15 April 2016, Volume 42 Issue 4
-
Select all|
-
Computer Engineering. 2016, 42(4): 1-6. https://doi.org/10.3969/j.issn.1000-3428.2016.04.001Deadlock detection is in general difficult in concurrent program.Aiming at this problem,this paper proposes a method of deadlock detection in multi-threaded program based on Petri net.It defines a Petri net model which describes the lock operation in multi-threaded program.Based on the current deadlock detection algorithm using Mixed Integer Programming(MIP),it proposes an improved MIP algorithm to detect whether the model exists deadlock.Experimental results show that the improved MIP algorithm can detect the deadlock in Petri net model.It has higher computational efficiency when dealing with larger programs,compared with flag matrix algorithm and reachability graph.
-
Computer Engineering. 2016, 42(4): 7-13. https://doi.org/10.3969/j.issn.1000-3428.2016.04.002The storage subsystem is a bottleneck as the server has to handle a lot of I/O operations in a diskless workstation system.In order to solve the problem,this paper presents a data distribution method based on Solid State Drive(SSD) and Hard Disk Drive(HDD) hybrid storage.It allocates OS and application image files to SSD while allocating the modified data of the image files generated by running workstations to HDD.It presents a block distribution model based on I/O features of diskless network server,and designs a dynamic migration method between SSD and HDD.The distribution method computes data distribution probability according to data access features of SSD and HDD,and then determines where a data block should be placed.Experimental results show that this method can improve the workstation boot speed,widen bandwidth of the shared storage and cut down the count of writing operations to SSD.
-
Computer Engineering. 2016, 42(4): 14-19. https://doi.org/10.3969/j.issn.1000-3428.2016.04.003Under the serious threat of earthquke to defense engineering,the evolution in the brittleness of the complex risk system often leads to system crash,and the complex risk system has the limitations in the fuzzy characters association explaination and reasoning computing operations.Therefore,according to the evolution discrimination and reasoning computation problem in the risk system,a new earthquake-damage risk assessment method is proposed for defense engineering on the basis of complex brittleness and Bayesian Network(BN).Through discriminating the brittleness environment and structure characteristics of the earthquake-induced complex risk system for defense engineering,the brittleness structure elements are designed as the node variables of BN model,and the network structure and parameters are determined,which achieves the implicit expression of the uncertain logic association knowledge and multi risk status for the earthquake -damage risk system.Finally,the earthquake-damage risk BN assessment model and components are established for defense engineering based on complex brittleness,and are applied into the risk analysis system software about earthquake-damage defense engineering.Case results show that the brittleness risk report obtained with this method includes the probability of all risk levels and that of overall risk,and has high scientific and practical value.
-
Computer Engineering. 2016, 42(4): 20-26. https://doi.org/10.3969/j.issn.1000-3428.2016.04.004Missing data filling is an important issue in the field of data cleaning.For the vast majority of local filling methods realize classification on the basis of all attributes without considering the correlation between object attributes,this paper puts forward a missing data filling algorithm based on incomplete set biclustering.This algorithm fills missing data based on the theory that the mean squared residue of biclustering perfect cluster is 0 and the fluctuation of the cluster’s attribute values is consistent.This paper translates the problem of finding the maximum perfect cluster which contains the missing values into the problem of finding out the maximum similarity attribute sets between the missing object and other objects through mathematical analysis,then the majority of missing values which is used as the filling value can be calculated by the same maximum similarity attribute sets.This paper takes experiments using 4 groups of UCI data sets,and it is demonstrated that the proposed algorithm averagely improves the accuracy of 77.13% filling values compared with ROUSTIDA algorithm.
-
Computer Engineering. 2016, 42(4): 27-30,36. https://doi.org/10.3969/j.issn.1000-3428.2016.04.005Current Linux systems prevent abuse of administrator privileges by restricting root login.Ordinary users are allowed to be temporarily elevated to administrator privileges when there is a need to perform privileged operations.However,the existing user privilege elevation mechanism only supports elevation to root,so the administrator decentralized operating system can not distinguish between elevations of different administrator privileges,so that the users can not handle the appropriate management operations.This paper designs an administrator decentralized user privilege elevation mechanism on the basis of customizing SELinux policy to decentralize administrator and authenticate users according to privilege requirements of different applications.Only authenticated users can elevate to the specified administrator privileges.The mechanism adopts DBus to isolate ordinary users and administrator and secures the user privilege elevation service with SELinux policy.Experimental result shows that ordinary users can effectively elevate their privileges to the specific administrator’s with the support of this mechanism.
-
Computer Engineering. 2016, 42(4): 31-36. https://doi.org/10.3969/j.issn.1000-3428.2016.04.006
In view of the low accuracy and limited parallel scale of the major Computational Fluid Dynamics(CFD) software,a new workbench software based on China CFD software CCFDV1.0 is developed in this paper.The software is based on the principle of Open Service Gateway Initiative(OSGI),uses the architecture of Micro-kernel+Plugins and provides batch processing,double-model control arguments,computing monitoring and other user-friendly functions.The parallel scale of the software is improved by integrating 7 core solving modules and an optimized design module.Experimental results show that the parallel workbench software has higher accuracy and larger parallel scale than the major CFD software.
-
Computer Engineering. 2016, 42(4): 37-43. https://doi.org/10.3969/j.issn.1000-3428.2016.04.007When the network coding technology is applied in Wireless Sensor Network(WSN),it can improve reliability and throughput in the network,also reduce energy consumption,balance network load,realize communication security and improve network security.Multipath routing has the characteristics of strong stability and high resource utilization.Based on network coding,this paper mainly discusses three kinds of multipath routing models of WSN including the normal multipath routing model,the non-intersecting multipath routing model based on network coding,and the intersecting multipath routing model based on network coding.It deduces formulas of the Successful Delivery Ratio(SDR) and the Normalized Redundancy(NR) on these models,and then simulates the three models and makes comparison.The simulation results show that the intersecting multipath routing model based on network coding has high reliability and balanced energy consumption and is more suitable for WSN.
-
Computer Engineering. 2016, 42(4): 44-49. https://doi.org/10.3969/j.issn.1000-3428.2016.04.008The Continuous Time Markov Chain(CTMC) is adopted to construct the cognitive radio system under multi-channel non-ideal spectrum sensing.Based on this,the relationship of the state transition rate,system detection rate and user arrival rate is formulated.The universal solution method of stationary distribution of the system states is given.Numerical simulation result indicates that the conflict probability and the channel utilization are proportional to the strength of second user arrival rate,and opposite to the number of channels and the false detection rate.In the realistic design,parameters can be set according to the conclusion in this paper as desired.
-
Computer Engineering. 2016, 42(4): 50-54. https://doi.org/10.3969/j.issn.1000-3428.2016.04.009The compressed sensing theory and its channel estimation algorithm are researched in this paper.Combined with the characteristics of 60 GHz system pilot structure,the heterogeneous pilot compressed sensing channel estimation algorithm is proposed in this paper.In this algorithm,the pilot sequence is multiplied by PN sequence,and the obtained pilot sequence matrix is inserted in the Orthogonal Frequency Division Multiplexing(OFDM) symbols,then the channel can be estimated by three symbols with Orthogonal Matching Pursuit(OMP) compressed sensing theory.Simulation results show that the channel estimation estimated accuracy and the system error rate of this algorithm are as the same as the traditional Golay sequence channel estimation algorithm.The sampling points number of the proposed algorithm is only 1/42 of the traditional channel estimation algorithm,which greatly improves the system bandwidth utilization.
-
Computer Engineering. 2016, 42(4): 55-59. https://doi.org/10.3969/j.issn.1000-3428.2016.04.010Because the ZigBee network and WiFi network work together in the 2.4 GHz band,it produces inter channel interference,which seriously affects the performance of ZigBee.In view of the above problems,a Carrier Sense Multiple Access(CSMA) control algorithm based on probability is proposed.The Packet Reception Rate(PRR) is calculated at the receiving node,the feedback packet scheme is adopted to transmit the packets to the sending node,the sending node determines the strength of the network interference according to the PRR,and decides to open or close the CSMA with pre-set probability function.Experimental results show that the proposed algorithm can effectively guarantee the network performance of ZigBee,regardless of whether there is WiFi interference.
-
Computer Engineering. 2016, 42(4): 60-64. https://doi.org/10.3969/j.issn.1000-3428.2016.04.011
It is difficult in selecting wavelet parameter when using the original wavelet ridge extraction algorithm in different signal environments.Therefore,an adaptive algorithm to extract wavelet ridge of Morlet wavelet transform spectrum is proposed.By analyzing wavelet ridge algorithm,this method takes the success of iteration as the feedback in order to adjust the shape parameter of Morlet wavelet,which makes the wavelet basis function change in real-time to adapt to a variety of situations.Simulation results show that wavelet-ridge of all kinds of signals can be accurately extracted with the same initial parameter settings.Meanwhile,the new algorithm has a higher probability of detection and noise suppression compared with the original wavelet ridge extraction algorithm.
-
Computer Engineering. 2016, 42(4): 65-69. https://doi.org/10.3969/j.issn.1000-3428.2016.04.012Based on the real-time task set scheduled by the Earliest Deadline First(EDF)algorithm,the bandwidth transfer caused by the request of inserting a new task or accelerating the current task is studied.The bandwidth transfer mode of a task pair is presented,that is,a curent running task frees part of its bandwidth for the insertion of a new task.A time point is proved from which on the insertion of the new task of the pair is smooth,and it means no deadline missing will be caused in the system.The theorem in the paper shows that this time point can be evaluated on a simple expression and is very applicable to the applications at run-time,if the period and the execution amount of each task in the pair are available and the remaining execution time of the one that frees bandwidth can be gotten in current period.Assuming the remaining bandwidth(the bandwidth of the system minus the one used by the task pair) must totally be reserved for other tasks during certain stages,then this time point is the earliest feasible time for the new task.The correctness of the theorem is verified by a simulation example.
-
Computer Engineering. 2016, 42(4): 70-74. https://doi.org/10.3969/j.issn.1000-3428.2016.04.013In view of the application scenarios that the non anchor nodes and sensor nodes can not communicate with each other,this paper develops a self-localization algorithm for Wireless Sensor Network(WSN).Signal source is introduced,so that the sensor node can receive the signal emitted by the signal source.According to the Time of Flight(ToF) of the signal from the signal source to the sensor node to obtain the observation distance between the two,the linearization of the observation equation is expressed as the form of linear matrix multiplication.The rough estimates of sensor and source locations are found using Singular Value Decomposition(SVD) after linearizing measurement equations and the Weighted Least Squares(WLS) estimation is applied to refine the estimates.The formulation of the WLS estimator takes the effect of TOF noises on SVD results into account.Simulations show that the new algorithm improves localization accuracy over the original method that uses Least Squares(LS) only.
-
Computer Engineering. 2016, 42(4): 75-78. https://doi.org/10.3969/j.issn.1000-3428.2016.04.014Interference cancellation method without feedback is proposed for Multi-input Multi-output(MIMO) interference channel with three users.Space-time code with Alamouti structure is designed ofr each user.The multi-user interference at the third receiver is cancelled by introducing zero matrices into each code,and the multi-user interference at the other two ceceivers is eliminated by the orthogonal property of Alamouti code.Compared with existing schemes for the same scheme,the proposed scheme requires no feedback information and maintains the same transmission efficiency.
-
Computer Engineering. 2016, 42(4): 79-82,87. https://doi.org/10.3969/j.issn.1000-3428.2016.04.015It is difficult for the existing congestion control strategies in Content Centric Network(CCN) to obtain the feedback of the network congestion status directly.Besides,it can not guarantee the Quality of Services(QoS) for multi-service of each business.To solve this problem,a congestion control strategy that can improve the QoS in multi-service situations is proposed.This strategy monitors the arrived interest packets by the intermediate routing nodes,judges the congesting degree according to the business conditions of the interest packet queue,and feeds back the link state information to receivers,so as to adjust the sending rate.Experimental results show that this strategy can effectively control congestion in the network without increasing the completion time of services,and ensure QoS of different busiress.
-
Computer Engineering. 2016, 42(4): 83-87. https://doi.org/10.3969/j.issn.1000-3428.2016.04.016
This paper presents a fast routing algorithm for the new double-layer Network on Chip(NoC) to make full use of the NoC resources.According to the frequency and total volume of the inter-core communications,the proposed algorithm firstly divides them into four types.Considering that circuit network is much more suitable for communications of large volume,the proposed algorithm assigns these communications to circuit network layer,and then uses the fast routing method to configure the paths for remaining communications in packet network layer.Experimental result shows that the proposed algorithm uses 74% less running time and results in 21% more resource usage of circuit network than simulated annealing algorithm.In addition,this algorithm can improve application performance by more than 16%.
-
Computer Engineering. 2016, 42(4): 88-93,100. https://doi.org/10.3969/j.issn.1000-3428.2016.04.017With the rapid growth of network usage,mobile communication network is operating under restrictive constraints of resources,seriously threatening the stable operation of network.High-frequency users,who activate/delete data session frequently,consume much more signaling resource and phone power.This paper performs the real-network investigation of high-frequency users from the aspects of resource consumption,activation rule,behavior pattern,etc.,and observes that they are extremely low resource utilization ratio according to low data traffic volume and high signaling overheads.Also,they tend to activate periodic sessions on session activation interval.Besides,a large number of periodical high-frequency users(45.38%) behave abnormally,which greatly threatens the network.
-
Computer Engineering. 2016, 42(4): 94-100. https://doi.org/10.3969/j.issn.1000-3428.2016.04.018Aiming at the problem that the existing fixed spectrum allocation method of shortwave radio access network can not satisfy the new requirement of applying intelligent frequency hopping technology,the spectrum demand of shortwave radio access network based on intelligent frequency hopping is analyzed,and a spectrum allocation model based on graph coloring theory is established for the spectrum allocation issue of shortwave radio access network users.Besides,dynamic spectrum allocation algorithm applicable to intelligent frequency hopping shortwave radio access network is proposed.According to the simulation results,compared with fixed frequency communication of using fixed spectrum allocation method,intelligent frequency hopping of adopting this method can increase reach ability of subnet and network throughput,reduces the jamming rate,and effectively solves application problems of intelligent frequency hopping technology in shortwave radio access network,and meanwhile quickly,efficiently and accurately completes allocation of usable spectrums.
-
Computer Engineering. 2016, 42(4): 101-104,111. https://doi.org/10.3969/j.issn.1000-3428.2016.04.019The modulation recognition technology of communication signal can be used in signal confirmation,interference identification,electronic warfare combat and intersatellite link communication.Aiming at the problem of the low signal modulation recognition rate under low Signal-to-noise-Ratio(SNR),an automatic modulation recognition algorithm based on neural network is proposed.By simplifying the identification feature parameters,the sensitivity of the parameters to noise is reduced and an identification process based on decision theory is presented.The BP neural network algorithm is improved by realizing adaptive learning rate and by choosing the optimal number of hidden layer nodes.Thus,an automatic modulation recognition scheme is gived based on the combination of the neural network algorithm and the decision tree.Simulation results show that,when SNR is no less than 0 dB,the average recognition ratio of the proposed algorithm is above 98%.
-
Computer Engineering. 2016, 42(4): 105-111. https://doi.org/10.3969/j.issn.1000-3428.2016.04.020According to the demand of interference detection in Global Navigation Satellite System(GNSS),and the energy detection principle,starting from the signal model,GNSS interference detection algorithm of iteratively updating threshold is designed,and the detection probability and false alarm probability in frequency domain are derived.The detected spectrum is sorted by Quicksort,and the initial threshold is set.The spectral line less than the threshold is updated to the new threshold,continuing to detect the updated spectral lines more than the new threshold until the threshold is converged.And the detected result is output with Gausstan white noise,the simulation results of 6 kinds of jamming show that the designed algorithm converges quickly.The algorithm can output the frequency,bandwidth and power of the interference signal,and can distinguish continuous wave interference,pulse interference,narrowband interference and wideband interference.Compared with traditional methods,the designed algorithm has better detection performance.
-
Computer Engineering. 2016, 42(4): 112-118. https://doi.org/10.3969/j.issn.1000-3428.2016.04.021Aiming at PHP program,this paper proposes an Structured Query Language(SQL) vulnerability detection method based on the injection analysis technology.This method makes a detailed analysis on the injection in the aspects of data flow and program behavior,on the basis of the combination of dynamic and static analysis technique.A lexcial feature comparison-based decision algorithm is designed and implemented to detect SQL vulnerability through simulating data.Combining alias analysis technology,behavior model and decision algorithm,the SQL vulnerabilities detection system based on lexical feature comparison is designed and realized.Experimental result indicates that,compared with Pixy and RIPS system,this system has higher SQL vulnerability detection capability and lower time cost.
-
Computer Engineering. 2016, 42(4): 119-125. https://doi.org/10.3969/j.issn.1000-3428.2016.04.022For a Wireless Sensor Network(WSN) using directional antennas for information transmission,a SEIRS malware propogation model based on rotating the directional antennas is established.Differential dynamic system is used to study the behavior of malware propogation in WSN.The results show that when the basic reproductive number of malware,R0,is less than or equal to 1,malware in network eventually eliminates and the disease-free equilibrium of system is globally asymptotically stable.When R0 is greater than 1,malware in network exists continually,and the endemic equilibrium of system is globally asymptotically stable.Analysis on the thresholds of the maximum communication radius,direction angle and retational angular velocity verifies the correctness of the theory.
-
Computer Engineering. 2016, 42(4): 126-130. https://doi.org/10.3969/j.issn.1000-3428.2016.04.023Aiming at the energy consumption inequality of sensor nodes,and to improve data security protection intensity,this paper puts forward a data fusion privacy-preserving algorithm with balancing cluster node energy.Through clustering routing to balance each node energy,it prolongs the network life cycle,uses the homomorphism Hash function to protect data integrity,introduces cluster group key to reduce the probability of node data to be deciphered,and strengthens the protection of private data.Experimental results show that,compared with the existing protection data fusion algorithm based on vector secret key,the algorithm has higher security and longer network lifetime.
-
Computer Engineering. 2016, 42(4): 131-136,142. https://doi.org/10.3969/j.issn.1000-3428.2016.04.024Based on the characteristics of Ad Hoc networks,such as self-organization and high mobility,a more robust identity-based key management scheme for Ad Hoc network is proposed.In the proposed scheme,the mutual authentication at the key distribution moment between the users and server nodes is provided through the public cryptosystem based on bilinear pairing.Threshold cryptography is used to realize distributed key management and avoid single point of failure.And the auxiliary nodes can effectively solve the key renewal problem which is caused by network partition.The proposed scheme guarantees the security of key management system,and reduces communication overhead and storage cost.Simulation results show that the scheme can not only complete renewal of the key reliably,but also enhance robustness of the key management service effectively.
-
Computer Engineering. 2016, 42(4): 137-142. https://doi.org/10.3969/j.issn.1000-3428.2016.04.025The space Delay Tolerant Network(DTN) has characteristic such as high delay and frequent disconnectivity.Aiming at this problem,this paper designs a non-interactive key exchange protocol,and it is proven secure in the random oracle model.The proposed protocol enables two parties to agree on a symmetric shared key without requiring any interaction.The protocol eliminates the dependency on certificate used in Public Key Infrastructure(PKI) and avoids the key escrow problem in ID-based Public Key Cryptography(ID-PKC).Analysis result shows that the proposed protocol does not utilize the expensive pairing computation and achieves higher efficiency compared with other Non-interactive Key Exchange(NIKE) protocols, and thus proves to be suitable for the space environment with restricted computing and few power resource.
-
Computer Engineering. 2016, 42(4): 143-146,154. https://doi.org/10.3969/j.issn.1000-3428.2016.04.026In most existing threshold proxy re-signature schemes,threshold value is essentially immutable.However,depending on the significance of re-signature message,the value should be changed dynamically in many practical applications.Based on proxy re-signature scheme Suni,this paper designs a new threshold proxy re-signature scheme which is unidirectional and variable.The system sets the new value of threshold before each re-signature,each agent re-calculates originally received sub-key partially,then each proxy can generate corresponding re-signature sub-key and verification key non-interactively.In order to protect the trustee security and avoid risk,system provides the public key when generating re- signature key.Compared with other unidirectional and variable re-signature schemes,this scheme can provide better efficiency in computation cost and signature length,and it has high security.
-
Computer Engineering. 2016, 42(4): 147-154. https://doi.org/10.3969/j.issn.1000-3428.2016.04.027The development of mobile devices makes users remotely control their desktops with phones or tablets,but vulnerabilities in mobile operating systems put mobile remote control under great security risk.This paper introduces TrustRFB,a secure remote control system leveraging the security extension TrustZone of ARM processor.By encrypting the network connection and putting privilege related part of code into the secure world provided by TrustZone,TrustRFB can effectively prevent attackers from stealing users’ privacy with malware,compromised system or even malicious remote control client.A prototype of TrustRFB is implemented on Exynos4412 development board and the evaluation shows that TrustRFB can defend against many attack methods like Rootkit and repackage attack while introducing a bandwidth dropdown of 1.2% in normal use.
-
Computer Engineering. 2016, 42(4): 155-159. https://doi.org/10.3969/j.issn.1000-3428.2016.04.028Membership functions of traditional Fuzzy Support Vector Machine(FSVM) are mostly designed based on the distance between the samples and the class centers,which decrease the effect of support vectors.This paper combines two membership functions and presents a new membership function to solve this problem.The new membership function divides the samples into three parts:support vectors,non-support vectors and outliers.It assigns large membership values to the support vectors which are far away from their class center.Small membership values are assigned to non-support vectors and outliers which are also far away from their class center.Experimental results show that the FSVM with the proposed membership function is more effective in classification.
-
Computer Engineering. 2016, 42(4): 160-167. https://doi.org/10.3969/j.issn.1000-3428.2016.04.029To resolve the synergy conflict of supply chain production-marketing,this paper proposes a self-adaptive multi-Agent negotiation method.This method considers the equity concern behavior of Agents with the background of one manufacturer Agent negotiating with many retailer Agents.It adopts the inequity aversion model and uses the Radical Basis Function(RBF) neural network to optimize the Actor-Critic learning algorithm to predict and adjust the concession magnitude of Agents.The effectiveness and stability of the algorithm are verified through comparative experiments under different numbers of retailer Agents and different concern degree.The method can enhance the self-learning and adaptive ability of Agent,overcome the disadvantages of low learning efficiency and lack of equity concerns,shorten the negotiation time and improve the efficiency of resolving conflicts.
-
Computer Engineering. 2016, 42(4): 168-172. https://doi.org/10.3969/j.issn.1000-3428.2016.04.030In order to handle the local optimum problem of the existing multi-objective differential evolution algorithms,the jumping genes operation is introduced and a novel multi-objective differential evolution algorithm based on jumping genes is presented.Different from the existing algorithms,the proposed algorithm performs jumping genes operation after the classical crossover operation,in order to improve the population diversity.Numerical experimental results indicate that the proposed algorithm is capable of dealing with the local optimum problem and exhibits significantly better convergence performance than the existing algorithms on ZDT and DTLZ function tests.
-
Computer Engineering. 2016, 42(4): 173-178. https://doi.org/10.3969/j.issn.1000-3428.2016.04.031In community mining process of complex network,traditional modularity definition depends on the adjacent matrix of a network,and only takes the relationships between nodes into consideration,which cannot reflect the structure of the real network well.Aiming at this problem,this paper proposes a path-based modularity analysis method.It constructs the modularity matrix based on paths between any two nodes and their probability distributions,substitutes the edge-based modularity matrix,i.e.adjacent matrix,and gets the modularity value that reflects the network structure.Using the path-based modularity analysis method for mining the communities of a complex network,this paper predicts labels for unlabeled nodes according to the community structure.Experimental results on Flickr dataset show that,compared with ModMax,LableDiffusion and EdgeCluster methods,the proposed method has better prediction performance when predicting labels for unlabeled nodes,and it can reflect the community structure more accurately.
-
Computer Engineering. 2016, 42(4): 179-183. https://doi.org/10.3969/j.issn.1000-3428.2016.04.032In the research of quantifying neighboring area’s dialect difference,this paper makes people read the independent word text A in dialect to form the sound file M,uses the HTK tool to structure the acoustic feature parameter set S_M for the file M,calculates and forms the diversity factor.Using this method in continuous i neighboring regions forms the homologous parameter set Si_Mi,while using the sound file Mi to compare to the sound-character (word) mapping table of the sample area (i=0),obtaining the text Ai of the village i.The ratio of text content differences between Ai and A0 (sample area or village) is defined as diversity factor ξ.By analyzing ξ feature of continuous villages,the paper finds that the dialect has less difference when the ξ value is between 0.88 and 1 in neighboring 3 villages(geographic location),while in 9 villages distance,the ξ value(synthesize) less than 0.6,and the ξ value of phrases less than 0.2,this difference changes quickly,so this paper establishes the dialect distance and proposes the concept of dialect radius,confirming that the dialect radius is eight(eight villages).
-
Computer Engineering. 2016, 42(4): 184-189,196. https://doi.org/10.3969/j.issn.1000-3428.2016.04.033Aiming at the limitation to effectively realize the view fusion in the multi-view clustering task for Maximum Entropy Clustering(MEC),this paper proposes an improved view-weighting MEC algorithm by introducing partition fusion and view-weighting.This method assigns a weight of each view to show the importance of each view.And it sets the partition matrix of each view,and the view-fusion in each view partition is made by a view-fusion weighting matrix.Finally,it proposes a new integration strategy to obtain the global partition result.Experimental results on synthetic datasets and UCI datasets show that the proposed algorithm outperforms MEC algorithm and CombKM algorithm in dealing with multi-view clustering task.
-
Computer Engineering. 2016, 42(4): 190-196. https://doi.org/10.3969/j.issn.1000-3428.2016.04.034Traditional communication network alarm correlation rules are often manually done by experts and coded into network fault management systems.However,the artificial maintenance method is difficult to meet the huge amounts of data processing requirements of real-time communication alarm rules.To solve this problem,this paper proposes an automatic alarm rule mining method based on Weighted Frequent Pattern-tree(WFP-tree) algorithm.It uses the sliding window method to convert raw data into alarm transactions,and employs BP neural network,Support Vector Machine（SVM） and Analytic Hierarchy Process(AHP) methods to generate the weight information of alarm equipment.Finally,it uses WFP-tree algorithm to automatically generate the weighted frequent itemset.The experimental results show that,the WFP-tree algorithm performs better in frequent itemset compression and important domain correlation rule finding compared with Apriori and FP-growth algorithms.
-
Computer Engineering. 2016, 42(4): 197-201,208. https://doi.org/10.3969/j.issn.1000-3428.2016.04.035A microblog recommendation method based on label correlation relationship is presented via analyzing mircoblog features and the deficiencies of existing microblog recommendation finding algorithms.Label retrieval strategy is adopted to add label for unlabeled users and users with small number of labels,and the user-label matrix is built and user-label weights are obtained.The label information is used to represent user interests.In order to solve the problem of high-dimension sparsity of the matrix,the correlation relationship between the labels is investigated to update the user-label matrix so as to get the final user interests.Experimental results show that compared with unlabeled microblog recommendation algorithm,the presented recommendation method is more effective in microblog information recommendation.
-
Computer Engineering. 2016, 42(4): 202-208. https://doi.org/10.3969/j.issn.1000-3428.2016.04.036Aiming at the problem that the test data do not belong to the training data in pattern classification task,under the sparse representation-based classification framework,this paper presents a pattern recognition with reject options model based on Fisher discriminant dictionary learning.This model adds the Fisher discriminant constraint to the objective function of dictionary learning,so that the sparse decomposition coefficients under the learned dictionary have a larger scatter between classes and a smaller scatter within one class,and builds several local linear manifold subspace for the coefficients of the training data to make an approximation of the nonlinear manifold space which the coefficients belong to.It makes a hypersphere cover to this built subspace to describe the distribution of the coefficients.For the test data,the model can make the decision accept the sample or reject it according to the hypersphere cover border.Experimental results on MINST show that the method has high correct recognition rate and can process with reject recognition to non training samples.
-
Computer Engineering. 2016, 42(4): 209-214,220. https://doi.org/10.3969/j.issn.1000-3428.2016.04.037With the development of high throughput technology,the size of Protein-protein Interaction(PPI) network grows larger and larger,thus high speed PPI network global alignment algorithm is needed.Aiming at this problem,this paper analyzes the SPINAL algorithm.It modifies the coarse-grained phase which occupies more than 95% time cost to adapt for distributed platform named Blogel,and keeps the fine-grained phase unchanged.Hence,the Blogel-SPINAL algorithm is gotten.Theoretical analysis and experimental results show that Blogel-SPINAL is much faster than SPINAL,and its scalability is better.
-
Computer Engineering. 2016, 42(4): 215-220. https://doi.org/10.3969/j.issn.1000-3428.2016.04.038Aiming at the problems of high cost and low efficiency in existing semantic Web service composition method based on graph search,this paper proposes an automatic composition optimization method of semantic Web service based on user request input closure and Complete Backward Tree(CBT).This method uses Request Input Closure Constructor(RICC) algorithm to construct user request input closure and determines whether the user requests can be met in the semantic Web service rule repository based on user request input closure in Closure_Complete Backward Tree_Optimal Valid Generation Path(C_CBT_OVGP) algorithm.When user request is not satisfied,it illustrates the request without a solution and then ends the algorithm.Otherwise,it prunes the backwark tree by user request input closure and setting object sets of new nodes so as to avoid unnecessary repetition to build the same node on a number of different branches.Example analysis result shows that the optimization algorithm improves the efficiency of semantic Web service composition,especially for dissatisfied user request.
-
Computer Engineering. 2016, 42(4): 221-225,234. https://doi.org/10.3969/j.issn.1000-3428.2016.04.039By fusing Kernel Principal Component Analysis(KPCA) and Minimum Distance Projection(MDP),a new method based on the original minimum distance differential projection is developed to address the face recognition problem.Different from the classical minimum-distance discriminant projection algorithm,the new algorithm is based on dimensionality reduction algorithm.By using KPCA to reduce the dimension of these samples,they are mapped to a high-dimensional nonlinear space.The corresponding projection matrix is calculated in the dimension-reduced nuclear space by differential projection method.The nearest neighbor classification method is adopted to classify these samples and the face recognition is succeeded eventually.Experimental results on the ORL,FERET,and YALE face databases show that the proposed algorithm can outperform other algorithms.It can be avoided by using the nuclear technique to extract the nonlinear feature of the face image,and on the same time,the neighbor relationship of the samples is also kept by the defined similarity weight of the nuclear space.
-
Computer Engineering. 2016, 42(4): 226-234. https://doi.org/10.3969/j.issn.1000-3428.2016.04.040In order to study how relationship strength affects the adoption of viewpoint in aggregation groups,this paper designs a new method which is based on the relationship strength theory.It uses the single source of infection with Susceptible-Infected-Recovered (SIR) model,assuming the infection rate is λ=βα,to analyse how relationship strength affects the adoption of viewpoint.Experimental results show that strengthening the authority of viewpoint and friend rewmmendation strategies are conducive to enhancing the probability and velocity of the adoption of viewpoint.It has a certain practical value to enhance advertising effectiveness on aggregation groups,accelerate information diffusion and inhibit the spreading of rumors.
-
Computer Engineering. 2016, 42(4): 235-241. https://doi.org/10.3969/j.issn.1000-3428.2016.04.041In estimating the perimeter of a target,the blurred and complicated target boundaries seriously influence the accuracy of perimeter estimation,and contribute to a long computing time.In order to solve this problem,combined with granular computing,an adaptive algorithm to calculate different boundary perimeter is proposed.A probability model based on image grains is designed to define the boundary thickness of the target object.The image is preprocessed by blocking and granulating based on the optimal size of image grain that is derived from the boundary thickness,which results in dimension reduction of the image data.And the classical algorithms are applied to the preprocessed image to estimate the perimeter of the boundary of the target object.Experimental results show that,compared with traditional algorithms like Digital Straight Segment(DSS),Minimum Length Polygon(MLP) and Gray-Level(GL),the proposed method can adaptively deal with different boundary thickness of the target objects,which uses less time in the case of maintaining the accuracy of perimeter estimation,and has better accuracy and adaptability,especially for images with blurred and complicated boundaries.
-
Computer Engineering. 2016, 42(4): 248-254. https://doi.org/10.3969/j.issn.1000-3428.2016.04.043In order to increase iterative efficiency and image segmentation quality,an improved Kernel Mapping(KM) model is presented.Both definition and method of image data discretization are proposed based on the rough set theory.The image after discretization will enhance intensity homogeneity and every region parameter sequence deduced by the weighted KM model converges,so that the region parameters can better express the gray value among the evolving regions.Therefore,the image can be more accurately segmented.During each iteration,the minimum of the weighted value of the level set elements is one according to the weighted KM model.The larger value for the weight becomes,the faster the level set element is updated and the evolving curve converges.The level set of KM model without weighted can be taken as the level set which the weighted value of every element equals one.Thus the weighted KM model can fulfill an image segmentation with high efficiency.Experimental results on synthetic and natural images show that the proposed method has better segmentation quality and iteration efficiency.
-
Computer Engineering. 2016, 42(4): 255-258,265. https://doi.org/10.3969/j.issn.1000-3428.2016.04.044Cross-dataset image classification is a common problem in the real applications of image classification.Even though training data and testing data are related in the cross-domain classification,there are some differences between them.And this leads the performance of traditional classifier in cross-dataset classification dramatically reduced.In order to solve this problem,this paper proposes a novel cross-dataset image classification method.The new method introduces the idea of feature information reservation of Principal Component Analysis(PCA) into Logistic return based on L1 logistic regression,so that it can keep high information features in dataset when selecting image features.Experimental results show that in commonly used cross-dataset image classification,the method can obtain good image classification effect.
-
Computer Engineering. 2016, 42(4): 266-275,281. https://doi.org/10.3969/j.issn.1000-3428.2016.04.046This paper proposes a dynamic social network model for competition among three languages with Agent based modeling method and social circle theory.The topological structure parameters of the constructed social network are closer to the parameters of actual social network.The individual Agent in the network can move,die and reproduce,so the network is endowed with dynamic characteristics.The model for competition between two languages of Agent nodes in language competition network is extended to the model for competition among three languages of Agent nodes.The Agent nodes in the network represent individuals having learning and forgetting abilities.Each Agent can acquire the second or third language through learning to become bilingual or trilingual Agent,and they can become monolingual or bilingual Agent by forgetting one or two languages.Actual linguistic researches show the expansion and penetration of language has directionality in the communities with competition of more than two languages.Simulation analysis of the proposed model shows that the proportion of Agents with the lowest language state who have to learn both the strong and sub-strong language is more than the proportion of Agents with the highest language state or higher language state who have to learn the language of lowest state,the proportion of Agents with sub-strong language who have to use the strongest language is more than the proportion of Agents who have to use the language of lowest state,the proportion of Agents with the lowest language state who have to learn other two languages at the same time is the highest,so the living patterns with different proportions of three languages can affect the proportion of bilingual or trilingual Agents,but it can not change the direction of language evolution,which is consistent with actual language survey results.
-
Computer Engineering. 2016, 42(4): 276-281. https://doi.org/10.3969/j.issn.1000-3428.2016.04.047To differentiate data streaming services of different types and different requirements for Quality of Service(QoS) in intelligent Road(iRoad) systems, this paper proposes a resouce reservation model based on priority. The model classifies data packets of different QoS requirements into different priority queues, and the schedule algorithm decides data output sequence. When the node gets into channel successfully, it will restrain the transmission of other nodes by resouce reservation model, until the transmission of remaining data packets with the same priority is completed. Experimental results show that the proposd model effectively reduces the probability of collisions and improves channel utilization, and can realize service differentiation between data streams of different priorities.
-
Computer Engineering. 2016, 42(4): 282-287,294. https://doi.org/10.3969/j.issn.1000-3428.2016.04.048A single machine scheduling problem with sum of actual processing times dependent learning effect and unequal release time consideration is investigated where the objective is to minimize the total completion time.According to the NP-hard characteristic of the problem,two dominance rules are developed,comined with Tabu Search(TS) algorithm and priority rule,this paper proposes an algorithm of hybrid TS algorithm,which improves the ability of jumping out of the local optimal trap,both retains the excellent gene and expands the scope of the search area.Experimental results demonstrate that the proposed hybrid Tabu search algorithm has a better performance than the benchmark algorithms in the literature and the advantage becomes more obvious with the number of jobs increasing.
-
Computer Engineering. 2016, 42(4): 288-294. https://doi.org/10.3969/j.issn.1000-3428.2016.04.049The traditional Doppler fetal heart rate detector has the problems of low measurement accuracy,slow response,poor stability and high power consumption.Aiming at these problems,combining with smartphone,this paper proposes an improved fetal heart rate detection system.Using the smartphone audio interface,fetal heart signal detected by ultrasound Doppler detection terminal is passed to smartphone,and the improved fetal heart rate detection algorithm is running in it.At the same time,the hardware architecture of the Doppler detection terminal is simplified to reduce power consumption and cost.A signal detection module is added before the signal processing module,starting from the signal source,which effectively eliminates mutation or pure noise signal.Experimental results of detecting the fetal heart from a simulated signal verify the accuracy and usefulness of the proposed system.
-
Computer Engineering. 2016, 42(4): 295-300,306. https://doi.org/10.3969/j.issn.1000-3428.2016.04.050In order to solve the problem of low efficiency in the use of data sources and data transmission of the traditional data intensive scalable computing systems,this paper proposes a big data throughput optimization algorithm based on full C-order moment model with parallel flow prediction.To improve the prediction accuracy of parallel flow,it takes the utilization efficiency improvement of the bottleneck link for the purpose,and designs the method of equivalent parallel flow algorithm.According to the partial C-order moment model and full second-order moment model,it constructs the full C-order moment model,and designs the low sample throughput optimization framework,which can reduce the computational complexity.Experimental results on the data set with different size show that the full C-order moment model with parallel flow is more suitable for the transmission of big data,and more efficient.
-
Computer Engineering. 2016, 42(4): 301-306. https://doi.org/10.3969/j.issn.1000-3428.2016.04.051Cover query plays an important role in maintaining the integrity of files and the consistency of data in distributed systems.Although Bloom filter can support quick membership query,it can only represent a discrete data set.This paper utilizes a set of prefix to represent the range rule and proposes a prefix-code transformation function to transform each prefix into an unique binary code.To facilitate cover query,this paper further designs a BFrange system to store the two tuple containing rules and specific storage elements by combining the counting Bloom filter with a set of list.The query time complexity under BFrange is only O(1),which does not depend on the number of the rules.Simulation results demonstrate that the BFrange system can support efficient and accurate cover query.
-
Computer Engineering. 2016, 42(4): 307-312. https://doi.org/10.3969/j.issn.1000-3428.2016.04.052Considering the existing Advanced eXtensible Interface 4(AXI4) master interface can only work in half-duplex mode and can not apply for multiple transmissions simultaneously,a design scheme of AXI4 master conversion interface supporting pipeline transmission is proposed in this paper.By adding an address information storage module which allows the conversion interface to store the new apply transmission requests before current transmission ends,it gives out the address data on the bus until it is allowed.This scheme bases on separated state machines of reading and writing,it supports full-duplex operation and pipeline transmission.Simulation result shows that,when the host requests for two times of a 4-data continuous read and an 8-data continuous write transmission,compared with the interface supporting half-duplex mode,the interface designed by proposed scheme can save 16 clock cycles.The conversion interface can reduce the transmission cost effectively and ensure the accuracy of data transmission.
-
Computer Engineering. 2016, 42(4): 313-317. https://doi.org/10.3969/j.issn.1000-3428.2016.04.053To eliminate the noise mixed in the Electroencephalogram(EEG),the paper presents a kind of EEG de-noising method based on Ensemble Empirical Mode Decomposition(EEMD) and improved lifting wavelet.Firstly,the noise-added EEG signals are decomposed into several Intrinsic Mode Function (IMF) components by EEMD.Secondly,the high-frequency IMF components dominated by the noise component are extracted through the properties of auto correlation function method,and de-noised by improved lifting wavelet.Finally,the high-frequency IMF components processed and low-frequency IMF components are reconstructed to get the de-noised signal.Experimental results show that this method has better Signal-to-noise Ratio(SNR) and smaller Root Mean Square Error(RMSE) compared with the traditional method and the improved lifting wavelet de-noising method.
-
Computer Engineering. 2016, 42(4): 318-321. https://doi.org/10.3969/j.issn.1000-3428.2016.04.054The traditional high order feature representation methods are based on feature matching,which can cause combination explosion problem.In this case,this paper proposes a novel high-order feature representation method without feature matching.The image is divided into multiple grids.The scale Invariant Feature Transform(SIFT) features in each grid are combined into one high-order feature representation.The feature of maximum tf-idf value is regarded as the major feature of the grid,while the other as auxiliary features.The angle between,the major feature and the auxiliary features is used as the information of the feature space.High-order feature matching is based on vision matching and geometry matching,and Support Vector Machine(SVM) is used for image classification experiments.Experimental results show that the proposed method can improve the classiffication accuracy by 4%,and effectively avoid the combination explosion problem.
2024
Volume 50 No 4
Volume 50 No 4
(Started in 1975, Monthly)
Sponsored by：
East-China Institute of Computer Technology Shanghai Computer Society
Edited & Published by：
Editorial Board of Computer Engineering
Chief Editor：Jiang Bo
Executive Chief Editor:
Hu Bing
Tel：021-67092217
E-mail: ecice06@ecict.com.cn
CN: 31-1289/TP
ISSN: 1000-3428
Sponsored by：
East-China Institute of Computer Technology Shanghai Computer Society
Edited & Published by：
Editorial Board of Computer Engineering
Chief Editor：Jiang Bo
Executive Chief Editor:
Hu Bing
Tel：021-67092217
E-mail: ecice06@ecict.com.cn
CN: 31-1289/TP
ISSN: 1000-3428