Abstract:The popularity of multiple-core devices has made parallel programming a necessity to harness the abundant hardware resources. However, due to the non-determinism of parallel software, writing robust parallel software is notoriously hard. Therefore, how to debug concurrency bugs efficiently has become an issue that needs to be urgently deal with. In this paper, we have surveyed the parallel debugging technologies systematically. Further, based on the survey, we have made classifications and comparisons. At last, we have presented the prospects of the possible development direction of concurrency bug debugging approaches.

Abstract:Trusted internet has been a research hotpot in recent years. This paper introduces trusted identification of inter-domain routing, DNS resolving and trusted-application, including PRKI, DNSSEC and trusted website identification. Finally, we emphasize specially on the importance and significance of trusted website identification.

Abstract:To meet both the needs of reusability and semantic interoperability of limits and fits knowledge and the needs of limits and fits data access performance, the tier structure of shared concepts and data storage is introduced to organize the limits and fits knowledge. Considering the limits and fits ontology as the shared concepts, we manage the limits and fits data with relational database management system and express the relationship between the data and the concepts by mappings. Then the limits and fits data can be accessed through the limits and fits ontology. In this way, the needs of limits and fits knowledge reusability and semantic interoperability are met, while ensuring the efficiency of limits and fits data access. The feasibility and effectiveness of ontology based limits and fits data access are demonstrated by a prototype system and a practical example.

Abstract:Aiming at the demand for monitoring and change research on environment and resources in Kashgar of Xinjiang province, a rapid order-style dynamic monitoring and evaluation system on environment and resources was constructed. Based on ArcEngine components and rich functional classes, the system was built upon the combinations of orderly spatial analysis operations, "virtual raster" in.afr format as a decent medium for band operations on remote sensing images, and multithreading technology. The architecture of the system and its core functions were exhaustively introduced, including task management, information extraction on water bodies, vegetation and ice, dynamic monitoring and evaluation, and thematic map production. Furthermore, the system application examples were presented. Meeting the requirements on operational applications, the system provided a good way to timely monitor changes and developments on environment and resources in a region, thus contributing to the regional economic development. The system can be widely used in the field of monitoring on environmental factors and resources, which displayed certain versatility and scalability.

Abstract:Information retrieval (IR) is widely used in automatically discovering requirement traceability. However IR will miss some correct artifacts which have low text similarity with the requirement. There are accuracy issues in requirement traceability based on IR. To solve the problem, we propose an approach of using the developer collaborative relationship to improve the accuracy of the traceability links recovery between requirement and source code. Meanwhile, we develop a requirement-to-code traceability system. When the system is tracing, it retrieves the source code artifacts of the highest text similarity with the requirement and extracts the developer collaborative relationship from code commit logs. Then the system recommends some relevant code artifacts by developer collaboration relationship. Users can choose the correct code artifacts from the recommend result. The experiment shows that the requirement traceability system could improve the accuracy and the efficiency and reduce errors.

Abstract:To address the problem of anchor speakers tracking in broadcast news, an algorithm that integrated effectively speaker segmentation and clustering and speaker verification is presented. An anchor speakers tracking system based on the proposed algorithm is also designed. The system firstly uses audio activity detection algorithm in order to remove silence segments. Secondly, a speaker segmentation and clustering process converts multi-speaker speech waveform into several speaker homogenous segments. Finally, speaker verification approach based on GMM-UBM is dedicated to decide whether a segment belongs to one targeted anchor speaker. Furthermore, the impact of T-norm on system performance is also analyzed. Experiments on CCTV Mandarin Broadcast News demonstrate the effectiveness of the proposed algorithm. The tracking system achieves precision and recall at 93.03% and 84.34% respectively.

Abstract:Most shopping sites, represented as a routine of sameness, are dull in shopping environment, shopping in supermarket, we can be responsed to social, cultural and environmental factors. In order to make up these deficiencies, Visual Environment Simulation and information management tools have been presented in this work. We build a 3D business enviroment and vision roaming based on OSG technology to realize the 3D graphic based shopping process.

Abstract:An A2P architecture centered on venue agent was proposed. It consisted of server, management terminal, venue agent and player terminal. The network nodes were organized into several groups by venue agents. The algorithms of both grouping and downloading were proposed to reduce the complexity of system management, at the same time, it improved the speed of spreading files. The function of scheduling task was managed by template, and released in B/S mode which can increase their flexibility. Based on the architecture, the multimedia information display system was designed and achieved good results.

Abstract:Considering the picture has the futures that a strong interpretation of events and convenient disseminating, this paper studies extraction of data from a large number of news web pages, and organizational structure chart presented to the users. It achieves dynamic pages based on page template extraction and analysis, processing converted to the corresponding sets of datastructure. Based on the news cosine correlation graph data sets from different sites are differentiated, and in accordance with the appropriate standards for data collection to score sorted. This system is based on hadoop distributed platform, considering the large number of users and imgsets. This paper will describe the design and implementation of our system in detail, and report the results of running the system on Baidu news image column.

Abstract:In order to improve the utilization of medical resources further, this paper improves the quality of medical service, which can reduce the medical cost and risk. To construct a classification treatment model more perfect and reasonable, it is need to build a regional medical information sharing and collaborative services platform which uses the medical service institutions as the main body. The medical resources and information are sharing as the goal. Integrate the generic technology and medical service techniques, and implement the sharing and the collaborative about the regional medical information, resources and the service. In this paper, according to the demand of the regional medical information sharing and collaboration service requirements, we propose the overall architecture of the platform, and we also put forward solutions to propose the key issues which are the main index of patients, the document sharing crossing regional, the clinical document's construction, the term service, data acquisition and transmission, the hospital pre service, document storage and so on.

Abstract:The existing cloud computing resource scheduling strategy did not consider the overall requirements of resource pool resource requirements, causing the loss of resource and affecting the normal operation of the virtual machine. This paper presents a cloud computing active scheduling method based on the resource monitoring statistics, giving full consideration to overall resource demand of resource pool. It introduces monitoring data acquisition module and the monitoring data analysis module which adds the monitor database as the foundation of the system. We customizes a set of monitoring data statistics and analysis the method for resource scheduling based on the virtual machine, making the virtual machine scheduling method based on the monitoring data can be easily realized.

Abstract:Virtual disk snapshot is a prevailing technology to support virtual machine failover. It's usually organized into linked structure, which may cause virtual machine performance degradation because this method would introduce additional disk I/O overhead. In this paper, we analyze the efficiency of disk snapshot technology with non-linked structure. By integrating ZFS and OCFS2 file system based on non-linked structure, we implement an efficient management system for virtual machine snapshots. In order to improve the efficiency of snapshot retrieval, we also propose a virtual machine snapshot storage organization model and optimize a snapshot search algorithm. The result of comparative experiment indicates our system's effectiveness.

Abstract:An implementation of Telnet server for non-os embedded system is proposed in this paper. User is able to log in to the target embedded system through Telnet client program in local computer, and remote debugging and controlling can be realized. The designed Telnet server is divided into two modules: server module and shell module. Server module handles connection requests of clients and negotiates options, while shell module is responsible for processing the interactive data.

Abstract:E-government extranet is a window to display the image of the government, service people. Its purpose is resource sharing, efficient convenience, the security is particularly important. In this paper, in order to solve the security problem of e-government extranet, we put forward the design scheme of security system. Analysis shows that, by applying cryptographic service, unified user management, unified identity authentication, unified resource and authorization management and a series of safety measures can effectively guarantee the safety and reliability of e-government extranet.

Abstract:A new design of the digital broadcast system based on 3G means of transmission and a cloud computing is proposed to replace the traditional one, which has such disadvantages as the coverage restriction, the high covering and maintenance costs, frangibility to interference and poor capacity of emergency information dissemination. Consisting of three parts including the broadcast control platform, the control and the receiving terminal, the new system uses the 3G network infrastructure of Telecom to achieve grouping and fixed broadcasting, thus solving the urgent broadcasting issues in emergency. Experiments have shown that this system owns many advantages, such as resource pool, elastic supply, simplified terminal and rapid deployment.

Abstract:In order to generate self-check signal and speed instruction of radar antenna system test signals, design a kind of mixed signal generation system based on C8051F206 and FT245R.The system uses FT245R USB chip to realize the communication of upper machine, and rely on C8051F206 high SPC and C language to response the instructions of upper machine in a reliable way and send real-time measured parts required for the self-check instruction,speed instruction,etc.This paper introduces the principle of hardware design and software design of mixed signal systems. Practical application shows that the system can accurately generate radar antenna system self-check signal, speed signal and real-time feedback test results.

Abstract:This paper presents an improved niche genetic algorithm applied to multimodal function optimization for finding all the extreme solutions. This algorithm is pre-selected niche based and similarity based on the mechanism of eliminating the niche combination. We improve the adaptive crossover operator and mutation operator according to the probability, crossover probability and mutation probability the fitness value to dynamically adjust the individual. And the algorithm is used to solve a typical multi peak, the experimental results show that the niche genetic algorithmcan searchall themultimodal functions' optimal solutions and extreme solutions, and has faster search speed. At the same time, this method is universal in the multi peak function for the other.

Abstract:Due to the wireless sensor network (WSN) is provided by battery, energy management becomes a fundamental problem in such network. In view of the maximization lifetime problem of the restricted multiple mobile base stations network, we will propose the MMBEC algorithm in this paper. Since the mobile station was limited by the actual road and its own energy, we first balance the energy load of base stations by sub loops division, and then we move the base stations periodically in order to achieve the data traffic balance of nodes adjacent roads. Because the balancing of sub loops problem belongs to NPC problem, in this paper we proposed an approximation algorithm for the purpose of finding an approximate solution. Control the mobile base stations stay periodically ensures the energy of the nodes adjacent to roads exhaust almost at the same time, and therefore, prolong the lifetime of network. The results of simulation experiment show that the algorithm prolongs the lifetime of network and improves the throughput capacity of data when compared with the existing algorithms.

Abstract:After a failed test is encountered and the error trace is generated, a significant amount of effort is often required for programmers to manually examine the program code and localize the failure's root cause. In this paper, we propose a fault localization method based on program execution trace. This method computes a set of possible error statements based on a combination of weakest pre-condition computation through program's reverse execution. All possible error statements are organized in a fault localization tree to help developer identify the root cause. Our experiments on SIR datasets demonstrate that our method can not only efficiently compute the possible causes, but also provide sufficient information to help programmers quickly locate the root cause.

Abstract:Due to the large amount of training data and the high complexity of its recommend algorithm, the updating cycle of recommendation system tend to be long. However, the data on the system is growing all the time, and a lot of data is produced during the cycle, which is useful for the recommendation of next moment, and recommendation system can't use these data in time. In order to use these data in time to improve the quality of recommendation system, a new approach to hybrid recommendation based on incremental data was proposed. The approach mainly divided recommendation into offline and online module, the offline module is used to produce the personalized recommendation list, while the online recommendation module maintains a list of popular trend momentum based on real-time and incremental data. Then, combining with the results of the two modules, based on which give users anonymous or personalized recommendation. Experiments show that the approach is simple, effective, feasible, and can improve the performance of recommendation system better.

Abstract:When a bug was reported to the bug tracking system, it should be assigned to a developer who is responsible for its resolution after it is confirmed. This processing is called bug triage. With increasing number of bug reports submitted to the bug tracking system, it is more and more difficult to assign appropriate developers to the reported bugs manually. In this paper, we propose an approach called BUTTER (BUg Triage by topic modeling and heTERogeneous network analysis) to automatically assign bugs to developers. Different from existing work, BUTTER not only uses topic model to analyze the text information from bug reports, but also innovatively takes structural information into consideration by constructing a heterogeneous network which includes relationships among submitters, bugs and developers. Experiment shows that BUTTER outperforms other methods on automated bug triage.

Abstract:With the application of the thermal manikin and fire manikin, mass data are produced during the experiment research in the clothing engineering area. The advantages of the larger samples can not be revealed using the conventional analysis method. Thus, in the paper, by using the Clementine software, data mining method is used to explore the data produced by flash fire experiment. The decision tree method and the neural net method are used to determine the key influence factors of the thermal shrinkage, which are then used in Kohonen cluster to divide protective clothing into different parts. Research shows that heat flux is the most important factor to the degree of shrinkage and the arm and leg are the key parts to be protected. It suggested that data mining is an effective tool to explore the character and function of the protective clothing.

Abstract:In order to reduce the dimension of high-dimensional data, raised edge semi-supervised marginal discriminant embedding and local preserving algorithm for dimensionality reduction is proposed. By minimizing the distance between sample and the center of its category, the local topology of samples is maintained in the projection subspace. And by maximizing the distance between the edges of different categories, the inter scatter of classes is increased in the projection subspace. Experimental results show that the dimensionality reduction algorithm of semi supervised marginal discriminant embedding and local preserving can get a better projection subspace of the initial feature space.

Abstract:A dynamic optimization model is proposed based on different response time. The model considers the features of multi-retrieval depots, multi-resource and multi-objective emergency scheduling problem, and combines with the constraints of response time led by the sensitive areas of oil spill. With the earliest emergency rescue time and the fewest number of retrieval depots as the optimization goal, the perfect point method and structuring elimination assemble are used to get the optimal scheduling scheme according to different response time slots. Finally, a numerical example is given to verify the feasibility of the proposed scheduling method.

Abstract:Network traffic had long related and nonlinear characteristics, in order to improve the prediction accuracy of network traffic, this paper proposed a network traffic prediction method based on particle swarm algorithm optimizing the parameters of least square support vector machine. Parameters of least square support vector machine were taken as the position vector of particle, and then the particle swarm algorithm is used to find the optimal parameters of the model, finally, the prediction model of traffic model is established based on least square support vector machine with the optimal parameters. The simulation results showed that the proposed model had improved prediction accuracy ompared with other network traffic prediction models and could more accurately describe the change rule of network traffic.

Abstract:With the fast development of cloud computing technology, more and more users choose cloud storage to store personal files. Storage and share technique allows users share files and visit others' files with different kinds of client on the cloud. Storage and share technique brings the demand of large-scale share scene for the versioned files. This is a big challenge for the performance of simultaneously I/O. In this paper, according to the characters of the share scene in cloud storage, we try to dig the relationships between the versions of the file and take the increment based data transmission technique. By doing this, we optimized the performance of rolling checksum skill in increment algorithm and reduce the transmission quantity and improve system storage performance. In addition, this technique can help data transmission work in limited bandwidth and network instability scenario and large-scale share-synchronization scenario.

Abstract:An Improved Cuckoo Search(ICS)algorithm is presented for Lot-streaming Flow shop Scheduling Problem(LFSP)with objectives of makespan. Ordering rule is applied to enable the continuous cuckoo search algorithm to be applied to discrete scheduling problem. Then, after the CS-based exploration, a simple but efficient local search, which is designed according to the LFSP' landscape, is applied to emphasize exploitation. Simulation results show the feasibility and effectiveness of the proposed algorithms.

Abstract:In order to effectively solve the sensors nodes mobility difficult problem in the water, the paper presents a localization algorithm based on mobile nodes—MNLS(mobile nodes localization algorithm). The main reasons of nodes mobility are the waterflow and random noise. The algorithm is on the basis of the analysis of the existing relevant localization algorithms, we present the idea which firstly forecasts nodes' trajectory, then calculates the nodes' distances and coordinates. The algorithm is simulated by MATLAB, and the results show that compared MNLS with Chan algorithm. The single node's positioning accuracy and different speed nodes' precision are all improved. So it has an application value for the underwater sensor network localization.

Abstract:In the human-computer interaction, the keyboard far cannot satisfy the requirement of people. In this paper, a method based on gesture recognition algorithm is proposed, and combine with keyboard keys to realize intelligent input. In the algorithm, the hand region is extracted by using skin color segmentation. According to the hand shape to detect the thumb and the number of outstretched fingers, and their position, then gestures are converted into binary language computer can accept. Combined binary number with the keyboard keys, gestures are defined as the different buttons to achieve compute intelligent input. The experimental results show that the algorithm has certain robustness, and it can achieve a high success rate.

Abstract:The performance of software process is related with the software process models and the resource allocations of software processes. If a correct model is impertinently allocated with the limited resource of a software organization, as a result, the performance of the software process may fail to reach the actual requirement, followed by delay, over-cost, and even failure. The existing approach based on process automata isn't fit for the analysis of instantiation model. A model with only correct structure can't ensure a successful enactment, because it lacks schedule information. This paper presents an approach to verify the process instantiation model with time and resource constraints, which is an extension of existing s-TRISO/ML process modeling language. This paper also presents an approach to convert from an s-TRISO/ML model into timed automata and explains converting algorithm. Finally, making use of UPPAAL to carry out the function verification on the converted timed automata, we can get a reasonable instantiation model to provide guidance for the actual process development.

Abstract:There are a lot of data redundancy in wireless sensor networks. By compressing the original sampling data, the data fusion technology eliminates redundancies in data, reduces the amount of data sent by nodes effectively and prolongs lifetime of sensor networks. This paper proposed a data fusion algorithm that combined data forwarding and compressed sensing. During the process of collecting sampling data in sensor networks, the algorithm selects using compressed sensing to compress original sampling data or simply storing and forwarding sampling data according to the amount of nodes' child nodes. Simulations indicate that compared with the data fusion algorithm based on compressed sensing, the data fusion algorithm that combined data forwarding and compressed sensing achieved both network load balance and data compression effectively.

Abstract:In order to solve the maldistribution problem for wireless sensor network, topology control algorithm based-on regional assignment with switched-beam directional antennas, SRADTC is proposed. Through the number of the key neighbor nodes, the whole network can be divided into several sparse area and dense area. In sparse area, network topology is controlled by the minimum spanning tree algorithm. In dense area, network topology is controlled by the K-Neigh control topology algorithm. Pareto distribution node model will be built to compare the effectiveness between SRADTC and traditional topology control algorithms. Simulation results show that, under the precondition of ensuring the network connectivity, SRADTC improves the performance of network.

Abstract:The human brain functional connectivity detection is an important technique in neuroscience research. The restricted boltzmann machine (RBM), modeling on a large amount of multi-subject functional magnetic resonance imaging (fMRI) data, it can discover the brain functional connectivity. However, the former method with restriction of the huge training data, it can not detect the functional connectivity on single-subject data effectively. In this research, a novel functional connectivity detection model taking advantage of the sparsity is presented, which is an effective combination of the spatial-domain sparse approximation theory and the RBM technique. The experimental results demonstrated that the proposed model could effectively discover both the temporal dynamic model and the corresponding spatial functional maps on the single-subject data, which settled the the bottleneck of RBM.

Abstract:In view of solving linear programming problems with parameters both in objective function and constraints, a computational method based on novel smooth exact penalty function neural networks is proposed. First, the error function is introduced to constructing the approximate function of unit step function, which is used to give the smooth penalty function that more accurately approximates the L1 exact penalty function, and its basic properties are discussed. Second, the neural network model for parameter linear programming problems is constructed based on the proposed smooth exact penalty function and the stability and convergence of the neural networks are proved. Moreover, the specific calculation steps of our proposed neural network model for the optimization are given. Finally, a numerical example is given to illustrate that the proposed method possesses the smaller penalty factor, easier construction and higher accuracy.

Abstract:In equipment maintenance activities, a great amount of maintenance work have to be completed by more than one person in cooperation. In this case, an object may be operated equally by multi-users simultaneously, which is called "concurrent operation". For efficiently handling conflict, many concurrency control methods were analyzed. Then the combination of multiple concurrency control mechanisms was used and a set of effective conflict solution was put forward to shorten the time of conflict resolution. Finally, the feasibility and availability of the method is verified through a practical instance application.

Abstract:Based on the test for anti-evade attack of IDS(Intrution Detection Systems), traditional software testing method was improved, and a simple and effective software method was designed. This improved method use VMware virtual maching, test software were running on physical and virtual machine, the network cards of the virtual maching were set in different operation mode to build the test environment, simplifying the testing method, saving the testing equipment and the test result is correct and effective. Meanwhile, compared with the instrumentation BPS(Breakingpoint Systems) expensive instrumentation isn't required.

Abstract:With the development of e-commerce, blog, social networking sites and micro-blog become much more flourishing. The Internet has entered a new era, and the sentiment classification of online comments related to individual decisions, business management and also social security. A model of sentiment classification based on interval-valued intuitionistic fuzzy sets and the interval-valued intuitionistic fuzzy operator was proposed to calculate the interval-valued intuitionistic fuzzy numbers of feature words. Meanwhile, by using membership, non-membership and hesitation to quantitatively describe the feature words. It can get the sentiment tendency through the sentiment synthesis, in order to take more accurate analysis results of sentiment tredency. Finally, throughing the comparative experiment based on same corpus, it proves that this model has high feasibility, correctness and classification performance.

Abstract:To improve the safety level of large and heavy freight transportation, and to improve the efficiency of rail freight transportation, an automatic out-of-gauge identification system for large and heavy freight has been developed in the present paper. The developed system adopts information including the three-view drawing of the freight and the user specified type of train vehicle, and then performs comprehensive analysis to obtain three-dimensional structure of the freight by comparing with the gauge of railway, which at last results in the automatic out-of-gauge identification technology for freight loading. The system can help choose train vehicle before transport of goods, and provide the level of out-of-gauge to ensure the safety of the transport.

Abstract:Human brain is one of the most complex systems in the world, and the interaction between different regions has constituted brain network. To aid scientific research, scholars build brain functional networks with variety of methods. In this paper, brain functional networks were estimated by using AAL (Anatomical Automatic Labeling) template and Canonical Correlation, and some topology properties such as small-world property, global efficiency, local efficiency, etc. were studied. The result indicates some characteristics present significantly statistical difference between normal population and abnormal psychological sailors. The prominent conclusion demonstrates the proposed way of building functional networks is feasible.

Abstract:In Linux multi-core environment, network driver adapter is an important factor that affects network performance. When receiving a data package, network driver adapter responses with hardware interrupt firstly. Then it schedules with NAPI mechanism and uses software interrupt to forward the data package upward to the network layer. By analyzing to know the data receiving process of network driver adapter, this paper tries to optimize the network driver adapter. Furthermore, it designs appropriate experiments, and the experiment result approves the network performance is improved to a certain extent.

Abstract:To better understand the relationship between the main factors affecting the mental health of college students as well as psychological symptoms between a university's 2011' students' psychological test data, the research uses statistical analysis and association rule mining two species method. From gender, only-child or not, native place, student cadre or not, family structure, family's monthly income to analysis research. According to the research results will help educators to get a deeper understanding of students' mental health problems and provide a basis for them to make plans and decisions about college studnets' psychological educaiton.

Abstract:A hybrid algorithm based on improving QPSO and nonlinear programming is proposed to improve the speed and success rate of PID controller parameter optimization. The Cauchy random disturbance is introduced to the algorithm to increase the diversity of the particle population and enhance its global search capability. Nonlinear programming function is integrated in the algorithm to improve the capabilities of the local search, thus improving the accuracy and convergence rate. The algorithm is applied to fourth-order object of PID controller parameter optimization. Use MATLAB programming and run it, the results show the response time is short, the overshoot is small, the stability is good, and has certain practicality and promotional value.

Abstract:The purpose of this paper is to develop a suitable method for the defect region segmentation of strip surface images. Firstly, the traditional segmentation methods based on edge detection and global thresholding are studied and compared. Then, according to the gray-scale characteristics of strip surface defect images, a new segmentation method is proposed based on gray-scale morphology and adaptive thresholding. Finally, experimental results show the effectiveness of the proposed segmentation method compared with the traditional ones. Therefore, the proposed segmentation method has some practical value.