In order to improve processor core simulation modeling efficiency, a virtual processor modeling method based on SimpleScalar architecture is proposed, and the model aiming at Godson-1 processor reaches 1 000 000 per second with average IPC error of 2.3%. A controllable random event Bus Function Model(BFM) is presented, providing active stimuli generation and on-chip bus verification function for SoC design. Experimental result proves that the solution has broad applicability in processor core modeling and can be seamlessly integrated into mainstream SoC flow.
This paper obtains an exponential stability criterion of Networked Control System(NCS) by constructing a delayed difference inequality. On the basis of this new established criterion, it designs a feedback controller in the form of Linear Matrix Inequality(LMI). It does not contain any free-weighting matrices, and can be solved by LMI toolbox in Matlab software. Numerical value simulation result shows that this design has less computational complexity and conservatism, and it is convenient to implement.
According to the properties of Artifact, semi-structured data is used to represent it. Taking advantage of Predict/Transition Petri-Net(Pr/T-Net), a new method is proposed to formalize the business process model, in which the static predict is the set of Document Type Definition(DTD) which gives Artifact structure and the transition is a kind of first-order logic language which operates the XML. The arrival, persistence and uniqueness of the Artifact are analyzed.
This paper presents a Reverse Nearest Neighbor Queries(RNNQ) algorithm based on Auto-Selection Grid Index(ASGI). ASGI divides the space into an equal size network, and each cell stores moving objects as a bucket. RNNQ algorithm uses the strategy of the bucket’s object number and the grid geometry characteristic to reduce the access node the RNNQ must visit. ASGI will repeat the previous action to decrease the cost again if objects around the query point are too many. Experimental results show that RNNQ algorithm based on ASGI has good query performance and outperforms RNNQ algorithm based on TPR tree.
Aiming at the problem of the inaccurate information in the monitoring based on the current middleware load balancing mechanism, and of the large size of the particle, this paper designs a more precise, more fine-grained load balancing system and its associated load balancing mechanism with the combination of fuzzy logic technology. Record list of load information records load state information, and load transfer list records relationship between service nodes and service requesters. Experimental result shows that the load balancing mechanism can effectively target node overload to better achieve load balancing system services of the cluster.
Aiming at the problem of constraint on network protocols or incurring embedding error in the existing methods, a scalable proximity search method based on cover tree CPS is raised. It is comprised of two parts: cover tree constructing protocol and maintaining protocol, and k nearest neighbors searching algorithm. Attending nodes compute the layers of themselves, so as to construct a layered tree. The neighbor maintaining protocol is responsible for maintaining the structure of cover tree, ensuring the adaptability in the network environment. k nearest neighbors searching algorithm constructs candidate nodes for every level by pruning the cover tree in order to improve the searching efficiency. Simulation results indicate that the accuracy of CPS is better than Tiers.
In the process of integrating “3S”, the choice of a favorable data structure is the key point. This paper analyzes the advantages and disadvantages of both vector and raster data structures, and builds an optimal data structure based on the secondary dividing strategy, which merges the advantage of simultaneity vector and raster and can fulfill the requirement of vector data accuracy. It gives a logical representation of this data structure, and rasterizes the space data simultaneity. Based on that, it discusses the strategy of the collection and overlay analysis of space data based on this data structure.
Greedy linear shift balancing algorithm is presented for ring or linear array to reduce balancing time and the amount of the transferred load. The algorithm can be used for balancing any system in which internetwork contains at least one Hamilton path. It need not a large amount of load transferred among nodes or a large amount of the consumed time. In addition, two-stage greedy linear shift balancing algorithm is presented to improve the greedy linear shift balancing algorithm for mesh etc. Experimental results show that the two-stage linear shift balancing algorithm can reduce execution time when the balancing condition is weaken.
This paper proposes a novel Fair Aggregator based on Flows Isolation(FAFI) that combines both Earliest Deadline First(EDF) scheduling and Worst-Case Fair Weighted Fair Queuing(WF2Q). By assembling individual flows in EDF and forwarding aggregated traffic in WF2Q, FAFI overcomes disadvantage in flow aggregating that the end-to-end delay of an individual flow is related to the burstiness of other flows sharing its aggregated traffic. An end-to-end delay bound is derived in the network equipped with proposed aggregator. A performance comparison is done with previous works on the aspects of end-to-end delay bound, network utilization, and scalability.
This paper proposes a threshold multi-secret sharing scheme based on key agreement to assure that secret data can not restored by attackers have diddled subsecrets from all the participants. The cheating of dealer and the cheating of participants can be detected. The subsecrets are chosen by the participants themselves and can be used repeatedly. System does not need a security channel. It is easy for the participants to join or leave system dynamically, and all the secrets can be recovered in one time.
This paper presents a forgery attack on the threshold signature scheme proposed by Zhang Youyi et al that can resist conspiracy attack. Everyone can generate a valid threshold signature, which proves the scheme is not secure. So an improved scheme is proposed to overcome the weakness of the signature scheme, and its validity and security are proved.
The researches on combination of identity, threshold proxy signature and blind signature are rare presently. This paper proposes an ID-based threshold proxy blind signature scheme by combining proxy signatures, threshold signatures, blind signatures and using bilinear pairing on elliptic curve. It solves the problem that potency is too centralized in proxy signature, and realizes the anonymity by blind signture. Analysis shows that the scheme is easy to carry out and has high robustness, and it is suitable for application in practice.
To realize multi-sensor tracking for maneuvering target state, this paper presents a multi-sensor fusion algorithm based on DSmT and particle filtering. On the basis of observation date which is delivered by multi-sensor and filtered by particle filters. It chooses DSmT as the combining tool. The observation data is transformed into the elements and masses of the frame of discernment, and gets the finally result. Experimental results show that this algorithm can reduce distance error and improve tracking accuracy, and the appropriate computational complexity can satisfy the demand of fusing on-line.
In order to avoid the obstacles in the surrounding environment, some changes of formation style should be applied on the traditional-controlled formation. Those changes of formation can not satisfy those situations in which high degree of accuracy is required, like joint transportation. This paper introduces a strategy of rigid-formation control which regards the whole formation as a virtual unity. Theoretical analysis and experimental results show that this strategy can successfully smooth the track during avoiding obstacles, increase the degree of formation-maintain accuracy, and solve the contradiction between avoiding obstacles and keeping formation.
This paper analyzes Chinese character structure feature and evolution, combines the character recognition process together on the basis of Chinese character database and changing in recognition target, researches and comprises the advantages, disadvantages and the application range of five different kinds of Classification recognition method such as Bayesian method, quadratic discriminate function approach, Markov model, neural network classifier and Support Vector Machine(SVM). Results shows that Chinese character recognition method including recognition should be improved on the aspects of target, feature extracting and train set preparing.
This paper proposes a new approach to improve collaborative filtering recommendation. It applies the attitude-behavior relationship theory in psychology to predict the users’ rating using the Web usage data and adopts the predicted rating in collaborative filtering algorithm to generate recommendations for users. It verifies the prediction model, and compares the method to traditional collaborative filtering algorithm which only records the users’ clicks as their interests. Experimental results demonstrate that the method is more effective at recommending items for users than the traditional algorithm.
Basedon the directed acyclic graph structure of gene ontology, this paper proposes a novel method to measure the semantic similarity of gene ontology terms. By calculating the common ancestors of two terms and the disjunctive ancestors according with the condition, the average information content of disjunctive ancestors and the average information content of the two terms are obtained. The semantic similarity of two terms is the ratio of the average information content of disjunctive ancestors to the average information content of two terms. Experimental results show that the method has high accuracy.
Some models of events in the campaign process of multichannel surface to air missile weapon system are built, and based on them, the campaign simulation system of multichannel surface to air missile weapon system is founded with Arena which is a visual environment for modeling and simulation. This simulation system is applied to evaluate the multichannel surface to air missile weapon system’s capability of anti-airplane saturation attack. The result proves that the simulation is valid.
Automatic parallelization has limitations on the parallelizing capabilities and the scope of applications, while the interactive parallelization method makes up the deficiencies. This paper presents an interactive parallelization system named ZJU Interactive Parallelizing System(ZIPS), and describes the mechanism in detail. The system uses a computation-driven algorithm as the basis, obtains the program information through the powerful interaction interface, combines the technology of automatic parallelization, and finishes the source to source transformation for two different computations. Experiments show that the method can achieve good performance.
The wireless netowrk position technologies based on ZigBee are developing rapidly in recent years, but in pratical application their precisions are also greatly affected by signal fluctuations caused by the multipath effect, human interference and so on. For the signal fluctuations in pratical application, the paper uses a vibration sersor to judge the motion state of position terminal, and estimates the motion path and direction by stable position fingerprints, finally searches for the nearest neighbors in the neighborhood of the estimated path, and gets the position result. Experimental results show the method weakens the affection caused by signal fluctuations, the precision statistics is about 2 m to 6 m.
Based on the space-time eigenstructure of signals, an improved space-time Direction Of Arrival(DOA) estimation matrix method is proposed to estimate the high-resolution 2-D DOA of narrowband signals. The method obtains two DOA matrices through creating another sub array, which is the primary sub array shifted down X axes. Because these two matrices are different in angle ambiguity curves, the method can get accurate DOA; therefore advantages of the traditional space-time DOA matrix method, such as automatic parameter alignment and no need of 2-D search, are still available. Simulation results show that the method is effective.
In allusion to the storm surge system’s requirement: it is necessary to figure out the shortest path from the disaster area to multi-settlements, then to evacuate the victims based on the number of disaster victims and the capacity of the settlements, this paper presents two improvements of Dijkstra algorithm, including sorting the shortest paths and evacuating to multitargets. An emulation is carried out based on Matlab, and it turns out that the improvement of the algorithm is effective and feasible, which can reduce the times of figuring out the path of evacuating to multitargets from several to one.
Focusing on the problem of effectively organize and retrieve the distributed massive remote sensing images, this paper presents a well extensible remote sensing catalogue data model based on the remote metadata standard draft which is proposed by the Federal Geographical Data Committee(FGDC). A distributed remote sensing images retrieval algorithm is designed based on the proposed catalogue data model. It accomplishes a remote sensing data catalogue and retrieval system which is successfully applied to a massive data management system.