survey
WU Ying,JIANG Jian-hui,ZHANG Rui
Computer Science. 2011, 38 (1): 20-25. 
Abstract PDF(706KB) ( 761 )   
RelatedCitation | Metrics
System call based intrusion detection is currently a hot subject of research all over the world.The existing system call based intrusion detection techniques and theories with their respective challenges and research trends were discussed comprehensively,especially those(that are) newly developed. We hold that with the advent of the Tide-based commercial intrusion detection system(IDS) SanAPT, how to improve detection performance, to decrease error alarm rate and to solve issues on multiplatform,lightweight,and distribution related to practicality of the IDSs will be hot topics in this field.
Construction of Boolean Function with Maximum Algebraic Immunity
XIONG Xiao-wen,QU Liong-jiang,LI Chao
Computer Science. 2011, 38 (1): 26-30. 
Abstract PDF(415KB) ( 493 )   
RelatedCitation | Metrics
In this survey paper,we reviewed the recent constructions of Boolean functions with maximum algebraic immunity( MAI),and classifcid those into several different classes by the construction idea. Further,we also presented some results and developments of these methods, including some results of the authors.
Internet of Things: Semantics, Properties and Category
HUANG Ying-hui,LI Guan-yu
Computer Science. 2011, 38 (1): 31-33. 
Abstract PDF(375KB) ( 675 )   
RelatedCitation | Metrics
It is very important for Internet of Things as a new kind of global information engineering to understand the concept itself (including the semantics, the properties and the category) clearly. Analyzing the semantic meaning of the word Internet of things can help to understand the properties of Internet of Things. Comparing the similarities and differences of the entity Internet of Things and other Internet applications can help to classify the entity. The actual semantics of Internet of hhings is Internet of product information, a new type of Internet application. Chinese translation of "Internet of Things" is the failure to highlight the two key points "information" and "based on Internet". The properties of Internet of Things include the dominance and the attributes. The former means the interconnection of product information in different parts of the world based on Internet, while the latter further explains the three restrictive characters of the product information, namely the electronic tag form, the standardizing restraint, and the ontology supporting. Internet of Things can be deemed as a specific application form of Semantic Web.
Research Summarization of Humanoid Expression Robot with Emotion
WANG Zhi-liang,WANG Wei,GU Xue-jing,ZHENG Si-yi
Computer Science. 2011, 38 (1): 34-39. 
Abstract PDF(578KB) ( 967 )   
RelatedCitation | Metrics
Humanoid expression robot with emotion is as one of hot topics in robot rscarch. To this area, summarization was given. First, after pectination of the road of research on emotional humanoid expression robot, related theories and technologies of representative robot were summarized synoptically. In additon, their system construction, mechanical structure, sensors, controlling system and another features were compared and analyzed. Second, theories and technologies to this area, such as facial action coding system, artificial psychology and artificial emotion, sensoring technlogy, image processing, voice recognization and synthesis were expounded in detail. Based on these, some future research directions to several existing problems were proposed. Lastly, developing of humanoid expression robot with emotion was forecasted.
Algorithms for Feedback Set Problems:A Survey
WANG Jian-xin,Jiang Guo-hong,LI Wen-jun,CHEN Jian-er
Computer Science. 2011, 38 (1): 40-47. 
Abstract PDF(725KB) ( 636 )   
RelatedCitation | Metrics
Feedback Set problems arc classical NP problems, which include Feedback Vertex Set(FVS) and Feedback Vertex Set(FAS) problems. There are important applications of these problems in many fields, such as circuit testing,deadlock resolution, analyzing manufacturing processes and computational biology. People have designed many different approximate algorithms based on linear programming and local search approaches, and have found exact solutions by Branch-Prune and Measure-and-Conquer techniques. Recently, Parameterized Feedback Set problems have received con- siderable attention. The development of parameterized complexity motivates the studies on parameterized Feedback Set problems, especially on parameterized FVS problem. A chain of dramatic improvements on FVS problem in undirected graphs were obtained using different methods, such as tree decomposition, branch-and-search, iterative compression. In this paper, the approximation algorithms and parameter algorithms about FVS problem in general undirected graphs were introduced firstly. Then the research on the FVS problem in directed graphs and special graphs were presented.Moreover, the FAS problems were also discussed. Finally, some future researches with considerable attention on FVS problems were proposed by analyzing the researches on feedback set problems.
Performance Analysis Method for Access Authentication Mechanism
LIU Wei,YANG Lin,LI Quan-lin
Computer Science. 2011, 38 (1): 48-50. 
Abstract PDF(332KB) ( 361 )   
RelatedCitation | Metrics
Authentication mechanism has important meanings to network security defence. Most researchers mainly focus on putting forward new authentication methods and analyzing the security of authentication protocols, but ignoring the research of quantitative analysis on authentication method. By analyzing the system of authentication as a quasibirth-and-death process, establishing a two-dimensional queuing model, the steady probabilities distribution can be obtwined to compute some important indices. The experimental results prove that the general analytical method can effectively evaluate the performance of authentication method.
Application of PKI in Spatial Information Networks
REN Fang,MA Jian-feng,ZHONG Yan-tao
Computer Science. 2011, 38 (1): 51-53. 
Abstract PDF(281KB) ( 407 )   
RelatedCitation | Metrics
The structure of spatial information networks was introduced in this paper, the characteristics and security threats were analyzed and the security goals were proposed. We applied the technology of hierarchical PKI to spatial information networks and constructed a model of multi-level CA.In this model, the security goals of confidentiality, authentication, integrity and non-repudiation were reached.
Swarm-Evolution-based Reputation Model for Peer-to-Peer Networks
BAO Yi-ping,YAO Li,ZHANG Wei-ming,TANG Jiu-yang
Computer Science. 2011, 38 (1): 54-56. 
Abstract PDF(348KB) ( 394 )   
RelatedCitation | Metrics
Reputation system is the key approach to fight against the trust crisis of Peer-to-Peer networks. A novel reputation model based on swarm evolution was proposed in this article. In the model, the Peer-to-Peer network is viewed as social ecosystem, the process of reputation evaluation is modeled as an evolutionary process for peers. By using the strategy of crossover and sclection,the peers can rapidly and efficiently evaluate the trust value of potential partners,and finally improve the ability of their fitness. In order for verifying its performance, a simulation experiment for this model was also provided.
Localization Algorithm Based on Dynamic Hop Distances in Wireless Sensor Networks with Holes
ZHANG Song-tao,JIANG Hong-bo,LIU Wen-yu
Computer Science. 2011, 38 (1): 57-61. 
Abstract PDF(477KB) ( 311 )   
RelatedCitation | Metrics
Range-free positioning technology in wireless sensor networks is a challenge. It is difficult to estimate the dislances between sensor nodes which are more than one hop away in anisotropic sensor networks with holes. We introduced a range-free localization approach in anisotropic wireless sensor networks with holes,which can estimate the ranges accurately between unknown node and reference nodes. First, the algorithm estimates different average distances of one hop along the shortest path between different anchor pairs. Then some unknown node picks up the anchor pairs,whose average distance of one hop along the shortest path is larger than some threshold and the shortest path passes corresponding unknown node,as reference nodes to estimate its position. The new algorithm can filter out the anchor nodes with higher errors of estimation distances as reference nodes. Simulation results show that the localization accuracy of new algorithm is higher than previous algorithms.
Secure RFID Communication Protocol in Supply Chains
DENG Miao-lei,HOU Ying-chun,WEI Bao-jun
Computer Science. 2011, 38 (1): 62-65. 
Abstract PDF(363KB) ( 379 )   
RelatedCitation | Metrics
Radio frectuency identification(RFID) technology has greatly facilitated the supply chain management. The application of secure RFID protocols is one important approach for protecting RFID enabled supply chain system Sc curity requirements for RFID communication protocols in supply chain environments were described. A new secure RFID communication protocol in supply chains was proposed. The new protocol imposes lower computation load and storage cost on RFID tags,and has higher efficiency.
Reuleaux Triangle-based k-coverage Algorithm in Wireless Sensor Networks
XING Xiao-fei,SUN Ze-yu
Computer Science. 2011, 38 (1): 66-68. 
Abstract PDF(357KB) ( 390 )   
RelatedCitation | Metrics
Coverage is one essential task in sensor deployment for the design of wireless sensor networks. This paper mainly focused on solving k-coverage under the case that the sensor nodes are deployed in randomly. A Reuleaux triangle-based k-coverage algorithm(RTCA) was proposed to ensure that coverage can satisfy the requirement of network.By slicing the sensing range of sensor into six overlapping Reuleaux triangle with same length, the RTCA algorithm can judge whether the region is k-covered. The simulation results show that RTC;A algorithm can reduce the total number of active sensors and satisfy the requirement of coverage, which can improve the energy efficiency of sensor nodes. This algorithm can be applied in k-coverage of heterogeneous sensor networks in future.
Mathematic Analysis of Semantic Overlay Network's Optimal Scale
LIU Zhen,ZHOU Hao-hao,DENG Su
Computer Science. 2011, 38 (1): 69-72. 
Abstract PDF(281KB) ( 344 )   
RelatedCitation | Metrics
At present, a lot of researchers have focused on semantic overlay network's construction for improving query efficiency in P2P networks.But the research on optimal scale of semantic overlay network is lacked. This paper tried to use mathematic method to model the semantic overlay network. And then model solution method of properties which impacts the routing arithmetic was researched. At the same time, the relationship between semantic overlay network's scale and the capability properties of routing arithmetic was analyzed. In the base of this, the optimal scale of the community would appeare, which supports the construction and research on semantic overlay network effectively.
Data Randomization to Defend Buffer Overflow Attacks
YAN Fen,YUAN Fu-chao,SHEN Xiao-bing,YIN Xin-chun,MAO Bing
Computer Science. 2011, 38 (1): 73-77. 
Abstract PDF(408KB) ( 462 )   
RelatedCitation | Metrics
Code injection attack has become a typical representative of the attacks against memory,in which buffer over-flow attacks are the most commonly used. It relics on the change of control-flow, lets the program point to the malicious code in order to obtain the root rights. This paper presented a method using randomization based on data protection, wiich can defend buffer overflow attacks effectively, through the protection of pointers and arrays.
LEACH-CS:A Customizable Zone-spanned Multi-hop Routing Protocol in WSNs
GU Yue-yue,BAI Guang-wei,TAO Jin-jing
Computer Science. 2011, 38 (1): 78-82. 
Abstract PDF(445KB) ( 361 )   
RelatedCitation | Metrics
This paper, based on the existing LEACH routing algorithm, proposed a novel customizable zoncspanned multi-hop routing algorithm(LEACH-CS) for large-scale, energy limited wireless sensor networks. At first, we introduced the concept of "zone". On this basis,an additional cluster-head may be elected on demand during cluster-head selection process. The purpose is to optimize multi-hop routing with the constraint of multi-zone distance threshold. Our simulation experiment demonstrated that, compared with the existing LEACH protocol, the LEACH-CS could reduce the energy consumption of the cluster-heads significantly and extend the lifccycle of largcscale networks due to more uniformly distributed dead nodes.
Reliable Maintanace of Wireless Sensor Networks for Event-detection Applications
HU Si-quan,YANG Jin-yangWANG Jun-feng
Computer Science. 2011, 38 (1): 83-86. 
Abstract PDF(355KB) ( 356 )   
RelatedCitation | Metrics
The reliability maintannace of the wireless sensor network is a key point to keep the alarm messages delivered reliably to the monitor center on time in a event detection application. Based on the unreliable links in the wireless sensor network and the network charateristics of an event detection application, MPRRM, a multiple path redundant reliability maintanace algoritm was proposed in this paper. Both analytical and simulation results show that the MPRRM algorithm is superior to the previous published solutions in the metrics of reliability, false positive rate, latency and message overhead.
Real-time Detection of Malicious Web Pages Based on Statistical Learning
WANG Tao,YU Shun-zheng
Computer Science. 2011, 38 (1): 87-90. 
Abstract PDF(459KB) ( 448 )   
RelatedCitation | Metrics
Malicious Web pages impose increasing threats on Web security in recent years. Currently, there are mainly two client side protection approaches including anti-virus software packages and blacklists of malicious sites. Anti-virus techniques commonly use signaturcbased approaches which might not be able to efficiently identify malicious HTMI codes with encryption and obfuscation. Furthermore, blacklisting techniques are difficult to keep up-to-date. This paper presented a novel classification method for real-time detecting malicious Web pages which is independent with the contents of Web pages. Our approach characterizes malicious Web pages using HTTP session information. With representafive statistical features and decision tree algorithm in machine learning,we built an effective classification model for online real-time detecting malicious Web pages. Experiment results demonstrate that we arc able to successfully detect 89. 7% of the malicious Web pages with a low false positive rate of 0. 3%.
High Energy Node Driven On-demand Distance Vector Routing Protocol
JIAO Zhen-zhen,TIAN Ke,ZHANG Bao-xian,ZHONG Guo,LI Xiao,ZHANG Yan
Computer Science. 2011, 38 (1): 91-93. 
Abstract PDF(383KB) ( 390 )   
RelatedCitation | Metrics
We designed a High energy Node driven Ad hoc On-demand Distance Vector routing protocol(HN-AODV) for wireless ad hoc networks. The design objective is to prolong the network lifetime. To achieve this goal, in our protocol design, the path to reach a destination is biased towards those paths constituent of nodes with high energy for balancing energy consumption in the network. Our protocol embeds this high energy node driven feature into the on-demand path discovery process of an existing routing protocol AODV. Detailed protocol design descriptions were provided Simulation results show that our protocol can greatly prolong the network lifetime as compared with AODV. Moreover, our idea can also work well with other existing on-demand routing protocols for wireless ad hoc networks.
Analysis on Ad Hoc Routing and QoS in Heterogeneous Wireless Networks
PEI Xue-bing,WANG Qing-ping,ZHU Guang-xi
Computer Science. 2011, 38 (1): 94-99. 
Abstract PDF(566KB) ( 514 )   
RelatedCitation | Metrics
We firstly gave a detailed analysis on the QoS routing protocols of ad hoc network. Then, the architecture of heterogeneous wireless networks and ad hoc multi-hop routing were illustrated in the environment of heterogeneous networks, including the position-based routing, the hierarchy routing, the multi-hop relaying routing to raise the network capacity, the routing for load balancing, and the cross-layer routing protocols. Finally, we summarized our previous research works about our proposed load balancing strategics based on the ad hoc multi hop routing in heterogeneous wireless networks,and analyzed the simulation results.
Verifiable and Efficient Multi-secret Sharing Threshold Scheme
BU Shan-yue,WANG Ru-chuan
Computer Science. 2011, 38 (1): 100-103. 
Abstract PDF(339KB) ( 709 )   
RelatedCitation | Metrics
In most of the multi-secret sharing schemes already published, RSA, ECC or other public key cryptosystems are used to improve security. But such schemes would take up lots of resources and result in low speed. We proposed a new multi secret sharing(t, n) threshold scheme based on shamir secret sharing scheme, using the Lagrange interpolating polynomial to split and reconstruct the secrets and the NTRU and oncway hashing function to verify the validity of data. The scheme is simple in design and requires limited calculation and limited storage space. It can detect effectively a variety of cheating or forgery behaviors and guarantee that the reconstruction of the secret is the secure and trustworthy.
Secure Overlay Network for Peer-to-Peer Full Text Search
HUO Lin,HUANG Bao-hua,BA0 Yang,HU He-ping
Computer Science. 2011, 38 (1): 104-106. 
Abstract PDF(365KB) ( 312 )   
RelatedCitation | Metrics
P2P(Peer-to-Peer) full text search utilizes resources of many peers, it is important to control the scope of peers involved in a search request Considering the requirements of security in full text search, SON(Secure Overlay Network) was proposed for organizing peers into overlay network according the domination of their security level. In SON, a search request of a peer should be transferred down to peers it can dominate. So the peers involved in a search request are only a subset of peers of the overly network, and the search result should accord with the security police certainly. Definition of SON was given and its properties were analyzed. Principle of P2P full text search based on SON and its security problems were discussed. The experiments show that P2P full text search based on SON is efficient.
Multicast Algorithm Based on Priority Route Performance in IPTV Channel lapping
JIANG Zhan-hua,SUN Dan-dan,WANG Ai-min
Computer Science. 2011, 38 (1): 107-109. 
Abstract PDF(271KB) ( 394 )   
RelatedCitation | Metrics
In order to reduce the channel zapping time of the IPTV, this paper presented a new multicast algorithm based on the priority route performance evaluation which maintains the multicast tree. First of all, the priority route set was got according to the distance between the candidate parent node and the source node, then it was described that the candidate path performance was evaluated by using the peer performance assessment algorithm, ultimately, the optimal parent node was obtained based on the level of path performance, and then the multicast tree management was completed. Analysis and simulations show that this algorithm improves the efficiency of the multicast tree effectively, reduces the channel zapping time of the IPTV and enhances the QoE(Quahty of Experience).
Efficient Scheme of (t, n) Threshold Group Signature
LIU Dan-ni,WANG Xing-wei,GUO Lei,HUANG Min
Computer Science. 2011, 38 (1): 110-112. 
Abstract PDF(255KB) ( 407 )   
RelatedCitation | Metrics
An efficient threshold signature scheme based on the difficulty of discrete logarithm problem and secret sharing method was proposed. Distribution Center(ID) shares the signature secret key with group members by the method of optional quotient.ID generates threshold Signature according to the sharing signatures after the identity and correctness verification.The identity and time stamp of the sharing signature members are recorded in the database prepared for arbitrating the dispute in the future. The analysis from correctness, security and efficiency shows that the proposed scheme is effective and secure.
Study of a Grid User Whole Life Cycle Management Model Based on State Chart
LIU Li-min,SHANG Chao-xuan,XU Zhi-wei
Computer Science. 2011, 38 (1): 113-115. 
Abstract PDF(328KB) ( 309 )   
RelatedCitation | Metrics
This thesis took the user's state and activity of grid user management as the studying matter.First,this paper introduced the location and role of user's state and activity in the user management architecture,presented a kind of user state chart based on meta and sub state and gave the classification and de-finition of each metastate.The definition and elements of each substate attached to each type of metastate and their relations were studied.Based on the state definition,this paper studied and defined user's activity-transition action and gave the grid user's state transition chart.Finally,this paper introduced the application of this model in the portal of high performance computing environment.
Design of Dynamic Approach for Data Path ID Distribution in WiMax Network
LI Jun,TANG Ding,ZHAO Zhi-jun,WU Jin-zhao,TANG Hui
Computer Science. 2011, 38 (1): 116-118. 
Abstract PDF(329KB) ( 345 )   
RelatedCitation | Metrics
URE(Ueneric Route Encapsulation) encapsulation is used as the tunneling protocol for the data plane in WiMax network, while the key option is applied on R4 and R6 for providing the DPid(Data Path ID) of the tunnel. DPid assigned by the Data Path Function uniquely identifies a particular tunnel for user payload, and it's area of validity is the NAP. This paper proposed a concept of DPid Server for the effectiveness of the DPid, which is used to store and manage all of the DPids in the NAP, and then a dynamic approach of DPid distribution was presented to guarantee the DPid's uniqueness and safety-reliablely distribution in the NAP. This approach simplifies the processes of DPid distribution,reduces the management difficulty and maintenance costs. Furthermore, this paper also analyzed the design and performance of this approach.
Research and Implementation of Mobile SSL VPN System Based on Socks5 Proxy
YU Ding-guo,SHU Ming-lei,TAN Cheng-xiang
Computer Science. 2011, 38 (1): 119-121. 
Abstract PDF(379KB) ( 1721 )   
RelatedCitation | Metrics
We studied the current weaknesses of the mobile VPN based on traditional IPSec and SSL technology to present a solution of mobile SSL VPN based on socks5 proxy. This paper analysed the principle and the implementation process of the mobile SSL VPN, and introduced the framework and workflow of this system. Finally, we implemented and tested this system, and analyzed its security performance, access and data transmission efficiency.
Real-time Scheduling Algorithm TC-LSF Used for the Switch Network
LIU Jun-rui,CHEN Ying-tu,FAN Xiao-ya
Computer Science. 2011, 38 (1): 122-124. 
Abstract PDF(257KB) ( 302 )   
RelatedCitation | Metrics
By analyzing the existing scheduling algorithm in the network and the real-time systems, the author put up a real-time scheduling algorithm TC-LSF(Tasks Combining-Least Start First) used for the switch network based on the LSF algorithm, to ensure the real-time restrictions of those tasks in the real-time network. The algorithm used the tasks combining strategy to combine multiple communications tasks, omitted the same routing and reduced the routing overhead. So,the network communication efficiency was greatly improved. This paper gave the implementation details and C fragments of the algorithm, and the performance of the algorithm was also analyzed.
Research on Workflow Access Control Temporal Policy Combine in Multi-domains
TANG Zhuo,LIU Guo-hua,LI Ken-li
Computer Science. 2011, 38 (1): 125-129. 
Abstract PDF(399KB) ( 552 )   
RelatedCitation | Metrics
In multi-domains environment, workflow access control policy is consisted of heterogeneous temporal policies in difference autonomic domains,and its requirement is special subjects can access special objects in perodic time or duration time. While XACML specifics policy combine, it does not contains temporal constraint. Based on the kinds of temporal constraint proposed by UTRBAC, this paper defined temporal constraints policies and illustrated by figures. It is extended from XACML by introducing correspond temporal constriant elements. Finally, this paper illustrated that the extended XACML can describe heterogeneous temporal policy combine conveniently through a example.
Analysis for Loop-freeness of Multipath Inter-domain Routing Based on Locator/ID Decoupling Architecture
LI Dan,WANG Bin-qiang,LIU Qiang,MA Hai-long
Computer Science. 2011, 38 (1): 130-135. 
Abstract PDF(586KB) ( 380 )   
RelatedCitation | Metrics
Multipath inter-domain routing has better effect in routing reliability and link utilization than traditional single-path routing policy, but how to guarantee the loop-freeness is still a new question. Through analyzing the incompleteness of AS_PATH detection mechanism,this paper proved the loop-freeness condition and offerd a method based on the collaborative operation of control and data plane(LFPM).By analyzing the theoretical conclusion and simulation experiments, LFPM could guarantee the loop-freeness demand in multipath routing. Moreover, the number of available routes could be increased greatly than equal cost multipath routing method and best routing method which could provide more probability for better routing selection results.
Research on Hybrid Simulation of OMNET++and H. 264 Codec
CAI Min-jie,LI Bo
Computer Science. 2011, 38 (1): 136-139. 
Abstract PDF(398KB) ( 410 )   
RelatedCitation | Metrics
As a good network simulation platform coming into vogue, OMNET++ lacks the consideration for video characteristics at application layer. Therefore,it can not evaluate the performance of protocols of wireless Ad hoc networks well under multimedia application requirements. In order to overcome this disadvantage, a new hybrid simulation method based on the combination of OMNET++and H. 264 codec was proposed in the paper. In the proposed simulation method, H. 264 encoder is used as the data source for OMNET++,and H. 264 decoder serves as the evaluator for video transmission quality. Based on the proposed method, we simulated the video transmission over wireless Ad hoc networks, and evaluated the performance of IEEE802. 11 MAC protocol. Simulations show that the proposed hybrid simulation method can provide good evaluation for the design and improvement of protocols and algorithms in wireless Ad hoc networks.
New Binary System for Detecting and Locating Integer-based Vulnerability on Run-time Type Analysis
XIAO Hai,CHEN Ping,MAO Bing,XIE Li
Computer Science. 2011, 38 (1): 140-144. 
Abstract PDF(486KB) ( 355 )   
RelatedCitation | Metrics
Integer-based vulnerability is an extremely serious bug for programs written in languages such as C/C++.Common Vulnerability and Exploit(CVE) shows that as the percentage of buffer overflow has declined,there has been an increase in related vulnerability types,including integer overflows and signedness errors. Here we presented the design, implementation, and evaluation of a tool for run-time detecting and locating integer-based vulnerability. We first translated the binary code into intermediate language VEX on Valgrind, then intercepted integer related statements at run-time, recorded the necessary information, and finally detected and located vulnerability based on the checking scheme. We chose several utility applications, which contain real integer-based vulnerability, to evaluate the effectiveness and run-time performance of our system. Preliminary experimental results are quit promising, it can detect and locate most of integer-based vulnerability in real software, and has very low false positives and negatives.
Software Reliability Metrics Selecting Method Considering Software Integrity Level
LI Hai-feng,MENG Ling-zhong,LU Min-yan,YIN Yong-feng
Computer Science. 2011, 38 (1): 145-149. 
Abstract PDF(470KB) ( 615 )   
RelatedCitation | Metrics
It is very important to select and use appropriate software reliability metrics(SRMs) for software intensive systems. Therefore there are some approaches presented for effectively selecting SRMs. However, almost all of these approaches didn't consider the software integrity which is a very important design attribute with the unique ability to conlain risk which combines the frequency and the consequence of software failure. Thus we suggested that the integrity is a significant factor which affects the final software reliability or safety and should be considered for selecting the suitable SRMs. This paper proposed an integrity-level-based framework with which software organizations can systemically and effectively identify and select the suitable SRMs respectively to satisfy the integrity requirement of the safety-critical software. Finally a preliminary case study was given by using this framework to select the basic, conditional and optional SRMs from ISO/IEC 9126 for various software integrity levels.
Program Chopping Approach for Java Program
JIANG Cao-qing,YING Shi,NI You-cong,GUAN Hua
Computer Science. 2011, 38 (1): 150-155. 
Abstract PDF(633KB) ( 458 )   
RelatedCitation | Metrics
Compared with program slicing,program chopping is a more focused approach of program analysis. And it is of great significance for the program understanding, analysis, debugging, testing. The existing chopping algorithms mostly are based on connected System Dependence Graph(SDG). However, the representation of SDG is very complicated, especially for larger program, which causes fault result easily. So this paper proposed a program chopping approach for Java program based on a representation for Java program with tags. In this approach, not only the accessory information on program dependency relation is utilized to analyze parameter dependency relation, and based on which this paper presented program chopping algorithms,but also inter-method program chopping can be converted to infra-method analyzing. Moreover the approach has much strongpoint such as fewer nodes of SDG and reusability of Program Dependence Graph(PDG),etc. Finally, this paper illustrated implementation process of this approach combined with an example and experiment.
Metric-based Research of Software Product Line Management
DING Jian-jie,HAO Ke-gang,HOU Hong,GUO Xiao-qun
Computer Science. 2011, 38 (1): 156-157. 
Abstract PDF(284KB) ( 364 )   
RelatedCitation | Metrics
Software product line is a new software development methods, production line management as one of the three basic activities directly affects the practice of software product line. In this paper,from the software metrics view,according to GQ(I)M model,a metri}based framework of software product line management was proposed and metrics which must be tracked in the management process was defined. This work has significance lead meaning for the effective implementation of product line management activities.
Autonomic Dependability Evaluation of Autonomic System for Service-oriented
ZHANG Hai-tao,WANG Hui-qiang
Computer Science. 2011, 38 (1): 158-161. 
Abstract PDF(358KB) ( 293 )   
RelatedCitation | Metrics
Based on scrviccoricnted aspect, a hierarchy evaluation system of service request-service process-service responsibility was proposed, which analyses autonomic dependability of autonomic system. This index system has extensibility and standardization. According to the result of test,the index system is valid and possible.
StarOSGi:A Distributed Extension Middleware for OSGi
SHI Dian-xi,WU Yuan-li,DING Bo,YIN Gang,WANG Huai-min
Computer Science. 2011, 38 (1): 162-165. 
Abstract PDF(467KB) ( 305 )   
RelatedCitation | Metrics
As the application domain has been expanded from a single node to distributed environments such as pervasive computing and enterprise computing, the OSGi technology should provide support for remote service access. On the premise of keeping the existing advantages such as service-oriented, dynamics and light weight, it is a great challenge for the researchers to provide effective distributed extension for the OSGi technology. Existing works in this domain has a set of deficiencies, such as the invasiveness to original OSGi programming model and lacking the support of interoperating with non-OSGi systems. This paper proposed a CORBA-based Distributed OSGi Model, which chooses CORBA middleware technology as a foundation and aims at non-invasiveness, generality and interoperation. Based on this model,this paper designed and implemented an OSGi distributed extension middleware-StarOSGi, which implements the remote service invocation capability with CORBA dynamic invocation and Java reflection technology. It can turn the central OSGi applications into distributed environment transparently while keeping the service-oriented programming model and lightweight feature of OSGi. Furthermore, StarOSGi supports the interoperation between OSGi and CORBA applications and also has performance advantages over existing works.
Modeling and Analysis of Cooperative Services Based on Price Stochastic Petri Net
WANG Hong-xia
Computer Science. 2011, 38 (1): 166-169. 
Abstract PDF(315KB) ( 352 )   
RelatedCitation | Metrics
Now an enterprise attaches great importance not only to system performance but also to system cost. In view of this situation the price stochastic Petri net was put forward in order to analyse the system price and performance. On one hand, price stochastic Petri net possesses the advantage of perfommnce analysis like stochastic Petri net. On the other hand, it can calculate system cost and guide the establishment, improvement of system macroly. An equivalent simplified method of collaborative services typical model based on price Stochastic Petri net was put forward. And corresponding certification was done. This method can improve the system analysis efficiency to a certain extent.
Model Checking of Timed Colored Petri Nets with Inhibitor Arcs
YANG Nian-hua,YU Hui-qun,SUN Hua
Computer Science. 2011, 38 (1): 170-176. 
Abstract PDF(703KB) ( 328 )   
RelatedCitation | Metrics
TCPNIA(Timed Colored Petri Nets with Inhibitor Arcs,TCPNIA) is a model for specifying embedded systerns. This paper proposed a structural transformation method from TCPNIA to TA(Timed Automata,TA). A collision mediation mechanism was introduced to ensure the semantics equivalence between TCPNIA and the transferred counter-part. I}he semantics equivalence was proved. The complexity of the transformation algorithm was analyzed. Hierarchical method was utilized to improve time and space efficiency in model checking. A case study shows the applicability and feasibility of the technique.
Trust Pattern Verification Based Trustworthy Judgment Method of Explanatory Web Text
HUANG Shuai-biao,ZENG Guo-sun,WANG Wei
Computer Science. 2011, 38 (1): 177-180. 
Abstract PDF(425KB) ( 311 )   
RelatedCitation | Metrics
There are much open Web information, which are composed of true and false information and useful and harmful information, and much information are untrustworthy. hherefore, it is more and more important how to judge the text credibility to choose the trustworthy and useful information from much Web information. This article introduced a kind of j udgment method of Web text content trust based on trust pattern verification. First, the author defined the trust pattern of Web text and described trust pattern by the formal method, then the author modeled Web text, and finally checked the Web text model to judge whether Web verification model satisfies the trust pattern, and designed the algorithm of calculating Web text trustworthy degree according to the result of model checking. Experiment proves this method is effective.
Intelligent Question-answer System Uniponse Based on Requirement
ZENG Cheng,CHEN Gui-sheng,DU Gang,LIN Chen
Computer Science. 2011, 38 (1): 181-184. 
Abstract PDF(515KB) ( 575 )   
RelatedCitation | Metrics
The current intelligent question-answer systems(IQAS) are restricted to find the best answer for a simple question. They are unable to meet increasingly complex personalized ctuestion-answer demands of users'. This papertransformed the user question to question-answer service requirement based on the requirement metes description framework RGPS, which passes through rectuirement recognition, rectuirement analysis, requirement verification and so on.Then a local services database constructed by dynamically mining software services in Internet was used to generate the intelligent question-answer service, which could answer some similar questions, with process services composition technology based on services relationship network. We developed an IQAS Uniponse in the traffic domain which verifies the effectiveness and feasibility of our method.
Model and Algorithms for Dynamic Workflow Mining
LIANG Jun-feng,ZHANG Zu-ping,LONG Jun
Computer Science. 2011, 38 (1): 185-189. 
Abstract PDF(423KB) ( 495 )   
RelatedCitation | Metrics
Workflow mining is a key technology in workflow re-design and analysis field. It is generated to improve the objectivity of workflow modeling by extracting the actual business related process from records of the running informalion systems in enterprises. However,most of Workflow Mining methods are based on records of information systems running in companies. On the other hand, they have bad performance in mining complex tasks and flexible workflow. To solve those problems, we proposed a Dynamic Workflow Mining model. In this model, there is no record from the beginning,workflow nets and records were dynamically generated. The given example was used to demonstrate improvements on model accuracy and description ability of the proposed algorithms.
Research and Implementation of Domain-specific Virtual Data Center Based on Semantics
LI Hua-yu,HU Chang-jun,OUYANG Chun-ping,YE Yin-zhu
Computer Science. 2011, 38 (1): 190-194. 
Abstract PDF(470KB) ( 313 )   
RelatedCitation | Metrics
The data involved in the field of oil-well-engineering are distributed, heterogeneous and autonomous, and there is also a complex semantic association among them Under this circumstance, effective global support for decisionmaking is difficult to ensure. Using integration technology of ontology and virtual view, this paper presented a virtual data center solution for oil-well data integration. Through global-ontology construction, ontology extraction, ontology mapping and query conversion, the data in oil-well engineering field can be semantically integrated. Consequently, a unified and semanti}based data query and shared services were achieved. By practical application, this virtual data center is able to supply production decision with comprehensive and real-time data services.
Clustering-based Improved K-means Text Feature Selection
LIU Hai-feng,LIU Shou-sheng,ZHANG Xue-ren
Computer Science. 2011, 38 (1): 195-197. 
Abstract PDF(280KB) ( 359 )   
RelatedCitation | Metrics
Text feature reduction is the key technology in text categorization. In addition, K-means is an partitioning method which usually be used. With regards to this arithmetic excessively incentive to the initial centers and the isolated points, the improved K-means arithmetic was put forward which is used in text feature selection. Text feature clustering was improved by optimizing primitive class center's options and the elimination of isolated point Following text classification test shows that the K-means arithmetic put forward in this paper has a good feature selection ability and high efficiency in text categorization.
General Extraction Engine Framework:Research of a New Approach for Web Information Extraction
GONG Ji-bing,TANG Jie,YANG Wen-jun
Computer Science. 2011, 38 (1): 198-202. 
Abstract PDF(482KB) ( 447 )   
RelatedCitation | Metrics
The large size of video collection not only provides an easy way for users to share information, but also brings a big challenge for managing them, in particular online monitoring. A critical rectuirement to monitor the video information is to accurately and adaptively identify the key information describing the video, which is also the first step for video analysis and video search. In this paper, we focused on the extraction problem of the video information from different websites. Specifically,we proposed an engine framework for information extraction. We formally defined the description model in the framework and implemented a customizable engine for information. The proposed framework has been applied to a real-world application of a national department and obtains promising results. Experimental results show that the proposed approach can effectively extract the video information and it significantly outperforms the baseline methods.
Maximal Clique Percolation Algorithm Based on Neighboring Information
CHEN Duan-bing,ZHOU Yu-lin,FU Yan
Computer Science. 2011, 38 (1): 203-206. 
Abstract PDF(344KB) ( 380 )   
RelatedCitation | Metrics
Maximal clique problem(MCP) is a classical and important combinational optimization problem with many prominent applications, for example, information retrieval, signal transmission, computer vision, social network and bioinformatics, etc. Researchers presented many algorithms to solve it by using various strategics, such as branch-and-bound, genetic algorithm, simulation annealing, cross entropy and DNA method. In this paper, a new clique percolation algorithm was presented based on neighboring vertices and edges of clique. From a given clique( it's a vertex at initial) at each step, investigated its all neighboring vertices and expanded it to a larger clique through a neighboring edge of clique. Two large scale author collaboration networks were used to test the performance of proposed algorithm and the clique distribution in large scale social network was also discussed. Experimental results demonstrate that the presented algorithm is efficient to percolation the maximal clique in network.
Probabilistic Earley Algorithm Based on Extended Viterbi Path
HAN Xi-wu,Roland Hausser
Computer Science. 2011, 38 (1): 207-209. 
Abstract PDF(268KB) ( 360 )   
RelatedCitation | Metrics
Probabilistic Earlcy algorithm applies Viterbi path to construct parse trees for input sequences, but the grammar limits often result in many empty trees.This paper exploited optional start states, new sequential beginnings and more subtrees to extend the traditional Viterbi path, removed almost all empty trees, and improved the general parsing performance of Earley algorithm.
Research on a Bilevel R-median Facility Interdiction Model for Air Attack Formation
ZHU Yue-ni,ZHENG Zhen,LIU Wei
Computer Science. 2011, 38 (1): 210-213. 
Abstract PDF(498KB) ( 338 )   
RelatedCitation | Metrics
Target selection and optimal resource allocation have become key factors of the effect of air-raid in modern warfare. We formulated a bilevel r-median facility interdiction model for air attack formation, based on an analysis of the main characteristics of attackers and defenders. In the bilevel formulation, the top level decision maker intends to make the most disruptive attack by identifying and interdicting the critical infrastructure in the target system; the lower level decision makers concerns the improvement of air attack formation safety and the accuracy of hit rate; the model is aimed to weigh the benefits of both top and low level decision makers and reach the optimal strategy of air attack formation allocation. Computational results on a simulation experiment show the effectiveness of the proposed methodology for providing optimal strategies of this problem.
Tree Kernel-based zero Pronouns Recognition
HUANG Li-wei,KONG Fang,ZHU Qiao-ming,ZHOU Guo-dong
Computer Science. 2011, 38 (1): 214-216. 
Abstract PDF(372KB) ( 308 )   
RelatedCitation | Metrics
This paper introduced a learning approach of Tree Kernel-based Zero Pronouns Recognition, and presented a system based-upon this approach. In order to get the syntax structure tree, which includes the zero pronoun, it marks the sentences of the CTB manually, and constructs a basccorpus.Then, it proposed three different policies to get syntax structure tree, which can be trained by SVM as a feature. At last, it also presented a baseline system based on rules, with which we can compares. Experimental results show that the recognition rate of the approach based on tree kernel is better than the one of the System based on rules.
Steganalysis of Compressed Speech Based on Statistics of Pulse Position Parameters
DING Qi,PING Xi-jian
Computer Science. 2011, 38 (1): 217-220. 
Abstract PDF(454KB) ( 381 )   
RelatedCitation | Metrics
For analysis-by-synthesis(AbS) compressed speech a steganalysis method based on statistics of the pulse position parameters was proposed. Speech compressed by G.729 coding algorithm was taken as an example to analyse the characteristic differences of the pulse position parameters of cover and stego speech, and a new feature, the histogram flatness was defined. In the steganalysis method,four statistics including histogram flatness,the center of mass of histogram characteristic function, histogram variance, and the probability difference of 0 and 1,were used as distinguishing features, and SVM classifiers were employed. Experiment results show the proposed algorithm has good performance for steganalysis of speech compressed by G.729 algorithm. As to compressed speech coded by other AbS-based coders, the proposed method is also applicable.
Novel Autonomous Clustering Method Based on Decision-theoretic Rough Set
YU Hong,CHU Shuang-shuang
Computer Science. 2011, 38 (1): 221-224. 
Abstract PDF(341KB) ( 398 )   
RelatedCitation | Metrics
This paper proposed an autonomous knowledge-oriented clustering method based on decision-theoretic rough set model. In order to obtain the initial clustering, the initial threshold values need to set in the knowledgcoricnted clustering framework. Thus, a novel method, sort difference, was proposed to produce the initial threshold values autonomously in view of physics theory. Then, a cluster validity index based on the decision-theoretic rough set model was developed by considering various loss functions, which can estimate the quality of clustering.The results of experiments show that the new approach is valuable.
Cartesian Product Rough Model of Approximation Spaces and Decomposability
WU Ming-fen,CAO Cun-gen
Computer Science. 2011, 38 (1): 225-228. 
Abstract PDF(337KB) ( 389 )   
RelatedCitation | Metrics
Pawlak proposed the rough set model, in order to processing data and knowledge which are imprecise or uncertainty in artificial intelligence. Then, the rough set model has been extended and many new rough set models have been put forward. Inhere are two main methods of extension, one method is to weaken the dependence of equivalence relation, the other is to expand the domain from one to two, and Y. Y. Yao ever proposed a rough set model of two-do-main. In this paper, we made some research for Cartesian product rough models based on two(finite) approximation spaces, and gave the concept of product approximation space. Afterwards, we described the upper(lower) approximation of decomposable subsets of a Cartesian product, and the approximate precision and roughness of decomposable subsets.Finally, we studied the decomposable problem of Cartesian product rough models, and obtained the sufficient and necessary conditions of decomposition of a product approximation space.
Incremental Algorithm for Attribute Reduction Based on Conditional Entropy
LIU Wei,LIANG Ji-ye,WEI Wei,QIAN Yu-hua
Computer Science. 2011, 38 (1): 229-231. 
Abstract PDF(325KB) ( 416 )   
RelatedCitation | Metrics
Rough set theory is a mathematic tool to deal with incomplete and uncertain information, in which attribute reduction is one of important issues. The changing mechanism of condition entropy was analyzed when a new object was added to the original decision table. Based on this mechanism, a new incremental algorithm for attribute reduction was proposed. In this algorithm we divided the added objects into three cases. Furthermore, by these different cases incremental attribute reduces could be calculated quickly. At last, the validity of the proposed algorithm was depicted by an experiment.
Method of Rule Extraction Based on Rough Set Theory
E Xu,SHAO Liang-shan,ZHANG Yi-zhi,YANG Fang,LI Han,YANG Jia-xin
Computer Science. 2011, 38 (1): 232-235. 
Abstract PDF(338KB) ( 436 )   
RelatedCitation | Metrics
Rule extraction is an very important and difficult process for an intelligent information system. To deal with the problem, the paper proposed a method based on rough set theory, researched attribute reduction, attribute values reduction and so on. According to the indiscernible relation in rough set, discernible vector and its addition rule were defined to calculate the core attributes and all attributes' importance. The core attributes set was taken as the start point to obtain an attributes reduction set by using the attributes' importance as the heuristic information. Based on the attributes reduction set, attribute value reduction was realized through gradually deleting the redundant attribute values in every rule of the information table depending on the correlation of condition attributes and decision attributes. Finally,a concise rule set was obtained. The illustration and experiment results indicate that the method is effective and efficient for rule extraction.
P-sets and Data Internal Search-application
XIE Wei-qi,YU Xiu-qing
Computer Science. 2011, 38 (1): 236-239. 
Abstract PDF(288KB) ( 338 )   
RelatedCitation | Metrics
Introduced by dynamic characteristics, P-sets can be obtained from the improved finite general sets. P-sets are a set pair which are composed of internal P-sets XF and outer P-sets XF,or (XF,XF) is P-sets. By using P-sets, the concept of data internal search, the measure and dependence relation about F-data were given, search iterative algorithm and criterion of F-data were proposed,finally the applications of data internal search were given. P-sets is a new theory and method to study dynamic information system.
Approach of Internet New Word Identification Based on Immune Genetic Algorithm
DING Jian-li,CI Xiang,HUANG Jian-xiong
Computer Science. 2011, 38 (1): 240-245. 
Abstract PDF(534KB) ( 444 )   
RelatedCitation | Metrics
The development of Internet leads the Internet new word coming into being. These unknown words are difficult to identify timely and accurately by the current Word Segmentation Method, therefore Internet new word identification method using Immune genetic algorithm was brought forward. This method is based on the analysis of characteristics of Internet new word, using the phenomenon of Chinese words and word groups to extract exemplary antibody, and injecting the antibody targeted during the process of genetic algorithm.The experiment results show that the method not only has a higher recognition rates of the new words consistent with the phenomenon of word groups in word fragments but the result of identifying ordinary Internet new word is adequate.
Three-way Decision-theoretic Rough Sets
LIU Dun,YAO Yi-yu,LI Tian-rui
Computer Science. 2011, 38 (1): 246-250. 
Abstract PDF(457KB) ( 379 )   
RelatedCitation | Metrics
A model of threcway decision-theoretic rough sets (DTRS) was presented based on the Bayesian decision theory. Based on the minimum expected risk, a detailed formulation of DTRS was given. Different types of errors in several probability intervals were examined. The conditions under which DTRS thruway method is superior to the Pawlak three-way method and two-way method were identified. DTRS threcway model was discussed for solving the practical decision problems.
Improvement of Dynamic Hand Gesture Recognition Technology Based on HMM Method
YU Mei-juan,MA Xi-rong
Computer Science. 2011, 38 (1): 251-252. 
Abstract PDF(195KB) ( 825 )   
RelatedCitation | Metrics
In order to enhance the efficiency and the accuracy of dynamic hand gesture recognition based on HMM method, for the computation's high complexity of HMM method in the training stage, a new HMM algorithm based on the dynamic programming was presented. It makes improvement to the HMM algorithm's training stage, and will enhance the accuracy and timeliness in human-robot interaction.
Generation and Identification of External-distortion Data
LI Yu-ying,LIN Hong-kang
Computer Science. 2011, 38 (1): 253-255. 
Abstract PDF(303KB) ( 306 )   
RelatedCitation | Metrics
P-sets have dynamic characteristics. P-sets are a new mathematic tool in dynamic information system. The concepts of external-distortion data and external-distortion data circle were proposed by using the outer P-sets and F-supplemented sets in P-sets structures. External-distortion data is a data pair which is composed of (x)F and(x)+,or ((x)F,(x)+).External-distortion data circle is a data circle pair which is composed of OF and O+, or (OF,O+).Using the concepts, this paper discussed the generation and identification about external-distortion. The theorems of generation and identification on external-distortion data were got The criterion of identification for external-distortion was given.The application on the identification of external-distortion data was provided.
Multi-structure, Multi-scale Morphology De-noising Combination of Genetic Optimization
WANG Yuan-ni,GE Fei
Computer Science. 2011, 38 (1): 256-258. 
Abstract PDF(366KB) ( 339 )   
RelatedCitation | Metrics
The traditional morphological filter, generalized morphological filter, adaptive weighted generalized filtering,generalized morphological filter of the multi structure elements or multi-directional or multi scale, all basically considering only one aspect, or improving the lack of only one aspect, regardless of what kind of filtering method that does not completely eliminate the noise. This paper presented a genetic algorithm based on adaptive multi scale multi structural morphological filtering method,the main consideration of the filtering window size,type and orientation of structural elements, as well as structural elements of the optimization of selection, using genetic algorithms to optimize the structural elements, and taking into account the convergence of genetic algorithm itself, using a strategy to retain the elite, while considering the choice of genetic algorithm parameters using adaptive strategies. At the same time, combined with the ideas of adaptive weighted generalized morphological filter to build the structure based on genetic optimization of multistructure,multi-scale adaptive weighted morphological filter,filter effects are better than the traditional morphological filter, the generalized morphological filter and others filters improved based on it.
Comparison with Attribute Reduction Algorithms in Information View
TENG Shu-hua,SUN Ji-xiang,ZHOU Shi-lin,LI Zhi-yong
Computer Science. 2011, 38 (1): 259-263. 
Abstract PDF(400KB) ( 430 )   
RelatedCitation | Metrics
Many types of conditional entropy reduction have been proposed and applied in information systems. It is thus important to clarify the relationships among the existing types of conditional entropy reduction. In this paper, the properties of three different conditional entropies were compared and analyzed, then we investigated the relations among three conditional entropy reduction based on different selection strategies,including reduction quality,run time and classification accuracy in cosistent and inconsistent decision table. The experimental result shows that three conditional entropy reducts have different advantages and disadvantages. The above work can give a valuable reference for application.
Robot Behavior Control Based on Multi-LCS and the Artificial Potential Field
SHAO Jie,YANG Jing-yu
Computer Science. 2011, 38 (1): 264-267. 
Abstract PDF(341KB) ( 319 )   
RelatedCitation | Metrics
Due to premature convergence, local optimal solution, accounting for a larger storage space and other shortcomings of genetic algorithms, a simulated control system for robot was designed by using distributed learning classifier system and Artificial potential fields to perform complex behaviors. A set of enhanced solutions of cover detectors problem was suggested and compared with each other in order to make the simulated robot more effective in choosing the appropriate behavior and improving the performance of the robot.
Feature-level Fusion Fault Diagnosis Based on PCA
WU Qian,CAI Hai-ni,HUANG Li-feng
Computer Science. 2011, 38 (1): 268-270. 
Abstract PDF(257KB) ( 851 )   
RelatedCitation | Metrics
In most of current application,fault diagnosis based on single sensor show bad performance and is difficult to achieve satisfactory accuracy owning to noises. In this paper,a multi-sensor multi-feature fusion fault diagnosis method was proposed. The method firstly collects part operating conditions using multiple sensors installed on it, Then a fusion model based on principal component analysis was constructed to fuse all extracted features from the different sensors.Finally,fault diagnosis was carried out according to the fused results. Experiments and results show that new method can achieve perfect performance which is also better than that achieved by singlcscnsor.
Noised Image Segmentation Effect Evaluating Method Based on Run Length Statistic
ZHANG Xue-feng,FAN Jiu-lun,XIE Xie
Computer Science. 2011, 38 (1): 271-275. 
Abstract PDF(461KB) ( 338 )   
RelatedCitation | Metrics
There is close relationship between segmentation effect of noise disturbed image and noise intensity the image contained. In order to quantize computing and analysis of segmentation effect of noise disturbed image, a evaluating function for noised image's segmentation results was presented based on segmentation image which is not noised firstly. Then a noise disturbed image segmentation effect evaluating method was presented that is based on run length testing of sequence randomness verifying. Run length function is defined based on run length statistic of segmented image's binary gray matrix and this index is quantifying computed when evaluating noised disturbed image's segmentation effect Simulation results of gray image,which is disturbed by noise,indicate that run length function is expressed simple, computing convenience, and can nicely reflect the changing relationship between segmentation effect of noise disturbed image and noise intensity the image contained.
Quality Assessment for 3-D Model Based on Surface Roughness
YANG Bin,LI Xiao-qiang,LI Wei,DING Guang-tai
Computer Science. 2011, 38 (1): 276-278. 
Abstract PDF(382KB) ( 598 )   
RelatedCitation | Metrics
In the process of 3-D model watermarking, emhedding watermark will make distortion on the 3-D model. Assessing the quality of 3-D watermarked model correctly can hchave as one of the principles of assessing 3-D watermarking scheme. In this paper, a quality assessment measure for the 3-D model was proposed. To compare with traditional quality assessment methods such as SNR ratio and Hausdorff, this method is more suitahle for 3D models. First, we calculated the roughness of the mesh model with the dihedrals which is the primitive of a 3D model. Second, we calculated the roughness of the watermarked model after watermarking. Last we calculated the increment of roughness between the original mesh and the watermarked mesh in order to assess the distortion made by watermarking. Moreover, the proposed method can estimate the distortion of watermarked mesh model made by various attacks.
Shape Recognition and Classification Based on Skeleton Junction Nodes Features
CHEN Jun,WANG Bo,ZHOU Yu,BAI Xiang,LIU Wen-yu
Computer Science. 2011, 38 (1): 279-281. 
Abstract PDF(253KB) ( 313 )   
RelatedCitation | Metrics
This paper proposed an algorithm using skeleton information to do shape recognition and classification. Using the information of skeleton junction nodes,the problem of shape similarity is converted to distance of skeleton junction nodes. Using method of neural network, the classification results can be got at the output of the network. The method can solve the recognition of nonrigid objects. The experiment results show that our method has better recognition results than other method.
New Fractal Model of Grid Overlapping Differential Box-counting
LI Chuan-long,LI Ying,YU Shui-ming
Computer Science. 2011, 38 (1): 282-285. 
Abstract PDF(366KB) ( 334 )   
RelatedCitation | Metrics
One of the most important feature of the fractal dimension is that its fractal dimension is largely correlated with the surface roughness of image feeled by the human eye. The fractal has many successful applications in analysis,segmentation and classification of images, the differential box-account method is often used in the fractal dimension estimation based on fractal theory. Least squares is better for detailed textures, but is not satisfactory for more coarse textures, in order to solve this problem, overlapping grid differential box-account model was presented, the grids which the numbers of the boxes arc calculated by arc overlapped, the numbers of boxes of the whole image arc calculated on the statistical self-similarity. Linear fit of least square of this model is better,in order to improve the precision of box-counting,non-integer box-counting and zoom-scale-fixing were presented,the fractal dimension of the image texture got by the model was validated, compared with differential box-account.
Research of Representation Based on Feature Semantic for Models
JIN Ying-hao,SUN Li-juan
Computer Science. 2011, 38 (1): 286-289. 
Abstract PDF(348KB) ( 402 )   
RelatedCitation | Metrics
To fit the request of feature semantic modeling system better, encapsulate more complex semantic with fealures in the CAD/LAID system, and increase the modeling efficiency, a feature-semanticbased representation was proposed. It built all features' faces by the feature semantic,managed all the features' elements on the cellular model,and verified the validity of the feature semantic by testing the wholeness of faces instead of constraint solving used in the modeling system which is based on the traditional representation. This representation can not only represent complex semantic of product models, but also enhance the performance of the semantic feature modeling system greatly. Experiments on computer show that this new algorithm is more adaptable and practicable.
Static Analysis of Soft Errors Effect in Register Files for Program Reliability
XU Jian-jun,TAN Qing-ping
Computer Science. 2011, 38 (1): 290-294. 
Abstract PDF(582KB) ( 475 )   
RelatedCitation | Metrics
Subsequently to the wall of performance and power consumption, the dependability of computing caused by soft errors has become a growing concern. Since register files are accessed very frequently and can not be well protected,soft errors occurring in them arc one of the top reasons for affecting the reliability of program.To access the effect of soft errors in register files, a static analysis approach for program reliability was presented based on the assembly codes. Firstly, all possible live interval of registers, which may degrade the program reliability, were sketched through the data flow analysis techniques; then the execution time and frequency of each live interval were analyzed according to the expression of execution path;finally the program reliability can be calculated under the occurrence of soft errors in register files. Experiments show that the analytical results arc compatible with the AVF methods'. Moreover, the execution time and frequency of all involved interval have been presented, which arc in favor of implementing the high effitient fault tolerance methods for soft errors in register files.
Cache Resource Management Mechanisms of Chip Multiprocessors
JIA Xiao-min,ZHANG Min-xuan,QI Shu-bo,ZHAO Tian-lei
Computer Science. 2011, 38 (1): 295-301. 
Abstract PDF(665KB) ( 505 )   
RelatedCitation | Metrics
With the trend towards Chip Multi-processors(CMPs) for the next leap in computing performance and the increasing of on-chip cache capacity, many architectures have explored cache resource management for better use and allocation of cache resource to accommodate various optimizing goals. Representive cache resource management mechanisms were evaluated and divided into two main catogaries, i. e. Cache Partitioning and Cache Sharing, based on their research focus. For cache partitioning, its main components and common form were described. For cache sharing, the key research directions were presented. Several mainstream cache partitioning mechanisms and cache sharing mechanisms were then investigated and compared respectively. The conclusion is that hardware and software co-designed page management is a promising field for future cache partitioning research, and also cache sharing mechanisms should be based on the thorough analysis of application background and the corresponding cache access behaviors.
Research of a Method for Hardware and Software Co-design Based on RSOC
LIU Zhao-yu,LI Ren-fa,CHEN Yu
Computer Science. 2011, 38 (1): 302-305. 
Abstract PDF(406KB) ( 332 )   
RelatedCitation | Metrics
This paper presented a hardware-software co-design method for overcoming the design complexity and programming difficulty of reconfigurable system on chip(RSOC).The overall framework of the method is based on specific hardware-software co-functions. We gave the basic processes of the method, involved key technology and implementation methods. Also,verified a key part of design method and feasibility. The method shields hardware-software implementation details of specific function modules for target application designers, improves operational efficiency and flexibility of the application based on RSOC.