Author Login Editor-in-Chief Peer Review Editor Work Office Work

15 July 2016, Volume 42 Issue 7
    

  • Select all
    |
  • CAO Guining,SHI Chengwei,CHEN Xiangxian
    Computer Engineering. 2016, 42(7): 1-15. https://doi.org/10.3969/j.issn.1000-3428.2016.07.001
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    In order to improve the reusability of existing mobile terminals for condition monitoring systems and reduce system development cost,this paper proposes a design scheme of universal mobile terminals for condition monitoring systems.The principle of the universal mobile terminal is introduced,and the interaction design of it is also explained.The universal realization of the mobile terminal is illustrated in detail,and the universality is verified by applying the mobile terminal to two different monitoring systems of tall buildings and green houses.The results show that the universal mobile terminal can automatically change display interface and display content according to the monitoring requirements of different condition monitoring systems and monitoring objects.It has strong improves the adaptability and universality.

  • SHI Longzhao,GUO Jimin
    Computer Engineering. 2016, 42(7): 16-21. https://doi.org/10.3969/j.issn.1000-3428.2016.07.002
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    This paper proposes a low power consumption FFT/IFFT processor for IEEE 802.11ac Multiple Input Multiple Output-Orthogonal Frequency Division Multiplexing(MIMO-OFDM) system. The processor can support the parallel processing of eight spatial data groups, and each group has 128 points. The RAM data reorder architecture is used to realize eight input and output spatial data groups in natural order, thus the processor can be used directly to communicate with other modules in the MIMO-OFDM system. The proposed FFT/IFFT processor is designed in Verilog and synthesized using TSMC 65 nm process library. The operating frequency is up to 100 MHz under 0.9 V and 125 ℃, the worst operating condition of the process library. Compared with the FFT/IFFT processor which uses register delay elements to adjust the order of input and output data, the proposed FFT/IFFT processor can realize 32.98% reduction in the gate count and 79.4% reduction in power consumption.
  • LI Yage,LAN Chengdong
    Computer Engineering. 2016, 42(7): 22-26. https://doi.org/10.3969/j.issn.1000-3428.2016.07.003
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Through analysis of business demands and performance of traditional mobile office systems,this paper proposes a mobile office system based on iPhone Operating System(iOS),to reduce mobile office system’s memory consumption and improve it’s operational performance.It uses Model-View-Controller(MVC),Delagate and Cell reuse mechanism to reduce the coupling between modules and system memory consumption,then improves the operational performance of the conventional mobile office system.Meanwhile,it uses the system’s background service architecture based on J2EE framework to provide mobile terminal with relevant data.Test results show that the operating performance of the proposed system is improved,and 19.6% system memory consumption is reduced,compared with traditional systems which do not use Cell reuse mechanism.
  • SUN Hao,XUE Xiao
    Computer Engineering. 2016, 42(7): 27-32,41. https://doi.org/10.3969/j.issn.1000-3428.2016.07.004
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Based on the characteristics of e-commerce ecosystem,this paper applies multi-Agent modeling technology and Java language to construct the model of e-commerce ecosystem on the RePast platform,and uses the method of computational experiment to conduct simulation experiment.Simulation results reflect that the system model can show the evolution process and the relationship between micro factors and macro evolution.By modifying the operating parameters,the model can simulate different system environment,and thus provide the basis for the strategy adjustment of relevant enterprise and the controllable evolution of the system.Therefore,applying multi-Agent modeling and RePast platform to construct system model and experimenting by the method of computational experiment are suitable means for the research on the evolution of e-commerce ecosystem.
  • WANG Lei,JIANG Jiulei,WANG Xiaofeng
    Computer Engineering. 2016, 42(7): 33-36,48. https://doi.org/10.3969/j.issn.1000-3428.2016.07.005
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    It is very meaningful for software development to identify design patterns automatically from Unified Modeling Language(UML) models.Formalization is the base of automatic identification of design patterns,so a method based on Petri net is proposed to describe design patterns.Conversion rules of design patters to Petri net are defined,and observer pattern is formalized by those rules.Preliminary process of automatic identification of design patterns is given.Analysis results show that this method can describe design patterns graphically.On this basis,automatic identification of design patterns can be achieved by means of the mathematical theory of Petri net.

  • MA Yunfei,YUAN Le,ZHU Feifei
    Computer Engineering. 2016, 42(7): 37-41. https://doi.org/10.3969/j.issn.1000-3428.2016.07.006
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The design and operation of Unmanned Aerial Vehicle(UAV)communication network are greatly influenced by man-made factors and have poor autonomy.To solve this problem,this paper proposes a cognitive network oriented design architecture for UAV system.Imitating from the vertebrates’ nervous system and the human social organization mode,this paper discusses the design of functional and organizational architectures of UAV system based on cognitive network.It designs a cross-layer interactive node-level architecture for UAV communication system,establishes a hierarchical cognitive behavior model based on Observe-Orient-Decide-Act(OODA) loop,and obtains a cognitive layered multi-routing method for UAV systems.Simulation results demonstrate that the proposed method can decrease the variation of system throughput and the fluctuation time.It improves the reliability of the system and the degree of intelligent decision.
  • JIAO Wenzhe,ZHAI Zhengjun,WANG Guoqing
    Computer Engineering. 2016, 42(7): 42-48. https://doi.org/10.3969/j.issn.1000-3428.2016.07.007
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    According to the characteristics of Time-triggered Avionics Full Duplex Switched Ethernet(TTAFDX),this paper proposes a scheduling strategy based on frame-length priority to solve the scheduling problem of Time-triggered Virtual Link(TTVL).It builds a typical network,and compares the instantaneities of TTAFDX and AFDX which uses Static Priority(SP) scheduling strategy.The result shows that,the delay of TTVL consists mainly of fixed delay,and the instantaneity of Rate Constraint Virtual Link(RCVL) is also improved compared with the low-priority VL in AFDX,which demonstrates that the scheduling strategy of TTAFDX can enhance the time-deterministic feature of the network while being compatible with AFDX.It can be implemented in the avionics system with hard real-time transmission requirement.
  • LI Bingqian,WANG Yong,Lü Ying,LIU Da
    Computer Engineering. 2016, 42(7): 49-53,58. https://doi.org/10.3969/j.issn.1000-3428.2016.07.008
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Specially designed for avionics systems,a flow and error control mechanism over fibre channel driven by credit update is proposed.By optimizing the frame header format,the receiving end obtains the credit value through halving and drives the sending end to check and update the credit counting.Full duplex communication realizes credit count parameter management for data sender,data frame transmission error control,and triggered credit update for the receiver,which avoids resource locking and timeout response.Experimental results show that the proposed mechanism has better performance than the mechanism with relevant credit update.It can make full use of the memory resources of airborne platform.In case of limited cache for data frame receiving,the mechanism can ensure the throughput rate of big data flow message sequence transmission and improve real-time performance,which satisfies reliable and efficient data transmission in avionics networks.
  • ZHU Linglei,YAO Yuancheng,QIN Mingwei
    Computer Engineering. 2016, 42(7): 54-58. https://doi.org/10.3969/j.issn.1000-3428.2016.07.009
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Currently widely used fractal Gaussian noise method is less effective in approximating Gaussian power spectrum and with low accuracy in generating self-similar traffic.To solve this problem,this paper proposes a high-order linear fitting method.Dividing the power spectrum of fractal Gaussian noise into two parts,which are finite and infinite accumulation terms,it uses the fitting toolbox cftool of Matlab to fit the infinite terms of power spectrum,reduces the amount of calculation,overcomes the complicated problems in mathematical deduction process,and makes up the error caused by using continuous integration instead of discrete summation in the process of mathematical derivation.Simulation results show that the relative error of the improved method is lower than that of the formula derivation method in generating self-similar traffic.The relative error is reduced to 0.05% by using wavelet method,and the computing speed of the improved method is fast.It takes 0.634 seconds to generate the self-similar sequence with the length of 220.
  • GAO Chunling,LI Zi,SUN Zeyu,XING Xiaofei
    Computer Engineering. 2016, 42(7): 59-64,71. https://doi.org/10.3969/j.issn.1000-3428.2016.07.010
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the process of multi-target coverage,the coverage for monitoring area becomes less effective and quite a lot of energy has to be consumed due to large amounts of redundant data.Based on this problem,an agreement of multi-target coverage preservation algorithm is proposed.The solution to expectation value for monitoring area coverage is given by establishing the subordination between sensor nodes and target nodes through the network model.With regard to the network energy transfer,the scheduling mechanism of sensor nodes is applied.When the network energy achieves balance,different coverage is realized by means of energy transfer of different nodes.Simulation results show that compared with optimized coverage algorithm based on event probability model(EPDM),multi-target coverage algorithm based on linear programming(ETCA),and other algorithms,this algorithm can get higher average network lifetime value and coverage rate.
  • JIANG Cheng
    Computer Engineering. 2016, 42(7): 65-71. https://doi.org/10.3969/j.issn.1000-3428.2016.07.011
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    IP networks are designed over optical backbone networks,so the failure of optical components often leads to multi-link failure in the IP network.The traditional failure recovery methods have higher computational complexity and are not practical.Aiming at the above problems,this paper analyzes the IP fast rerouting network model based on the Arc-disjoint Spanning Tree(ADST),uses packet forwarding algorithm to compare the average path length,and proposes an improved multi-link failure recovery method.Simulation results show that,for the k-edge connected network,the proposed method can achieve failure recovery as long as the number of concurrent link failures are less than k-1.The proposed method can reduce the path stretch and average path length while recovering multiple failures.
  • ZHAN Yiwang,HU Binjie
    Computer Engineering. 2016, 42(7): 72-76. https://doi.org/10.3969/j.issn.1000-3428.2016.07.012
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem that the mobile user travel pattern recognition is too complicated,an arbitrary shape clustering algorithm based on density and dynamic threshold is proposed.The data preprocessing technology is used to extract the speed of mobile users.User speed classification model is constructed through the MCFT tree.DBSCAN method is used to merge the core subclusters.The clustering results of the arbitrary shape of the MCFT tree are generated,so as to achieve the effective use of dynamic speed threshold to identify the user travel pattern.Experimental results show that it is feasible to identify the user’s travel pattern by using the DVTD method to get the dynamic speed threshold,which can effectively improve the accuracy of user travel identification.
  • LIU Xun,YANG Enzhong,ZENG Xiangchi,YANG Jian
    Computer Engineering. 2016, 42(7): 77-81. https://doi.org/10.3969/j.issn.1000-3428.2016.07.013
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Video transmission will be greatly affected if congestion link exists on the transmission path in single-path communication.This paper designs a multi-path transmission system of scalable video to solve the problem in OpenFlow network.The system consists of OpenFlow switch,media server and controller.Media server sends scalable video data.Controller acquires the path of each video streaming according to Edmonds-Karp algorithm,and generates and sends flow tables to OpenFlow switch.Then data forwarding and multi-path transmission are completed.Experimental results indicate that multi-path transmission can balance the loads of all network links and effectively reduce the packet loss rate of scalable video transmission compared with single-path transmission.
  • LIU Kan,ZHANG Wei,ZHANG Weidong,ZHANG Youmei,GU Jianjun
    Computer Engineering. 2016, 42(7): 82-85. https://doi.org/10.3969/j.issn.1000-3428.2016.07.014
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To deal with the variant and unpredictable wireless signals,this paper proposes a regressive wireless positioning method based on Deep Neural Network(DNN).It uses a four-layer Deep Neural Network(DNN) structure pre-trained by Stacked Denoising Autoencoder(SDA).This network structure is capable of learning reliable features from a large number of noisy samples,avoiding hand-engineering.The data required for the experiments is collected from the real world in different periods.Experimental results indicate that the proposed method can improve positioning accuracy in coping with the turbulent wireless signals.
  • XIE Ruiyun,HAI Benzhai
    Computer Engineering. 2016, 42(7): 86-93. https://doi.org/10.3969/j.issn.1000-3428.2016.07.015
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    Aiming at the problem that traditional Wireless Sensor Network(WSN) data compression algorithms cannot take both compression efficiency and data loss into account,a fast and efficient Lossless Adaptive Compression(LAC) algorithm based on adaptive Huffman coding and Golomb-Rice coding is proposed.Hybrid coding of adaptive Huffman and Golomb-Rice is used to solve the problem of variable length and dynamic.Heuristic method is used to simply estimate non-negative Golomb-Rice coding parameters proposed.A rice mapping function is used to transform the Laplace distribution error term so as to approximate the geometric distribution of nonnegative integers,which are used as the input of entropy encoder.Adaptive entropy coding is used to independently compress sampling data block.Experimental results on real environment WSN dataset from SensorScope show that the proposed algorithm acnieves a compression ratio of 4.11 per sample,and can realize power savings of up to 70.61%.Besides,compression performance and compression rate of the proposed algorithm are better than that of S-LZW,LEC and other compression algorithms.

  • CAO Shihua,WANG Qihui,WANG Lidong
    Computer Engineering. 2016, 42(7): 94-99. https://doi.org/10.3969/j.issn.1000-3428.2016.07.016
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to solve the problem of multi-dimensional dynamic elastic ranging in Wireless Sensor Network(WSN),this paper presents a node localization algorithm based on iterative mechanism of neighborhood rotation and hopping.It realizes node neighborhood clustering in WSN.The network neighborhood is composed of the first-order and the second-order neighborhood nodes.Network nodes are located according to the repulsion and attraction of neighborhood radius.The minimization of error is achieved by iterative mechanism.Finally,the exact coordinates of the nodes are obtained.Simulation results show that compared with node localization algorithm based on Result-sound Level Meter(RSLM) mechanism and Second-order Cone Program(SOCP) network node localization algorithm,the proposed algorithm can effectively reduce the node position error and network control overhead of multi-dimensional ranging,and improve node localizetion accuracy and network packet delivery ratio.
  • WANYAN Xuzhe,ZHANG Yangyang,CHEN Shuangwu,YANG Jian
    Computer Engineering. 2016, 42(7): 100-103. https://doi.org/10.3969/j.issn.1000-3428.2016.07.017
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of adaptive video transmission in OpenFlow network,an adaptive scalable video transmission algorithm is proposed.By leveraging OpenFlow,the control plane is detached from a router.The network information can be collected by the OpenFlow controller to instruct the dynamical adjustment of video bit-rate.The dynamics of instantaneous bandwidth on each link along the server-client path are recorded by the OpenFlow controller.According to the network status some time ago,the autoregressive model is applied to predict the network changes in the forthcoming period of time.The OpenFlow controller makes the rate adaptation decision and informs the client to adjust its requested video layer.Experimental results show that the algorithm is capable of predicting network changes and improving the perceived video quality,while keeping a low playback interruption.
  • LIU Ke
    Computer Engineering. 2016, 42(7): 104-108,116. https://doi.org/10.3969/j.issn.1000-3428.2016.07.018
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In a scenario of recovery from massive failures,a network is repaired through multiple restoration stages because availability of repair resources is limited.During repair transition,it is crucial to ensure the accessibility of important flow for network operators.Therefore,this paper discusses the problem of maximizing the flow recovery ratio involving transient repairing stages and minimizing the number of switched paths.It formulates the problem as Linear Programming(LP),proposes a heuristic algorithm based on the dividing and impaling idea for networks consisting of a few hundred nodes,and obtains pareto-optimal solutions of the flow recovery ratio and the number of switched paths.Simulation results show that the algorithm can produce sub-optimal solutions within a 4% difference from optimal solutions,produced by LP,and it reduces the number of affected routing entries by about 60% compared with the current method while ensuring equivalent flow ratio.
  • CHEN Yu,WANG Yadi,WANG Jindong,WANG Na
    Computer Engineering. 2016, 42(7): 109-116. https://doi.org/10.3969/j.issn.1000-3428.2016.07.019
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    Since various types of information security risk assessment standards are too complex,as an alternative,most enterprises can only choose to do construction in accordance with safety standards in implementation.This always leads to a substantial gap that security measures are not for particular systems and cannot be quickly adjusted according to the changes of the system.To deal with these problems,this paper proposes a risk assessment method with low difficulty of implementation.In this method,Fuzzy Cognitive Map(FCM) is used to capture dependencies between assets FCM reasoning process is used to calculate the value of systemic risks.An application of the method is studied using an example of a mobile office system.Results indicate that the proposed method is efficient and low-cost.It can reflect the risk status of the system promptly and appropriately.

  • HUANG Bin,LIU Guangzhong,XU Ming
    Computer Engineering. 2016, 42(7): 117-122,128. https://doi.org/10.3969/j.issn.1000-3428.2016.07.020
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    This paper proposes a two-way protocol to authenticate the security of nodes in the wireless sensor network based on clusters.The protocol is composed of lightweight verification based on solution of equation and digital signature verification based on bilinear mapping.The former is based on that it is difficult to solve a univariate equation of higher degree,and the latter is based on properties of bilinear mapping.In this protocol,nodes in the wireless sensor network are pre-constructed by the lightweight verification as a system structure in cluster units.Digital signatures are verified during the exchange of information between the nodes,so as to constitute a secure wireless sensor network system structure.In the process of generating digital signatures,both nodes achieve the exchange of symmetric key to encrypt information exchange.Security analysis results show that the nodes in the wireless sensor network can not only quickly join the cluster but also safely communicate with each other.

  • HUANG Baohua,WANG Tianjing,JIA Fengwei
    Computer Engineering. 2016, 42(7): 123-128. https://doi.org/10.3969/j.issn.1000-3428.2016.07.021
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the query efficiency of encrypted numeric data,this paper proposes a new method to build index on the numeric data in database.The monotonically increasing function is used to compute the comparison value from sensitive data,and the non-monotonic function is used to compute the confusion value from sensitive data.Then the comparison value and confusion value are connected together to form the index value to store in database.A two-phase method is used to query the encrypted data.In the first phase,the comparison value is computed according to the query condition.Then it is compared with the index value whose confusion part is removed to select out the encrypted result set that matches the query condition.In the second phase,the encrypted result set of the first phase is decrypted and queried to gain the final result set.Experimental results show that the proposed method can greatly improve the query efficiency of encrypted numeric data in database.
  • RONG Xuefang,WU Zhen,WANG Min,DU Zhibo,RAO Jintao
    Computer Engineering. 2016, 42(7): 129-133. https://doi.org/10.3969/j.issn.1000-3428.2016.07.022
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the existing researches,a fault is often required to be injected into a byte of a fixed position,and then special wrong ciphertext is obtained to carry out attacks.This assumption is difficult to achieve in the actual attack,so a differential fault attack method on SM4 based on random fault injection is proposed.Random faults are injected into the last four rounds of the SM4 encryption algorithm,and wrong ciphertext is generated by these faults.The sub keys of the last four rounds are got by using the candidate key screening method.The attack is made to an unprotected SM4 smart card.Compared with other attack methods,this method expands the scope of fault injection and improves the practicability of the fault attack.
  • WEI Lihao,AI Jieqing,LIU Shenghan
    Computer Engineering. 2016, 42(7): 134-138. https://doi.org/10.3969/j.issn.1000-3428.2016.07.023
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Recently,Identity-based Encryption(IBE) scheme based on lattices becomes one of the focuses of cryptographic research.But the IBE scheme based on standard lattice have large key size and high ciphertext expansion rate.So based on ideal lattices,this paper presents an efficient identity-based encryption scheme with small key size and low ciphertext expansion rate.The proposed scheme uses a method that combines the NTRU digital signature and the dual encryption based on ideal lattices.It is provably secure in the random oracle model.Analysis results show that the public key and the private key of key generation center contain one ring element and four ring elements respectively.The decryption key of each user is comprised of two ring elements.The ciphertext contains only two ring elements and the ciphertext expansion rate is a small constant.The encryption and decryption require four and two multiplications in the polynomial ring.Hence,the IBE scheme is more computationally efficient than that of traditional number theory based IBE scheme.
  • QI Yingchun,ZHU Hai
    Computer Engineering. 2016, 42(7): 139-145. https://doi.org/10.3969/j.issn.1000-3428.2016.07.024
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the Wireless Sensor Network(WSN),the Agent migration prococol based on the static path has high energy comsumption,and there are many malicious nodes.To solve these problems,a Directional Virtual Backbone-Mobile Agent Migration(DVB-MAM) protocol is proposed based on energy model and directional virtual backbone network.The protocol uses the static path to calculate Agent migration and determines the next node of Agent migration based on energy model and directional virtual backbone.In addition,cloning features are used to reduce the route length of the Agent.In the process of data aggregation,a trust evaluation framework is used to detect malicious nodes.And malicious nodes can be found at an early stage.Experimental results show that the energy consumption of the proposed protocol is decreased by about 26% compared with that of Tree Based Itineraries Design(TBID).When malicious nodes exist,compared with other existing protocols,the highest average delivery rate of the proposed protocol can be increased by 26%,which can effectively prevent node failures and the attacks of malicious nodes.
  • FEI Teng,ZHANG Liyi,CHEN Lei
    Computer Engineering. 2016, 42(7): 146-152,158. https://doi.org/10.3969/j.issn.1000-3428.2016.07.025
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In the late stage of basic artificial fish swarm algorithm, artificial fish has great potential to be aggregated in a non global optimum, causing the algorithm to fall into local optimum, thus the artificial fish swarm algorithm is seriously affected in search accuracy, convergence speed and stability. In order to overcome this shortcoming, an improved adaptive Levy distribution mixed mutation artificial fish swarm algorithm is proposed. By the characteristics of Levy distribution and chaotic mutation, the diversity of artificial fish in basic artificial fish swam algorithm is increased, which improves the ability of the basic artificial fish swarm algorithm to jump out of local optimum, so as to improve the basic fish swarm algorithm. Simulation results of the typical test function show the improved artificial fish swarm algorithm has better optimization ability than the particle swarm optimization algorithm and artificial fish swarm algorithm.
  • ZHANG Fei,HUANG Guoxing
    Computer Engineering. 2016, 42(7): 153-158. https://doi.org/10.3969/j.issn.1000-3428.2016.07.026
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to enhance the robustness on affine transformation and reduce false recognition rate of object recognition,an object recognition method based on Affine Scale-Invariant Feature Transform(ASIFT) and graph matching is proposed.It detects key points in the image by using ASIFT method,extracts the local features of each key point,including ASIFT features,texture features and color features,and uses three-tuple graph to represent these features.Objects are recognized by using separate layer matching,which executes graph matching after successful key point matching to reduce the false recognition.Parameter selection and recognition performance of the new method are evaluated comprehensively on two common datasets including PVOC-2007 and COIL-100.Results show that the new method has higher recognition rate and is a valid object recognition method.
  • XU Heli,NIU Lijun
    Computer Engineering. 2016, 42(7): 159-164. https://doi.org/10.3969/j.issn.1000-3428.2016.07.027
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Based on the hierarchical clustering algorithm and density clustering algorithm,a novel arbitrary shape clustering algorithm is proposed.In this algorithm,the dataset is divided into a large number of initial sub-clusters by finding the initial clustering center which is the density peak point.According to the clustering merging criterion,two sub-clusters are merged if their border density is greater than any one of the cluster average density.The algorithm uses the method of dynamic modeling to merge the sub-clusters,which can automatically determine the clustering termination point without the artificial input parameters.Experimental results on the test datasets and real datasets illustrate that this algorithm is robust to the choice of input parameters,and can effectively identify the clustering of arbitrary shape,size and density.It is suitable for the dataset with uneven density distribution.
  • PU Xingcheng,WANG Tao,ZHANG Yi
    Computer Engineering. 2016, 42(7): 165-172,180. https://doi.org/10.3969/j.issn.1000-3428.2016.07.028
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to solve the problem of poor robustness and low recognition rate in static gesture recognition based on the Kinect depth information,this paper presents a new gesture recognition method based on the improved Hu moment algorithm.It constructs a new invariant Hu moment by using the scale normalization method to eliminate the influence of scale on Hu moment,which can keep the static gestures’ invariability with shifting,scaling and rotation under the discrete states.Moreover,the Kinect sensor and Open NI are used to get the depth information of images.The obtained depth information is pre-treated,such as the noise elimination.Then the histogram method is used for gesture segmentation and feature extraction of the number of fingers.Finally,it uses the improved Hu moment algorithm to recognize static gestures.Experimental results show that the improved algorithm has stronger robustness and higher recognition rate even under complicated backgrounds.
  • QIU Yunfei,CHEN Yifang,WANG Wei,SHAO Liangshan
    Computer Engineering. 2016, 42(7): 173-180. https://doi.org/10.3969/j.issn.1000-3428.2016.07.029
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the wrong recognition of opinion target in online Chinese comments caused by diverse and non-standard language, a method of commodity opinion target extraction based on the part of speech features and syntactic analysis is proposed. According to the characteristics of Chinese language, rules are constructed to extract evaluation words by part of speech features of adjectives, adverbs, and verbs. Through the syntactic tree structure of clause sequence, opinion target candidates are extracted and filtered. Evaluation collocation is screened based on the core syntactic paths to reduce the noise of opinion target and polarity word introduced in the process of extraction, thus, extracting the real opinion target. Experimental results show that the introduction of syntactic tree structure and core syntactic path makes the F value of commodity opinion target recognition over 80%.
  • SHEN Jiquan,KONG Xiangjun
    Computer Engineering. 2016, 42(7): 181-188,193. https://doi.org/10.3969/j.issn.1000-3428.2016.07.030
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The existing Quality of Service(QoS) service composition methods fail to measure the interval QoS attributes, and overlooks the data distribution features in the QoS attributes and the description uncertainty for QoS demand, thus resulting in a big difference between the real composition results and the ideal ones. Therefore, a service composition method of QoS attribute interval number is proposed based on the improved Ant Colony Optimization(ACO) algorithm. User satisfaction and QoS utility function in interval numbers are used to construct the objective function of service composition. By using the improved pheromone updating and parameter selection strategies, the convergence speed is increased and the best service composition with high user satisfaction and service performance is found on the basis of satisfying the global QoS constraint. Experimental results show that the method can improve the efficiency and success rate of service compositon.
  • ZHENG Gang,QI Jia,CHEN Chen
    Computer Engineering. 2016, 42(7): 189-193. https://doi.org/10.3969/j.issn.1000-3428.2016.07.031
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Identification based on Electrocardiogram(ECG)signal is an emerging study hotspot in biometric identification.This paper proposes a feature selecting strategy.The fiducial point detection accuracy is used to modify the original contribution rate of the feature,so as to obtain the real applicable contribution rate and the key feature set for identification.This paper adopts the Physikalisch-technische Bundesanstalt(PTB)data sets and self-collecting ECG data sets.Experimental result shows that the proposed strategy increases the features of PTB data sets from 9 to 12 and that of self-collecting data sets from 17 to 18.Identification accuracy reaches 98.7% and 90.21% respectively.
  • CHENG Xi,CHEN Jun
    Computer Engineering. 2016, 42(7): 194-198. https://doi.org/10.3969/j.issn.1000-3428.2016.07.032
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of data sparseness and system scalability in traditional collaborative filtering algorithms,this paper proposed a new collaborative filtering algorithm.This algorithm firstly obtains the item rating matrix according to users’ rating on different items.Secondly,it utilizes naive Bayesian classifier to classify the items,and then searches for the items’ nearest-neighbor sets in the same class by the modified cosine similarity computation method.At the same time,it uses the MapReduce parallel computation framework on Hadoop to implement distributed data processing.Finally,it forms a rating prediction list and makes recommendations.Experimental results show that the algorithm not only effectively solves the problem of low prediction accuracy due to data sparseness,but also improves the accuracy of recommendation compared with collaborative filtering algorithm based on user classification and that based on item classification.It greatly improves the efficiency and scalability of the system by parallel computation.
  • QIU Renbo,LOU Zhen
    Computer Engineering. 2016, 42(7): 199-202,208. https://doi.org/10.3969/j.issn.1000-3428.2016.07.033
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Most of the improved AdaBoost algorithms based on margin theory diverty optimize the loss function to improve the gereralization ability of the algorithm.This paper proposes an improved parameterized AdaBoost algorithm.It improves the weight adjusting strategy of AdaBoost algorithm and increases the weights of the training samples of which the margin increment moves from positive to negative,so as to resist the negative movemen of the margin in crement and better optimize the loss function.Through 100 times of 5-fold cross-validation,results show that,compared with other algorithms,such as PAB,IPAB,etc.,the proposed algorithm has better classification accuracy and stability.
  • ZENG Jianping,LIU Hua
    Computer Engineering. 2016, 42(7): 203-208. https://doi.org/10.3969/j.issn.1000-3428.2016.07.034
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In many application fields of Chinese text analysis, name recognition is a widespread fundamental issue. Although there are several kinds of name recognition methods, they are mainly based on corpus statistics and language rules. A new name recognition method is proposed based on clustering coefficient according to the characteristics of the close relationship of the names in the event text. The method captures Chinese family names from the original text based on a standard list, and then candidates are selected according to the contextual information and statistical model constructed by the corpus information of personal names. Interpersonal semantic similarity and name probability are defined, and a personal name filtering method based on interpersonal network clustering coefficient is designed. Experimental results show that the F1 value is improved by 1.2% compared with the existing method based on hidden Markov model, and there is no need to manually label the corpus or use language rules.
  • LIU Wanjun,ZHAO Qingguo
    Computer Engineering. 2016, 42(7): 209-215. https://doi.org/10.3969/j.issn.1000-3428.2016.07.035
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the running efficiency of image dehazing algorithm and solve the color distortion problem in the bright area after dehazing, this paper puts forward a Fast Image Dehazing Algorithm based on Twice Filtering(TBFIDA).It gets the ambient light brightness with strong robustness and high stability by the improved quadtree method,and obtains a coarse scattering map using the discriminant filtering in minimum channel,eliminating the phenomenon of halo.And it makes full use of bilateral filter for further optimization on edge preserving to get more precise scattering map,followed by using the tolerance mechanism to correct the transmissivity and resolve color distortion in the bright area after dehazing. Finally it gains the fog-degraded image based on the atmospheric scattering model.Experimental results show that TBFIDA algorithm can well restore color,better preserve the edge details of the area with suddenly changed depth of field,and get a real clear no-fog image.It has higher accuracy,robustness,and real-time performance.
  • ZHANG Jinfeng,ZHANG Jiye
    Computer Engineering. 2016, 42(7): 216-219. https://doi.org/10.3969/j.issn.1000-3428.2016.07.036
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to achieve fast and accurate autofocus and reduce the effect of reverse backlash on autofocus,by using variable step and subsection control,an automatic focusing strategy is proposed in this paper.Through reducing the number of reverse motion,the focusing control method realizes both rough and fine adjustment in automatic focusing.The changing rate of the evaluation function is used as the judging function and the average changing rate of the first 5 points before the current point as the threshold.In rough focusing stages,fast focusing is achieved by adopting the large step.In fine focusing stages,a small step is used and reverse adjustment is reduced,thus achieving fine focusing.Experimental results show that the strategy is able to reduce the reverse backlash error,overcome local maxima and achieve rapid and accurate focusing.
  • DAI Qingyan,ZHU Zhongjie,DUAN Zhiyong,LI Weijie
    Computer Engineering. 2016, 42(7): 220-226. https://doi.org/10.3969/j.issn.1000-3428.2016.07.037
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Based on the classical graph cut theory,an image segmentation algorithm using super-pixel technique and improved iterated Graph cut algorithm is proposed.The super-pixel representation with the edges of foreground objects well preserved is obtained through an improved Simple Linear Iterative Clustering(SLIC) algorithm.Taking super-pixel as processing unit,segmentation is performed based on the framework of Graph cut,where the energy function is rebuilt by incorporating color and gradient information.Experimental results show that,compared with the Graph cut algorithm,the proposed algorithm not only has higher segmentation accuracy,extracts foreground objects with edges more complete and smooth,but also enhances the segmentation efficiency more significantly.
  • GUO Beibei,YI Sanli,HE Jianfeng,MIAO Ying,SHAO Dangguo
    Computer Engineering. 2016, 42(7): 227-231. https://doi.org/10.3969/j.issn.1000-3428.2016.07.038
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The non-local means filtering algorithm is an efficient denoising algorithm that can restore details of image texture well.However,the algorithm cannot adaptive adjust filtering parameters and the filtering result is easy to produce artifacts.Owing to these shortcomings,an adaptive non-local means filtering algorithm combined with isotropic algorithm is proposed.It estimates variance of image noise by using improved Laplacian operator.The estimated variance of the noise can be applied to the calculation of filtering coefficient,which can realize the automatic selection of filtering coefficient.Local variance is taken as adjustable factor to combine non-local means algorithm and isotropic local filtering algorithm.By this way,the artifact generated by filtering image is reduced and the quality of the image can be improved.Experimental results show that,compared with non-local means filtering algorithm,the proposed algorithm can better reserve details of image texture,at the same time reduce the artifact more efficiently,and improve the peak signal to noise ratio of the image,with a better noise reduction effect.
  • FU Hui,WU Bin,LI Linfei,ZHANG Hongying
    Computer Engineering. 2016, 42(7): 232-237. https://doi.org/10.3969/j.issn.1000-3428.2016.07.039
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Traditional dark channel prior algorithm needs a long time to deal with a single image,and the quality of the processed image can be further optimized.Aiming at this problem,this paper proposed an improved algorithm.It adopts Canny edge operator to locate the value interval of atmosphere vector A and determines A’s value precisely by the bipartite graph algorithm.Then it uses the median filtering based on Rank transform to process transmissivity,which greatly removes noise and maintains edge details.The transmission image is further optimized by dilation and erosion.Experimental results show that compared with the traditional dark channel method,Retinex algorithm,and other algorithms,the proposed method has higher processing efficiency,and the foreground color of the generated image is bright.A small amount of haze is kept,which makes the image more real.
  • LIU Yang,WANG Peng,YANG Rui,ZUO Xianyu,ZHANG Zhouwei,WU Xiaoyang,QU Jiantao
    Computer Engineering. 2016, 42(7): 238-243,250. https://doi.org/10.3969/j.issn.1000-3428.2016.07.040
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The traditional image classification algorithm suffers from low efficiency and is unable to meet the requirement of mass and real-time processing of high resolution remote sensing data.Based on the analysis and research of the Iterative Self-organizing Data Analysis Techniques Algorithm(ISODATA) for remote sensing images of ZY-3 satellite thematic production, this paper designs a Parallel ISODATA Based on OpenMP(PIsodataOmp).The OpenMP technology is used for the clustering of sample points and the calculation of the standard deviation of the clustering center, realizing parallel processing on a single multi-core computer based on the shared memory.Experimental results show that, when using the PIso data Omp classification to process the image data of ZY-3 satellite, the speed of classification is improved significantly while ensuring the accuracy of classification.
  • ZHU Shisong,FAN Jingfang,ZHU Hongjin
    Computer Engineering. 2016, 42(7): 244-250. https://doi.org/10.3969/j.issn.1000-3428.2016.07.041
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    In traffic videos,the impact of environment,installation angle of equipment and other factors may cause vehicle occlusion,resulting in the error of vehicle detecting and tracking.Hence,this paper presents a method based on contour feature points to detect and segment the overlapped vehicles.It uses the background difference method to obtain the target areas,detects the contour points of target area by Freeman chain code,and identifies the feature points by chain code pairs.The number of feature points and the occupancy ratio of target area are used to judge the overlapped vehicles.If they are overlapped,convex hull analysis is carried out in the target area to find the optimal segmentation points and segment the overlapped vehicles.Experimental results show that compared with related methods based on concavity analysis and ellipse fitting,this method can segment the overlapped vehicles more precisely and has better adaptability without any prior knowledge except the vehicle shape.

  • XU Ben,ZHOU Zhihu,FAN Liangzhong
    Computer Engineering. 2016, 42(7): 251-256. https://doi.org/10.3969/j.issn.1000-3428.2016.07.042
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    To solve the problem of weak real-time processing ability and applicability of most current video stabilization algorithms,a digital video image stabilization algorithm based on AKAZE features for complicated shakiness is proposed.Feature points of video images are extracted by AKAZE features,and feature points between video adjacent frames are matched by Fast Library for Approximate Nearest Neighbor(FLANN) algorithm.In order to improve the accuracy of motion estimation,a feature point screening mechanism is proposed to eliminate mismatched points.And random sample consensus algorithm is used to solve the motion parameters between video images.To obtain the stabilized video motion parameters,a Gaussian filter is applied to smooth out the motion parameters,and the stabilized video can be obtained from motion compensation process for video sequences.Experimental results show that in the case of many feature points,the motion estimation speed of the proposed algorithm is approximately triple as fast as Speed Up Robust Feature(SURF) algorithm.It has stronger robustness and higher real-time performance than the motion estimation algorithm with SURF points,and can achieve more reasonable stabilization results than the algorithms like Content Preserving Warps for 3D(CPW-3D) and Subspace Video Stabilization(SVS).
  • WANG Gang,WANG Xiaodong,CHEN Chao,WANG Chaolin
    Computer Engineering. 2016, 42(7): 257-260,266. https://doi.org/10.3969/j.issn.1000-3428.2016.07.043
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the deficiency of blockiness produced by using the contrast between regions for saliency detection,a multi-scale superpixel saliency detection algorithm is proposed.The algorithm uses a Simple Linear Iterative Clustering(SLIC) superpixel segmentation method to divide an image into regions of different sizes,and detects image saliency map according to the contrast between regions.An adaptive center surranding method is proposed to eliminate the blockiness of the salienly map.The final saliency map is obtained by multi-scale fusion.Experimental results on data sets MSRA-1000 show that,compared with classic algorithms like SEG,MSS and so on,the algorithm has higher precision and better recall rate.
  • YANG Jilian
    Computer Engineering. 2016, 42(7): 261-266. https://doi.org/10.3969/j.issn.1000-3428.2016.07.044
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    This paper proposes a novel unsupervised method for slight change information extraction from multitemporal hyperspectral image sequence based on the Minimum Mean Absolute Error(MMAE) index.To consider the spectral signatures in hyperspectral data,the Vector Autoregressive(VAR) and fitting models are exploited to create a prediction of single-band and multi-bands time series,while MMAE index is introduced to generate the Preliminary Change Information Image(PCII),which contains significant change information.Experimental results show that the proposed method is more effective in extracting the slight change information effectively in the hyperspectral image sequence preserving significant change details and suppressing random variation of the background.compared with the SCD,PCA,CVA and IR-MAD methods.
  • LI Jun,ZHENG Jiting,SHI Danlin
    Computer Engineering. 2016, 42(7): 267-272. https://doi.org/10.3969/j.issn.1000-3428.2016.07.045
    Abstract ( ) Download PDF ( )   Knowledge map   Save

    According to the problem that travel service mode and travel demand data sources are limited in current traffic information service, this paper proposes a method for modeling the behavior of a traveler by studying the travel behavior of individual traveler. It designs a Stay Point Recognition Algorithm(SPRA) based on conditional constraint, to construct the model of traveler’s mobile behavior. Finite Stay Point(FSP) clustering algorithm is put forward to eliminate behavior trajectory dissimilarity caused by GPS error. Intelligent Semantic Matching(ISM) algorithm based on Points of Interest(POI) is presented, which can establish a travel mode sequence containing POI. The effectiveness of the above methods is validated by using simulation experiments. Results show that the stay point recognition accuracy can reach 90% after using SPRA and FSP, while ISM algorithm has a higher recall ratio compared with the behavior recognition algorithm based on cosine similarity.

  • MA Lili,WANG Zhan,CHEN Jinguang
    Computer Engineering. 2016, 42(7): 273-276,286. https://doi.org/10.3969/j.issn.1000-3428.2016.07.046
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    The number of targets is under-estimated in the standard Cardinality Balanced Multi-target Multi-Bernoulli(CBMeMBer) algorithm when the detection rate is low.To address this problem,an improved CBMeMBer algorithm is proposed.In the previous filtering process,some particular Gaussian items are modified and saved.And these items are merged into the updated Gaussian items and used in the next filtering circle.As a result,the negative effect of low detection rate is reduced.Simulation results show that the improved algorithm can solve the problem of under-estimated target in some extent and improve the accuracy of target tracking under low detection rate.
  • YU Qunxiu,BI Guihong,ZHANG Shouming
    Computer Engineering. 2016, 42(7): 277-286. https://doi.org/10.3969/j.issn.1000-3428.2016.07.047
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    This paper proposes a complex Agent network model to study the interaction between social networks and social wealth.The Agents take advantage of different social radius to form acquaintance networks based on the theory of social circles in the sugarscape.Meanwhile,the Agents choose friends to form the wealth-sharing network through the rules in the acquaintance network.The Agents find wealth in the friendship network of wealth-sharing to achieve wealth distribution for the first time.The network adds social wealth trading mechanism to deal with the distribution of wealth.Through trading,the Agents complete the second wealth distribution.The Agent’s wealth makes its acquaintance network and friend network change,which finally leads to the co-evolution between social wealth and social network.Simulation results show that this model can reasonably adjust social relations.It can minimize the social inequality,and good social relation network can be developed.
  • DAI Dong,WANG Guo,WANG Lei
    Computer Engineering. 2016, 42(7): 287-292,298. https://doi.org/10.3969/j.issn.1000-3428.2016.07.048
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    At present,majority methods of tracking ground target using fixed-wing Unmanned Aerial Vehicle(UAV) need one or more strict assumptions,which limits the use of UAV.Aiming at this problem,a method of UAV tracking based on game theory is proposed.Firstly,a fourth-order dynamic equation is used to model and initialize the tracked objects and UAV.Then,discrete dynamic time of the whole system is obtained by Zero-order Hold(ZOH) of two dynamic equations,with getting the game objective function.Finally,dynamic programming is used to determine the optimal control strategy.The effectiveness of the proposed method is verified by simulation results under practical applications using a fixed-wing UAV with single camera.Compared with the methods based on segmentation guiding and visual servo control,the proposed method has smaller variances of elevation and distance,so the target is always in the visible range.In addition,wind is considered in the experiments.The experimental results show that the maximum wind speed resisted by UAV in a conservative estimation is 6 m/s around.
  • MAO Weiwei,ZHOU Feng,LIANG Qing,FENG Yu,YAN Yong
    Computer Engineering. 2016, 42(7): 293-298. https://doi.org/10.3969/j.issn.1000-3428.2016.07.049
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Aiming at the problem of the optimal control on biped robot’s walking stability,the scheme of active and passive vibration attenuation and landing impact elimination for crus structure design in biped robots is proposed,which can maintain stability of the upper platform to ensure sensor’s running conditions.The SimMechanics toolbox of Matlab is used to build the mechanical simulation model of the biped robot’s crus and the active vibration attenuation approach based on the Active Disturbance Rejection Control(ADRC) algorithm is applied to suppress low-frequency and high-frequency vibrations via the active and passive vibration attenuation respectively.Simulation results show that the active and passive vibration attenuation approach successfully attenuates the impact from the foot,and makes up for the disadvantage of traditional passive attenuation approach effectively.It can achieve stability of the upper platform of biped robot,which offers the basis for the stable walking of biped robots.
  • DONG Yongliang,CHAI Xuqing
    Computer Engineering. 2016, 42(7): 299-303,309. https://doi.org/10.3969/j.issn.1000-3428.2016.07.050
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the accuracy of similarity matching and ensure interpretability of retrieval results in multi-modal information retrieval,a two-layer multi-modal semantic network is proposed.Firstly,a sub-semantic network is built for the data of each single model,and the nodes in each sub-semantic network are clustered into different groups.Secondly,by assuming each group in the sub-semantic network as a node,a multi-modal semantic network is built based on semantic relationships,and the nodes in this multi-modal semantic network are further clustered into different groups.While retrieving information,the information can be retrieved by reversing steps of building the multi-modal semantic network.Experimental results show that the proposed method has higher retrieval accurary than the methods based on Hash index,low-rank matrix embedding or deep neural network.
  • CAO Xizheng,NIU Jingwen,WU Ruiqi,SUN Zhiyong
    Computer Engineering. 2016, 42(7): 304-309. https://doi.org/10.3969/j.issn.1000-3428.2016.07.051
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    Automatic accompaniment technology is mainly based on the seven tones and the major and minor modes of Europe,and it is still very weak in the research on the five-tone national mode of China.For this reason,an automatic accompaniment algorithm is proposed,which is oriented to the melodies of Mongolian songs.Firstly,the accompaniment chord generation algorithm is constructed according to the melody features of Mongolian songs.And the accompaniment chord sequence for the main melody is generated by this algorithm and the pre-constructed harmony progression rule base.Secondly,the accompaniment figure matching algorithm is designed,and the appropriate accompaniment figure is selected from their pre-constructed knowledge base.Finally,the whole accompaniment is generated by sequencing to the accompaniment figure.Experimental results show that the proposed algorithm is highly recognized by people with little or no musical knowledge,and can meet the accompaniment demands of them.
  • WEN Haoxiang,HONG Yuanquan,LUO Huan,ZHOU Yongming
    Computer Engineering. 2016, 42(7): 310-314. https://doi.org/10.3969/j.issn.1000-3428.2016.07.052
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    A new algorithm using Moving Window Integration(MWI) for locating active coefficients is proposed to improve the location accuracy.Combined with the dual-filter structure,the new algorithm uses the first filter W1(k) to identify the objective system.The square values and moving window integration of W1(k) are caculated.The region with the maximum is determined to be the active coefficient region.A second filter W2(k) is used to accurately identify the active coefficients belonging to the located region.Experimental simulation is conducted under the application background of echo cancellation.The results show that compared with the traditional location algorithm,the new algorithm improves the location accuracy and enables W2(k) to cover all active coefficients exactly,which eventually leads to faster convergence speed and lower misalignment.
  • ZHANG Zhuoran,YE Guangqiang,ZHAO Xiaolin
    Computer Engineering. 2016, 42(7): 315-321. https://doi.org/10.3969/j.issn.1000-3428.2016.07.053
    Abstract ( ) Download PDF ( )   Knowledge map   Save
    In order to improve the performance of Square Root Cubature Kalman Filtering(STSRCKF) algorithm to track maneuvering target in single observer passive tracking,a Strong Tracking Modified SRCKF(ST-MSRCKF) algorithm is presented.Target state variables and the square root of error covariance matrix are predicted by standard Kalman filtering replacing the method of approximate calculation with weighted sum of Cubature point.A new measurement equation is established with the first estimate value of state and the target state is estimated secondly by standard Kalman filtering,thus improving the filtering accuracy.Meanwhile,with reference to the Strong Tracking Filter(STF),by adjusting the gain matrix in real-time with introducing a time-varying fading factor into the square root of the error covariance matrix,the ST-MSRCKF has the capability of adaptive target tracking and its robustness to deal with sudden change maneuver is enhanced.Simulation results show that this algorithm has higher tracking accuracy than SRCKF algorithm in case of normal maneuver and sudden change maneuver.