Current Issue

• 2016, No. 06　Published:25 June 2016 Last issue    Next issue
  Columns in this issue: 论文

• 论文

  A coarse grain reconfigurable SoC design
  method based on architecture template   

  SHEN Jianliang1,2,LI Sikun2,WANG Guanwu2,L Ping1,LIU Lei2,LIU Qinrang1

  2016, 38(06): 1071-1077. doi:

  Abstract ( 156 )   PDF (646KB) ( 481 )   　　

  Since the traditional design methods of MPSoC have large design exploration space and high design complexity, we propose a coarse grain reconfigurable SoC design method based on architecture template. The architecture template can be reused, and its parameters can be configured, so the method can reduce the design exploration space, and improve the design efficiency and computational performance. The design templates are instantiated, and a cryptographic applicationspecific multireconfigurable instruction sets processor SoC(MultiRISP SoC) is constructed. Experimental results show that the performance of the MultiRISP SoC is equivalent to several typical reconfigurable platforms, but its system is more efficient.

  Design and verification of parallel algorithm of metadata
  index reorganization for meteorological data archiving  

  XU Jing1，REN Kaijun2

  2016, 38(06): 1078-1085. doi:

  Abstract ( 143 )   PDF (1039KB) ( 534 )   　　

  Meteorological archival and retrieval system (MARS) enables archival and retrieval management for meteorological data,  including many numerical prediction products. However, with the refinement of numerical forecasting techniques, the data under the MARS management is soaring in a massive growing trend, which presents a new challenge to archival and retrieval techniques. As an archival issue in the physical view of MARS, the metadata index currently has very low update efficiency. To address the problem, we propose a parallel approach to accelerate index update speed, which reduces the index reorganizing overhead during data archiving time. Experimental results show that compared with the original serial one, our parallel approach accelerates the maximum update speed by 3 times on average, hence improving the efficiency of meteorological data archiving.

  A parallel computation method of wall distance
  for largescale flow simulation  

  YANG Yongguo1,CHENG Xinghua2,WANG Wanjin1

  2016, 38(06): 1086-1090. doi:

  Abstract ( 122 )   PDF (621KB) ( 396 )   　　

  Based on MPI and Open MP, the recursive box method is parallelized. Then the preprocessing algorithm is improved. The scramjet combustor case, which has hundreds of millions of grids, is tested on the Tianhe2 system of National Supercomputing Center in Guangzhou. The results shows that the process box method and the boundary box method do not need to define the splitting box number, and the boundary box method has an excellent speedup, thus outstandingly enhancing the efficiency of the wall distance calculation.

  A topology optimization algorithm
  based on reciprocal capability in P2P networks    

  LIU Hao1，ZHANG Lianming2，HE Wenhua1

  2016, 38(06): 1091-1096. doi:

  Abstract ( 114 )   PDF (552KB) ( 327 )   　　

  Efficient topology optimization algorithms are one of the research hotspots in the domain of unstructured P2P networks. In view that the existing P2P network topology optimization algorithms are mostly based on an ideal network environment, and the selfcapacity and externalsurroundings of nodes are not considered comprehensively, we present a topology optimization algorithm based on reciprocal capability in P2P networks. The reciprocal capability of nodes is calculated according to two aspects: selfcapacity and externalsurroundings, based on which the topology of the unstructured P2P network is optimized. Analysis and simulation show that the proposed algorithm is capable of forcing nodes of lower reciprocal capability to the margin of P2P networks, reducing their impact on the overall performance of the network and effectively improving the search efficiency of P2P networks.

  An improved ElGamal digital signature scheme    

  LI Lijuan，GUO Yajie

  2016, 38(06): 1097-1102. doi:

  Abstract ( 144 )   PDF (385KB) ( 403 )   　　

  Digital signature plays an irreplaceable role in modern information security in recent years. ElGamal Digital signature is an important discrete logarithm digital signature scheme, however, the original ElGamal signature scheme has some safety and efficiency problems. We propose a new kind of ElGamal digital signature scheme and prove that it can withstand a new type of forgery attacks and homomorphism attacks. We also analyze and explain the time complexity of the improved digital signature scheme.

  A minimum connected dominating set construction
  algorithm based on multiple panning trees and
  sub networknode degree united weight   

  TANG Qiang，XIE Mingzhong，LUO Yuansheng

  2016, 38(06): 1103-1110. doi:

  Abstract ( 127 )   PDF (755KB) ( 307 )   　　

  We propose a minimum connected dominating set (MCDS) construction algorithm called static wireless networks’ minimum connected dominating set (SWNMCDS) based on multiple spanning trees and sub networknode degree united weight for static wireless networks. At first, a probability p is set, and every node randomly generates a probability. The nodes whose generated probability is less than p become root node candidates. The root node candidates within two hops exchange information with each other to determine the final root nodes. Based on the node weight, the root nodes generate multiple connected trees respectively. All the connected trees are connected with each other into a minimum connected dominating set based on the selected connected nodes according to the sub networknode degree weight. Analysis results show that the upper bound of the performance ratio is 2β(2+H(Δ)), and the time complexity as well as the message complexity is O(Δ2), where Δ is the size of the biggest one hop neighbor set, and β  is the number of spanning trees. Compared with the traditional algorithm MCDS, the SWNMCDS has a smaller connected dominating set.

  A review on public service platforms of
  collaborative industry chain     

  LI Ying1，WANG Chenxiao1，YANG Chen1,2，LI Xiao1

  2016, 38(06): 1111-1117. doi:

  Abstract ( 123 )   PDF (522KB) ( 333 )   　　

  We first analyze the difference of interrelated terms, including the connection and distinction among industry chain, value chain and supply chain; the difference between the ASP and the SaaS in corresponding relation, service fields and the userdefined requirement; the difference among the SaaS, the IaaS and the PaaS in service modes, service configuration and core technologies. Then we summarize the research status on architecture, component management, accomplished technology of public service platform, and divide the research status of accomplished technology into four parts: application configuration technology, DAAS technology, dynamic evolution technology and safety technology. We also analyze the difference between the public service platform of collaborative industry chain and the general public service platform in architecture, configuration technology, Dass technology and security technology. We find out that current research on collaborative industry chain lacks an integrated information service platform and collaborative product development, hoping our work can provide some basic documentation for further relevant research.

  Privacypreserving in data publishing based on fuzzy sets

  LIU Yinghua

  2016, 38(06): 1118-1122. doi:

  Abstract ( 123 )   PDF (449KB) ( 347 )   　　

  Privacypreserving data publishing becomes a hotspot technique in privacy preserving research, which mainly focuses on how to avoid leakage of sensitive data in data publishing, and at the meantime ensures efficient use of data. We propose a fuzzy kanonymity model for privacypreserving data publishing. We first calculate the prior probability of training sample data, and then achieve privacy protection through the Bayesian classification of a single sensitive attribute and two associated attributes. Theoretical analysis and experimental results show that compared with the classic kanonymity model, the proposal is more efficient, preserves more information, and has stronger practical applicability.

  A distributed time synchronization algorithm
  based on information fusion 

  TANG Bo，WU Shuangshuang,PENG Li,SUN Chao

  2016, 38(06): 1123-1127. doi:

  Abstract ( 159 )   PDF (651KB) ( 332 )   　　

  As a support technology of wireless sensor networks, time synchronization algorithms become a research hotspot in recent years. Centralized time synchronization algorithms have been widely studied, which use the reference node via broadcast, flooding or multi hop to obtain time synchronization of the whole network. They achieve good accuracy, but robustness and scalability are poor. We propose a distributed time synchronization algorithm, which realizes time synchronization by fusing the time information of the whole network nodes. The algorithm can eliminate the influence of the transformation of network topology and the change of nodes' density on centralized algorithms. Experimental results show that the distributed time synchronization algorithm achieves the whole network time synchronization on several typical network topologies and is robust to the change of network topology and the number of nodes.

  A personalized influence maximization
  algorithm based on influence path   

  YANG Shuxin,WANG Xi,PENG Qiuying

  2016, 38(06): 1128-1134. doi:

  Abstract ( 131 )   PDF (721KB) ( 367 )   　　

  Personalized influence maximization in social network has become a new branch of influence maximization study in recent years. Different from existing research that assumes equal propagating strengths of social network edges, our work aims to find out the topk most influential nodes for the target user without inappropriate assumption. We propose a maximizedinfluencepath algorithm (MIPA) based on the independent cascade model. It solves the problem through three stages. Firstly, to compute the propagating strengths from the nodes of social network to the neighbors of the target node, the strengths of edges are transformed into its logarithmic form for getting the maximized influence paths. Secondly, the strength of maximized influence paths which pass through different neighbors with the same source nodes are consolidated to calculate the node’s propagating strength on the target node. Finally, the seed set with high propagating strength is found out by selecting the topk nodes. We testify the algorithm on several realworld social networks. Experimental results validate the proposed algorithm.

  A user interaction processing strategy for overcoming blocking     

  CAO Zhonghua1，2，XIA Jiali1，2，PENG Wenzhong2，MAO Amin1，CHEN Weibin1，WU Dan1

  2016, 38(06): 1135-1140. doi:

  Abstract ( 126 )   PDF (915KB) ( 313 )   　　

  Mobile realtime interactive applications become a new kind of application mode of the mobile Internet. Interactive objects often unilaterally shield each other, which hinders direct interaction with each other. In fact, some of these interactive objects can achieve interaction through a third party. In view of this situation, we put forward an interactive state model and an interactive performance evaluation index. The interaction state judgment strategy is provided using the connected graph, the optimal path of interaction is identified by finding the shortest path, and the interaction processing strategy based on connected graphics (IPSCG) is proposed. Experimental results show that this strategy can significantly lower interaction cost, improve interaction efficiency, and enhance the stability of user interaction.

   An opportunity reliable transmission scheme based on
  cooperative  error propagation for wireless sensor networks 

  LIU Xicheng，XI Jianzhong

  2016, 38(06): 1141-1148. doi:

  Abstract ( 97 )   PDF (893KB) ( 353 )   　　

  In order to reduce the effect of error propagation in the process of cooperative communication in wireless sensor networks (WSNs), we propose a reliable transmission mechanism based on opportunistic error correction. First, three cooperative transmission schemes are proposed in the multinode cooperative WSNs, and the model of the cooperative error propagation is established based on channel quality and the symbol error rate. Then, aiming at the cooperative error, such as merging, scheduling, and interference, we propose an opportunistic error correction algorithm based on the channel, source and cooperative networks. Finally, based on the above conclusions, and taking the performance requirements of users, the complexity of scheduling and the model of cooperative communication into account, we propose an opportunity reliable transmission scheme. The results of mathematical analysis not only verify the higher reliability of the proposal than the static mechanism, but also prove its superior performance in real time, reliability, throughput and energy efficiency in end to end communications.

  Communication signal monitoring based on
  joint block diagonalization of matries 

  TANG Hui

  2016, 38(06): 1149-1155. doi:

  Abstract ( 114 )   PDF (895KB) ( 265 )   　　

  In the situation of noncooperative communication and military communication electronic warfare, multiple communication signals should be separated to extract some useful signals, which comes down to a problem of blind source separation of signal convolution. First the model of the received signals is built and transformed to the problem of joint block diagonalization. Then a new nonorthogonal joint block diagonalization algorithm is proposed. The new iterative algorithm is based on the steepest gradient descent method, and the optimization strategy and computation complexity are analyzed. Simulation results demonstrate the validity and reliability of the algorithm, which has a faster convergence speed than the existing methods.

  Artificial bee colony based ant colony
  optimization for continuous domains  

  ZHOU Niao1,2,GE Hongwei1,2,YUAN Yunhao2,SU Shuzhi2

  2016, 38(06): 1156-1163. doi:

  Abstract ( 129 )   PDF (617KB) ( 306 )   　　

  Continuous ant colony optimization is an important research direction of ant colony optimization algorithms. The ant colony optimization for continuous domains (ACOR) requires long computation time and is easily trapped into local optimal solutions, so we propose artificial bee colony based ant colony optimization for continuous domain algorithm (ABCACOR) to solve the problems. Firstly, an alternative mechanism instead of the original sortbased selection method is introduced to guide solution choice, which saves computation time and secures the diversity as long as possible. Secondly, the artificial bee colony search strategy is adopted to improve the  global search ability of the algorithm, thus the computation time is further reduced and the accuracy of solutions is improved. We evaluate the ABCACOR on a large number of test functions, and experimental results show that the ABCACOR outperforms some existing continuous ant colony optimization algorithms.

  An adaptive FA algorithm based on global information sharing  

  LU Kezhong1,2,SUN Jun3

  2016, 38(06): 1164-1170. doi:

  Abstract ( 106 )   PDF (729KB) ( 345 )   　　

  The firefly algorithm (FA) for high dimensional problems has some disadvantages, including slow convergence speed, low solving precision and unsatisfactory optimization effect. To overcome these disadvantages, we propose a novel adaptive FA algorithm based on global information sharing. Firstly, an adaptive control for gamma value is designed by the swarm distance. Secondly, the search process information of the firefly algorithm is updated to enhance its adjustment capacity of refinement. Thirdly, the global searching ability is improved by introducing the Delta potential well of the vector subspace based on the global mean location information. Simulation results show that the proposal has better convergence speed and precision than the basic FA and the PSO.

  A fruit fly optimization algorithm with
  CauchyGaussian dynamic reduction mutation

  DU Xiaoxin,ZHANG Jianfei,GUO Yuan,JIN Mei

  2016, 38(06): 1171-1176. doi:

  Abstract ( 165 )   PDF (703KB) ( 345 )   　　

  The fruit fly optimization algorithm (FOA) is easy to fall into local extremum and has slow convergence speed. To overcome these problems, combining the advantages of Cauchy mutation and Gaussian mutation, we propose the concepts of mutation effectiveness coefficient and CauchyGaussian dynamic reduction mutation factor and a fruit fly optimization algorithm with CauchyGaussian dynamic reduction mutation (FOACGDRM) as well, which takes into account of both the global exploration character and the local exploitation character, and which is  applied to improve the FOA. The FOACGDRM can enrich population diversity, effectively avoid local extremum and enhance the convergence speed. Simulation experiments on classic function instances and practical engineering instances verify the FOACGDRM, which show that the FOACGDRM is better in precision speed and stability.

  A unified mathematical model of population
   diversity for real and binary coded GA    

  ZHAO Hong1,2,LI Ying1,2,XIAO Wenjie1,2

  2016, 38(06): 1177-1182. doi:

  Abstract ( 167 )   PDF (731KB) ( 354 )   　　

  For the problem of lacking of unity and universality for existing population diversity definitions in  GA premature convergence research, we design a unified mathematical model of population diversity for real and binary coded GA based on the essence of gene population diversity. Firstly, the population matrix of real coded GA is converted into the same form as that of binary coded GA. Secondly, we define the concept of the homologous random variable and its characteristic measures, including the mathematical expectation, the deviation degree and the variance, based on which the unified population diversity model applicable to both of the two codes is established. The two representation methods (the evolution matrix and graphical representation) for the proposed model  are also given. Simulation analysis of GA test functions show that the proposed model can effectively reflect and analyze the change trend of GA population diversity in the process of evolution, as well as the convergence process and convergence results of genes. Finally, further analysis and research direction are pointed out.

  A particle swarm optimization algorithm
  based on local guidance and Gauss perturbation    

  WU Runxiu1,SUN Hui1,ZHU Degang2,ZHAO Jia1

  2016, 38(06): 1183-1192. doi:

  Abstract ( 117 )   PDF (806KB) ( 353 )   　　

  In order to solve the problem of low convergence rate and premature convergence of the particle swarm optimization (PSO), we propose an improved PSO algorithm which is based on local guidance and Gauss perturbation. Two measures on the particle velocity updating formula are proposed to improve the PSO. Firstly, social cognition is removed and the particles are only locally guided. Secondly, the Gauss perturbation term controlled by the global optimal particle is added. The combination of the two improvement measures can avoid the premature convergence problem and accelerate the convergence speed. Comparative experiments show that the improved PSO algorithm achieves better performance and stability than the classic PSO algorithm. Effect analysis experiments on the two improved measures and the comparative experiments on Gauss perturbation and social cognition further verify the effectiveness of the proposed algorithm.

  A quick autocalibration method of Kinect    

  MENG Bo,LIU Xuejun

  2016, 38(06): 1193-1199. doi:

  Abstract ( 287 )   PDF (635KB) ( 462 )   　　

  We propose a quick autocalibration method towards Kinect sensors to solve the problems such as large error and slow computation speed. Firstly, the corner points of the RGB image are extracted to form the corner point set. Then the point cloud of checkboard making from the RGB image is projected using the 3D  Hough transform detection method to the depth image in order to autoextract the point cloud corners. Then the point corners in the depth image are extracted. Thirdly, the point cloud corners are projected to the depth image using the least squaring plane fitting algorithm. Then the point cloud corners of the depth images are registered to the corner points of the RGB image. Finally, the 3D coordinates of these corner points are made and the pose transform matrix can be calculated. Experimental results show that the proposed algorithm can extract corner points automatically and decrease the average calibration time from 218 ms to 166ms, thus realizing a quick autocalibration of Kinect sensors.

  A license plate location method based on classifier voting      

  WANG Qinmin,LI Kuan,YANG Canqun

  2016, 38(06): 1200-1206. doi:

  Abstract ( 114 )   PDF (772KB) ( 349 )   　　

  We propose a votingbased multiclassifier method to improve the location accuracy of similar and distorted license plates. We enhance the license plate location accuracy from two aspects. First, two new plate descriptors are proposed according to the image characteristics of similar license plates and distorted license plates, which particularly promotes the location performance of the two types of license plates. Then several SVM classifiers are trained using different descriptors respectively, and a votingbased fusion mechanism is adopted to make the final decision, thus further enhancing the location accuracy.  Experimental results show that: 1) compared with the traditional wavelet and LBP license plate descriptors, the two proposed descriptors can effectively improve the location accuracy of distorted plates, and reduce the error rate of similar plates; 2) the votingbased fusion mechanism can effectively reduce the classification error rate of candidate plate regions from 3.05% to 0.8%.