Current Issue

• 2011, No. 12　Published:24 December 2011 Last issue    Next issue
  Columns in this issue: 论文

• 论文

  A Survey of Data Transport Protocols in Space Networks

  WANG Baosheng1,CAI Guilin1,2,CHEN Lin1,LIU Yaping1

  2011, 33(12): 1-6. doi:

  Abstract ( 322 )   PDF (457KB) ( 338 )   　　

  The transport protocol is the basis of effective operation for the network. The long propagation delay, high bit error rate and other characteristics raise significant challenges to the design of the transport protocols of space networks. In this article, we classify the typical transport protocols of space networks into groups according to their design principles, and discuss the mechanisms, characteristics and limitations of each protocol. Finally, we summarize the research status and point out the development direction of the transport protocols in space networks.

  A Congestion Control Scheme Based on the Price Mechanism  in Wireless Access Networks

  YAN Liming，NIU Yugang

  2011, 33(12): 7-11. doi:

  Abstract ( 324 )   PDF (540KB) ( 243 )   　　

  Due to the wireless access networks’ characteristics of large delay, small throughput and high packet loss rate, the current active queue management (AQM) is not applicable to the wiredcumwireless network. This paper analyzes the price production mechanism of Random Exponential Marking (REM) and points out its shortcomings. By improving the price with queuebuffer proportion, a new pricebased congestion control mechanism (PREM) is proposed. PREM uses queuebuffer proportion and virtual capacity as congestion measures. Finally, the proposed algorithm is validated on the NS2 simulation platform and compared with REM, PI and ProxyRED.It is proved that PREM can reduce the packet loss effectively.

  MultiAgent System SPIN Model Checking Based on the Temporal Logic of Knowledge

  LONG Shigong,WANG Kouwu

  2011, 33(12): 12-16. doi:

  Abstract ( 285 )   PDF (363KB) ( 365 )   　　

  A SPIN model checker is mainly used to check whether a system satisfies the specifications expressed in linear temporal logic, and the specifications of the distributed systems and protocols expressed in knowledge logic are more convenient. In this paper, the model checking approaches for the temporal logic of knowledge using SPIN are discussed. According to the theory of partial proposition, the problem of how to model check knowledge operators and public knowledge operators is converted into model checking linear temporal logic. These approaches make SPIN’s functions extended from model checking temporal logic to the linear temporal logic of knowledge. The article discusses the approach of model checking the linear temporal logic of knowledge by analyzing an instance of the RPC protocol.

  Research and Application of the Technology for Preventing Illegal Copying the ARM Microcontroller Code

  HUANG Yiping1,NONG Liping2,SU Jiande1,LUO Xiaoshu1

  2011, 33(12): 17-21. doi:

  Abstract ( 332 )   PDF (626KB) ( 240 )   　　

  After studying and analysing the application of the ARM microcontroller and the cracking technology, according to the problem that the program code is  fetched by hackers through illegal methods, a method of preventing illegal copying the program code for the ARM microcontroller is proposed. The information of encryption for a worldwide unique serial number inside the chip is authenticated in the program running, which ensures that only the system code of the legal controller runs correctly. The result shows that the method is of high security, and is convenient and practical, and has been used in the ARM microcontroller or the smart card system, and it plays a role of very good protection for system code and will be widely used in the future.

  L2-L∞Output Tracking ControlBased on Stochastic Uncertain Network

  QI Ji

  2011, 33(12): 22-26. doi:

  Abstract ( 311 )   PDF (439KB) ( 245 )   　　

  This paper studies the model reference output tracking problem for stochastic networkbased control systems with a single delay and parameter uncertainties residing in a polytope. Attention is focused on the design of controller, which guarantees that the stability of systems and a prespecified L2L∞ performance γ for the stochastic networkbased control systems with respect to all energybounded input signals. Sufficient conditions are proposed in terms of linear matrix inequality, and the corresponding controller design is cast into a convex optimization problem. A numerical example illustrates the effectiveness of the proposed L2L∞ output tracking design scheme.

  Research of the Heterogeneous MultiCore Processor Architecture Design

  CHEN Fangyuan,ZHANG Dongsong,WANG Zhiying

  2011, 33(12): 27-36. doi:

  Abstract ( 402 )   PDF (667KB) ( 677 )   　　

  Nowadays processors advance towards multicore architecture. In heterogeneous multicore processors, different kinds of computing tasks can be assigned to different processor kernels in order to perform parallel execution, which can provide more flexible and more effective principles. Nowadays the heterogeneous multicore architecture becomes the research hot. This paper introduces the design of heterogeneous multicore processors from architecture, and analyzes the challenges in terms of some important aspects: architecture of cores, interconnect, storage system, operating system support, test and validation, DVFS and so on. The challenges for the design of heterogeneous multicore processors are discussed. Finally we propose the design key points for high performance and embedded realtime applications: the performance and the number of cores for high performance heterogeneous multicores, realtime and lower power for embedded realtime multicores.

  Multi_MINT：A Multiprocessor Simulator Based on MINT

  CHEN Hu,LUO Weiliang,GAN Yunyun

  2011, 33(12): 37-43. doi:

  Abstract ( 323 )   PDF (958KB) ( 243 )   　　

  Multiprocessor simulators are very important tools to design, analyse and optimize the multiprocessor system structure. The Multi_MINT simulator we propose in this paper is composed of two parts: the front end and the back end. The front end adopts MINT to simulate the execution of the MIPS instructions, and the back end contains Cache unit, the connection_network unit and the memory controller unit. The two parts constitute a complete multiprocessors simulator system. This paper mainly introduces the general structure of Multi_MINT and the simulation framework of the main hardware components, and then discusses some details of the design and implementation of Multi_MINT, and finally illustrates how Multi_MINT plays the role of performance evaluation by an example.

  An Optimized Device Driver Model Based on MicroKernel Virtualization

  LIU Zhenyu,WU Junjun

  2011, 33(12): 44-51. doi:

  Abstract ( 320 )   PDF (625KB) ( 286 )   　　

  With the embedded system becoming more and more complicated, the efficiency and reliability turns to be an obvious and serious problem. In our paper, we analyze the present device driver models ,and propose a new optimized device driver model based on the L4 microkernel virtualization to serve the special requirements of embedded devices. With this model, device drivers can be shared among multiple virtual machines, and the reliability can be greatly improved ,making up the performance cost brought by virtualization, and providing a virtual environment with the relative balance of efficiency , function, and reliability.

  Design and Implementation of the Air-Conditioning Hardware in the Loop Simulation System Based on Modelica

  ZHAO Jianjun,HE Xiangliang,DING Jianwan

  2011, 33(12): 52-56. doi:

  Abstract ( 327 )   PDF (1032KB) ( 282 )   　　

  To analyze the whole performance of the airconditioning system containing different domains, the models of serial port communication and airconditioning are established based on the multidomain unified modeling language Modelica, the airconditioning hardware in the loop(HIL) simulation system is created by connecting and airconditioning controller and model with the serial port communication model and the signal process circuitry, and the result of simulation shows the effectiveness of the system.The HIL simulation system of airconditioning provides a new method of HIL simulation for multidomain systems, which improves the efficiency of production design and reduces the cost evidently.

  A Method to Check Biological Tissue Denaturation Based on Mode B Ultrasonic Image Processing

  ZOU Xiao,QIAN Shengyou

  2011, 33(12): 57-60. doi:

  Abstract ( 297 )   PDF (567KB) ( 271 )   　　

  A method using BP neural network was presented to check biological tissue denaturation based on grey value and wavelet transform coefficient of Mode B Ultrasonic Image in HIFU treatment. The digital subtraction was executed for mode B ultrasonic images obtained from fresh vitro pork before and after HIFU radiation, and then its image features were extracted and classified by neural network . Through analyzing the statistical properties and errors of plentiful experimental data, the distinguishing effects could be obtained with finished. The results show that combining the average grey with the wavelet transform coefficient to construct BP neural network, the denaturation of biological tissue can be monitored better during HIFU therapy.

  A Corner Match Algorithm Based on the TwoDimensional Gabor Transform

  ZHOU Delong，LIU Ying，WU Ginyi

  2011, 33(12): 61-65. doi:

  Abstract ( 320 )   PDF (575KB) ( 199 )   　　

  The core issue of image registration is to increase the speed and accuracy of registration, but the result of the registration much depends on the matching precision of features.In order to increase the matching precision of features, an algorithm based on the twodimensional Gabor wavelet transform for corners matching is proposed in this paper. Firstly,it adopts an improved Harris corner detector to extract corners and obtains the coordinates of corner positions,and utilizes multi masks of twodimensional Gabor wavelet to filter the reference image and unregistered image to extract the duplicate Gabor wavelet coefficient of corner coordinates from filtering the image, and takes this as the feature description of corners .Secondly it introducs two kinds of similarity measurement to match corners.Adopting the longest common subsequence is successful to extract more points of the same name and has a higher match rate ,which has been proved by multiexperiments when parameters are chosen appropriately.

  Information Analysis and Visualization of Social Network Based on Domain Ontology

  WU Peng,LI Sikun

  2011, 33(12): 66-71. doi:

  Abstract ( 306 )   PDF (748KB) ( 478 )   　　

  Social network analysis and visualization is an active area of study which lacks good comprehensions and organizations of social network information. This paper proposes a domain ontologies model focusing on social network information, which describes the impersonal existences and the relationships among them in the social network information domain. The model is suitable for describing a wide variety of the social network analysis and visualization methods and can extend easily for different applications of information visualization of the social network. It improves the force directed the layout algorithm for overcoming the disadvantage of traditional force directed layout algorithm in analyzing and visualizing the structure of social network. The graphic results intuitively display the layout of subgroup and the close degree between actors, display key attributes and link characteristics etc. The results with the case of terrorist information show that our model can be excellent in analysis and visualization of social network information.

  The Construction of a Selfadaptive MultiAgent System Based on Reinforcement Learning

  SHEN Le,MAO Xinjun,DONG Menggao

  2011, 33(12): 72-77. doi:

  Abstract ( 298 )   PDF (594KB) ( 236 )   　　

  The environment of selfadaptive systems is often uncertain, and the changes are difficult to predict. To develop such complex selfadaptive software systems has become a great challenge in the domain of software engineering. Reinforcement learning is an important branch of machine learning. A reinforcement learning system can learn the optimal mapping policy from the states of environment to the actions by means of trailanderror. Aiming to deal with the uncertainty of environments, this paper combines the agent technology and the reinforcement learning technology together, and proposes an adaptive mechanism based on reinforcement learning and the corresponding approach to construct complex selfadaptive systems that can adapt to the changes of uncertain environments. A case is illustrated to validate the effectiveness of the proposed mechanism and approach.

  A New Method of Detecting the Structural Clone

  GUO Jing，WU Junhua

  2011, 33(12): 78-83. doi:

  Abstract ( 267 )   PDF (578KB) ( 256 )   　　

  The structural clone is a clone with a similar principle and thinking structure, which does not rigidly adhere to the order of statements and expressions. Program dependence graph can better reflect the program's internal relations, and the expression becomes more intuitive to understand. So we first detect the clone code based on the program dependence graph and the process using the  program slicing techniques.Then we summarize  the characteristics of the structural clone, quantify and generalize the characteristics of the program dependence graph to calculate the similarity. It is useful to get a better assessment of similarity and improve the rate of clone.

  Convergence Analysis of Cloud Search Algorithms

  JI Yanfang,CAO Ju

  2011, 33(12): 84-86. doi:

  Abstract ( 278 )   PDF (324KB) ( 245 )   　　

  We analyse the global optimization problems of the cloud search algorithm, and the algorithm can be seen throughout the search process which includes drift operator, rain clouds of the Health operator, contraction expansion operator. If the contraction of the expansion operator fails, then the algorithm will compare the case of a bad search. In this case, we use the knowledge of probability theory analysis of the convergence of the algorithm. The results indicates that the search algorithm converges to the probability of a global optimal solution.

  Analysis of the Parallel Computing Performance of Ocean Model FVCOM2.6 Using TAU

  SONG Qian,HU Song

  2011, 33(12): 87-93. doi:

  Abstract ( 348 )   PDF (1023KB) ( 304 )   　　

  This study applies Tuning and Analysis Utilities (TAU) to analyze the parallel performance of the unstructured grid FiniteVolume Coastal Ocean Model (FVCOM) version 2.6 based on Message Passing Interface (MPI). Examples of ShenHu Bay FVCOM tidal models, with low resolutions (2108 and 10378 nodes) and high resolutions (15347 and 26033 nodes), are  tested using various processes on a linux cluster (Intel Xeon CPU E5450 and 10G InfiniBand). The results show that the advection subroutines occupied large proportion of running time as the models ran on a single process. The speed up of each test is examined; the grid number which affected the parallel performance as the models ran on multiple processes. Under the hardware condition of this study, each test had an optimal number of processes, which are  32 for low resolutions and 64 for high resolutions. The optimal number of processes is increased as the resolution increased. The total run time started increasing as the number of processes exceeded the optimal number. The TAU analysis shows that it is mainly due to the increasing times of calling MPI_Waitany subroutine so that the barrier time increased nearly proportionally to the total time, which provides information to improve the parallel performance for FVCOM in the future.

  Research of Web Page Classification Based on Improved Immune CloneSupport Vector Machine

  ZHANG Suqi1,LIU Enhai2,HE Ya2,DONG Yongfeng2

  2011, 33(12): 94-98. doi:

  Abstract ( 310 )   PDF (566KB) ( 201 )   　　

  Web page classification is an extended hot field for solving the problem of information overload ,with the excellent ability to learn, support vector machine shows a specific advantage in solving high dimensional problems. A new classification algorithm based on the combination of support vector machine and improved immune clone is proposed after the research of support vector machines and standard immune clones. As the standard algorithm achieves antibody variants through inverting  randomly some bits in antibody coding, so it is not strong in searching capability, for this deficiency, the paper distinguishes memory units and normal units, defines adaptive probability for the memory units, thereby strengthens search capability in the neighborhood of the current optimal solution, thus accelerates the speed to find the global optimal solution. A lot of experiments have shown that the improved algorithm which has a better parameters selection effect and a higher efficiency is a web page classification method with high accuracy and efficiency.

  An Ensemble Classifier for Mining Imbalanced Data Streams with Noise

  OUYANG Zhenzheng1,TAO Zijin1,CAI Jianyu2,WU Quanyuan1

  2011, 33(12): 99-105. doi:

  Abstract ( 299 )   PDF (828KB) ( 242 )   　　

  Many real world data streams mining applications involve learning from imbalanced data streams, where such applications expect to have a higher predictive accuracy over the minority class, however most classification models assume relatively balanced data streams, and they cannot handle imbalanced distribution. In this paper, we propose a novel ensemble classifier framework (IMDAP) for mining conceptdrifting and noisy data streams with imbalanced distribution by using an averaged probability ensemble framework and sampling technique. Our empirical study shows that the IMDAP is superior and have improves both the capability of the classifier and the accuracy in performing classification over the minority class.

  A New Method of Fuzzy Entropy of Vague Sets

  XU Fengsheng

  2011, 33(12): 106-109. doi:

  Abstract ( 309 )   PDF (329KB) ( 228 )   　　

  The drawbacks of the existing definition on the fuzzy entropy of vague set are pointed out and analysed. We give out the axiomatizing definition on the fuzzy entropy of vague set, and propose a new fuzzy entropy of Vague set, and prove its rationality and validity in this paper.

  A Weighted Rough Clustering Algorithm Based on New Similarity Measure

  SUN Xiaobo，LIAO Guiping

  2011, 33(12): 110-115. doi:

  Abstract ( 323 )   PDF (601KB) ( 255 )   　　

  Clustering is a major research orientation in data mining．Considering the drawbacks of the existing clustering algorithm, a new similarity measure is proposed firstly. Then the discernibility ability of the rough set theory is used to measure the importance of attributes, and thus a weighted rough clustering algorithm based on new similarity measure is proposed. Finally，we test our algorithm versus other algorithms on the UCI datasets, and the experimental results show the proposed clustering algorithm can deal with the categorical data, and does not need to be given the number of cluster, and especially, it improves the cluster quality.