In order to enhance reliability of X processor and save its power and chip pad, this paper designs a general Serial Peripheral Interface(SPI) Flash controller, which takes SPI Flash as program-loading memory. The component structure and realization method are given. Assert checks of Property Specification Language(PSL) are used to verify functions of the controller to decrease verification complexity and enhance verification quality and speed. Experimental result shows that function coverage of the controller obtains 100%.
In the context of semantic Web, knowledge bases sometimes are single and dispersed. As a result, it hampers the development of semantic Web. A method of integration of multiple knowledge bases based on minimal concept sets is proposed. The minimal concept sets are defined. The method of generating minimal concept sets is given and mapping’s strategies in ontology mapping model based on risk minimization are improved and integration algorithm of multiple knowledge bases on basis of minimal concept set is designed. The complexity of the algorithm is proved by an application example.
Based on the characteristics analysis of Anti-Radio Missile(ARM) echo wave, three parameters Chirplet Transform(CT) is proposed, choosing three parameters of five in CT. A new method of ARM detection technique is also proposed associated with adaptive linear prediction filtering algorithm. Implemented by FFT, it can narrow the searching range and reduce the requirement of input SNR effectively. Simulation results show that ARM can be detected fast and precisely in this way, in the surrounding of low SNR and high aircraft echo interference, so that a reliable early warning of ARM launching can be given.
This paper presents a Quasi-Birth-and-Death(QBD) model for damaged nodes in Ad hoc network. Since the communication in Ad hoc network is distributed, the destruction of nodes in the route between transmitter and receiver will break off the communication. Too many damaged nodes will result in some nodes unreachable. By assuming damaged vehicles’ arrival be subject to Poisson process, repair time and vacation time be exponentially distributed, it creats QBD model for Ad hoc network nodes damage and fix process, uses matrix analysis method to get stationary distribution, and then obtains damage nodes average queue length, the probability of no need to wait and so on. It is helpful on saving resources while keeping the network stability, can be a reference when make maintenance strategy.
In the Wireless Sensor Network(WSN) where nodes are randomly deployed, the proposed sensor scheduling schemes can not guarantee uniformly distributed and energy balanced at the same time. Aiming at this problem, a distributed Energy-Balanced Location-independent Coverage Protocol(EBLCP) is proposed. Based on virtual coordinate, EBLCP creates temp-sets in which working nodes are elected by residual energy of some neighboring nodes and its own. To compare with NSVC and DELIC, the simulation and analysis study demonstrate that the EBLCP not only provides high rate of area coverage, but also prolongs network lifetime to a large extent.
When handover bursts out, the generation of Care-of Address(CoA) of mobile routers, especially nested ones, and the duplicate address detection followed are the most time-consuming parts compared with the other processes. In this paper, after reviewing the existing schemes in the eye of delay related to CoA, it is based on access router tree model and Hash function, gives a novel scheme of CoA configuration and corresponding handover algorithm, to reduce the time cost and the signal overhead. Performance evaluation indicates that, compared with the existing solutions, lower delay can be achieved in CoA configuration of nested mobile routers and also in handover.
By analyzing the fundamental working tenet concerning the flow control agreement of the Fibre Channel(FC), this paper proposes an OPNET-based simulation model for controlling the flow of FC. The design program models the process nodes of sending and receiving. Simulation results show that the model proposed can abide by the flow control agreement of FC precisely with the advantages of simplicity, straghtforwardness and extensibility under the environment of OPNET and using C programming language as the development tool.
The defending ability of Dynamic Differential Cascade Voltage Switch Logic(DDCVSL) and SABL is restricted by layout and routing when they are used for protecting the circuit against differential electromagnetic analysis. Aiming at this problem, this paper gives an improved defend logic Masked Dynamic Differential Cascade Voltage Switch Logic(MDDCVSL) based on DDCVSL through introducing masking mechanism. NAND gates based on MDDCVSL, DDCVSL and SABL are designed with TSMC 0.18 μm technology. Experimental result demonstrates that the improved defend logic get rid of the electromagnetic information leakage and solve the problem that defend ability of MDDCVSL, DDCVSL and SABL are restricted by layout and routing. Meanwhile, MDDCVSL has less power consumption, area and delay than WDDL, MDPL.
To improve the data processing speed of intrusion detection system and reduce the data imbalance, this paper presents a Fast Hierarchical Nearest Neighbor(FHNN) resampling method by analyzing existing resampling methods commonly and according to the characteristics of network intrusion detection datasets. The novel algorithm experiments with KDD’99 datasets. Experimental results show that this method is efficient in tacking noise and majority class examples. It is a rapid and effective method for detecting many types of threats from massive audit data.
In order to protect the security of a digital image effectively, this paper proposes an image encryption algorithm based on wavelet expansion function and hyperchaotic system. The pixel locations of the image are scrambled with a wavelet function, and the relationship between the original image and the encrypted image is disrupted with hyper-chaos system. The many parameters are inserted in four-step Runge-Kutta method in which chaos sequence of hyper-chaos system is gotten, to expand the space of parameter. Simulation experimental result show that the random behavior of gray value distribution of the encrypted image is better and the method can defends attacks.
Aiming at the disadvantage of unsupervised method and supervised method, a linear dimensionality reduction method called Semi-supervised Local Discriminant Analysis(SLDA) is proposed. When there is no sufficient training sample, local structure is generally more important than global structure. SLDA utilizes the labeled data points to infer the local discriminant structure, as well as the intrinsic geometrical structure inferred from both labeled and unlabeled data points at each local area. Experimental results on ORL and Yale face recognition demonstrate the effectiveness of the algorithm.
In order to enhance the ensemble of the traditional AdaBoost algorithm and reduce its complexity, two improved AdaBoost algorithms are proposed, which are based on the correlation of classifiers. In the algorithm, Q-statistic is added in the training weak classifiers. Every weak classifier is related not only to the current classifier, but also to the previous classifiers as well, which can effectively reduce the weak classifier similarity. Simulations show that the algorithm is of better detection rate and lower false alarm rate.
Particle Swarm Optimization(PSO) algorithm is proposed to solve the problem of mining enterprises dynamic allocation. The mineable ore blocks can be marked according to mining conditions delineated, with a bit of ore particles to represent a mine block, and the selected ore mining block is represented by 1, and the constrained PSO particles computing and “flight” rule is re-defined and the dynamic allocation ore particle swarm algorithm is realized. The PSO algorithm implementation is simple, the optimum effect is obvious, the comparison of actual production in 2009 and the optimization results shows that the algorithm improves enterprise efficiency significantly when the cost of production is almost unchanged.
Manually creating multiple labels for each sample is very important but it is time-consuming. Manually creating multiple labels for each sample may become impractical when a very large amount of data is needed for training multi-label classifier. To minimize the human-labeling efforts, this paper proposes a weighted decision approach, the approach considers quantity and confidence of training samples, it can make the classifier need fewer samples, but achieve a comparative precision.
The construction of scenario map is one of key activities in the process of chance discovery. The formalization description of the structure of scenario map in chance discovery is presented. With this formalization description, the principle for the construction of event clusters in scenario map via clustering analysis based on Boolean correlation matrix and the construction of simple scenario map based on event clusters are given and the algorithm for the implementation of the principle is given. The evaluation criterion for the performance of methods for the construction of scenario map in chance discovery is discussed and efficiency coefficient is used as the criterion definitely. Experimental results show that the performance of the algorithm on the construction of simple scenario map is good with greater efficiency coefficient.
The performance of semi-supervised learning algorithm can be enhanced by incorporating the structural information of the dataset. Based on this assumption, a novel graph-based semi-supervised learning method is proposed. A local scale scheme is presented to define different scale parameters for edge weights in different density regions. The dataset is classified by a semi-supervised learning algorithm named graph Laplacian kernels. Experiments on several datasets show that the proposed method outperforms other kernel-based semi-supervised learning algorithms.
This paper presents a kind of parking cell detection method based on KL transform and kernel fisher discriminant. It preprocesses the parking cell images. The parking cell images is projected onto the eigen-parking cell-subspace which is constructed by KL transform, and the coefficients of the projection are the eigen-vector of the parking cell. It makes use of kernel fisher discriminant to determine the occupancy of parking cell. In the experiment, it uses three different kernel functions to compare the results of the fisher discriminant. Experimental results indicate that the parking cell detection method with Radial Basis Funtion(RBF) kernel function is superior to other methods, the correct detection rate reaches as high as 97.6%.
Aiming at the problem of fragment matching in 3D fragment reassembly, this paper proposes a rapid contour curve matching algorithm. In the algorithm, the curve is expressed by B-spline curve, and curvature, torsion and normal vector are calculated. In the matching process, the type of feature point is classified according to its neighbor surface, and the similarity of the two feature section is measured according to the Euclidean distance of the feature points. The normal vector is used to verify the matching probability of similar space curves. Experimental results show that the algorithm is stable and efficient.
This paper proposes a fuzzy-cut best wavelet packet image compression algorithm. The ratio of each high-frequency coefficient to the total high-frequency coefficients of wavelet packet decomposition and the ratio of each node’s children entropy to the original node entropy are defined as two inputs of the fuzzy system. The wavelet packet coefficients are divided into cutting part and the best wavelet packet is got by the reservations parts. The image is compressed by using the different thresholds on different wavelet packet transform level based on the best wavelet packet. Simulation results show that the algorithm has better performance than the rate-distortion and the whole thresholds wavelet packet image compression algorithm.
In order to meet high reliability and real-time requirement of embedded systems, this paper proposes a multi-task parallel IEEE1394 protocol stack model by using the hardware and software feature of IEEE1394 bus. And the data transmission of different quality of services is acquired by setting different tasks’ priority. The execution of task and command of IEEE1394 subsystem take on higher real-time performance. The feasibility is testified by implementing digital video decoding and Serial Bus Protocol 2(SBP2) mobile storage based on the protocol stack on VxWorks platform.
For solving the vehicle routing problem with disruption that may be vehicle breakdowns or traffic accidents in the logistics distribution system, this paper builds a mixed integer programming disruption management model. For the attributes of the vehicle routing problem with disruption, a series of solving-simplify strategies are given to simplify the solution space. On the basis of the characteristic of the model, the improved saving algorithm based on disturbed value is designed based on the disruption value. Numerical experiments from multiple wheel-driven vehicle scheduling problem and vehicle scheduling interference management problem are given to prove the efficiently of the disruption management model and algorithm.
Based on the modularized and parametric design, a model of module library management system is presented considering the special demand of collaborative sharing mode for supporting networked rapid design of automatic weapons. The design procedure for 3D models of rapid design platform of automatic weapons is described. In order to transfer data accurately and effectively, the multi-thread transferring method of models and compression program based on hierarchy of assemblies are put forward and realized ultimately.
This paper presents a simulation method for Taihu Lake blue-green algae outbreak process based on WSN and particle system theory. It adopts the technology of embedded system and designs an online measurement system based on temperature and transparency. According to the established mathematical models, it carries on the simulation of bloom in Taihu Lake with OpenGL. Practice result proves that the method which satisfies the general requirements of real-time display under general PC has good simulation performance for the outbreak process of algae.