Abstract:Efforts of this paper focus on the issues about the management and throughput of firewalls (or screening routers) applied in transit networks. On the one hand, manual configuration of large amount of firewalls distributed in many access points cannot meet the requirements of security management in the open and dynamic environment. On the other hand, the ordinal lookup of filtering rules in firewall results in decrease of throughput. Aimed at a typical transit network and its security policy requirements, a policy-based access control framework (PACF) is proposed in this paper. This framework is based on three levels of abstract access control policy: organizational access control policy (OACP), global access control policy (GACP) and local access control policy (LACP). The GACP, which comes from the results of IDS and search engines according to OACP, is automatically and dynamically distributed to firewalls as LACPs. Each LACP is then enforced by an individual firewall. Some algorithms for distribution of GACP and enforcement of LACP are described. A hashbased algorithm is proposed for lookup of filtering rules in LACP. PACF largely reduces the management labor of the security administrator for large transit networks. Under the environment with policy requirements described in this paper, the new algorithm reduces the time complexity of lookup from O(N) of traditional sequential algorithm to O(1), which increases largely the throughput of firewalls.

Abstract:In this paper the searching problem for an object in a set of N locations is considered. The goal is to allocate the resources so as to maximize the probability of locating the object. By using Lagrangian operator method the problem of optimal search with the unknown target distribution is studied. Some selection criteria and error estimate results are derived.

Abstract:A hierarchical automatic placement algorithm for mixed mode placement problem is presented. The so-called mixed mode is a combination of standard cell and macro block. The presented algorithm completes the placement in both block level and cell level. In block level, the random cells are firstly partitioned into soft blocks, then SP (sequence pair) based method is used to do block placement. In cell level, firstly, quadratic placement method is used to do inner placement within each soft block, then a placement improvement routine is done to the whole chip, and at last, a combined min-cut and enumeration based detailed placement procedure completes the final placement. The algorithm is tested on a set of circuits with different number of standard cells and macro blocks, and obtains satisfactory results.

Abstract:Due to the poor quality of ultrasound images, fully automatic segmentation methods are not feasible. This article describes a novel approach to the semiautomatic segmentation of ultrasound images.Although user interaction is not required much, it is used as an important factor and incorporated into the traditional B-spline snake models. The modified model is called an interactive B-snake model because the movement of the active contour is constrained through user interaction. By introducing a set of moving rules, B spline segments are moved to the desired boundary directly. The statistical models are trained on-the-fly by observing boundaries accepted by the user. The resulting algorithm is especially useful when dealing with successive slices and provides fast, reliable and verifiable segmentation in ultrasound images. The algorithm has been used successfully on the Liver Tumor Surgical Simulation System.

Abstract:Security protocols play more and more important role with the bursting development of Internet. Formal authentication is an effective way to detect flaws in those protocols. There are many automatic checkers for verifying protocols, but few can deal with digital signature protocols. This paper proposes a new tool for formal automatic verifying. It can have an effective check for digital signature protocols. It is based on intruder's role impersonate and backward deduction. It was implemented by using JAVA.

Abstract:The computational time complexity is an important topic in the theory of evolutionary algorithms. This paper introduces drift analysis into analysing the average time complexity of evolutionary algorithms, which are applicable to a wide range of evolutionary algorithms and many problems. Based on the drift analysis, some useful drift conditions to determine the time complexity of evolutionary algorithms are studied. These conditions are applied into the fully deceptive problem to verify their efficiency.

Abstract:In this paper, a face detection algorithm for color images based on skin color and template, which is composed of skin color segmentation, template matching, and neural network verifying is presented. First, a skin color model in HSI chrominance space is used for segmenting regions in which may have faces, and then the average-face based template matching and neural network verifying methods are used for searching faces in those regions. This algorithm integrates skin color information in color images with template matching and neural network classification model in gray level images, which results in not only faster speed, but also higher robustness of the algorithm. Experimental results demonstrate the efficiency and feasibility of this algorithm.

Abstract:The Chi calculus is obtained from the π calculus by unifying two classes of restricted names. The polyadic Chi calculus extends Chi calculus in that more than one pieces of information can be passed around in a communication. And the atomic Chi calculus is the subcalculus of the polyadic Chi calculus by removing the prefix combinatory. In this paper, the bisimulation equivalences for the atomic Chi calculus are investigated. The main result of this paper is that, in a certain sense, there is only one bisimulation equivalence on the atomic Chi processes.

Abstract:As the network applications develop, routers must support those functions such as firewalls, provision of QoS and traffic billing etc. All these functions need classification of IP packets, according to which it is determined how different packets are processed subsequently. In this paper, a new IP classification algorithm is proposed based on the Grid of Tries algorithm. The new algorithm not only eliminates original limitations in the case of multiple fields but also shows better performance in regard to both time and space. It has better overall performance than many other algorithms.

Abstract:Program slicing is an important method for debugging, testing, analyzing, understanding and maintaining object-oriented software. In this paper, a new hierarchy slicing approach is presented so as to replace old slicing methods based on SDG (system dependence graph). The main idea of hierarchy slice and algorithm are discussed in this paper, and the application of hierarchy slicing algorithm is also presented at the end of this paper.

Abstract:The classification algorithm based on SVM (support vector machine) attracts more attention from researchers due to its perfect theoretical properties and good empirical results. In this paper, the properties of SV set are analyzed thoroughly, and a new learning method is introdnced to extend the SVM Classification algorithm to incremental learning area. After that, a new improved incremental SVM learning algorithm is proposed, which is based on a sifting factor. This algorithm accumulates distribution knowledge of the training sample while the incremental training is proceeded, and thus makes it possible to discard samples optimally. The theoretical analysis and experimental results show that this algorithm could not only improve the training speed, but also reduce storage cost.

Abstract:MAS (multi-agent system) is a sociality which is composed of many autonomy agents. Under the social background, realizing the goal of MAS depend on social rationality of agent. Although Jennings has presented a mode of social rationality, he doesn't obtain the method of computing socially rationality. In this paper, a new computed framework of social rationality is constructed based on the principle of synergetic and point the physical significance of parameters. The authors shell parameter's computing divide into four steps; approximate liner, define order parameter; built differential questions and solve to it.

Abstract:Cooperative design is a complex group activity that involves participants with heterogeneous skill. In this paper, a cooperative design approach in a multi-agent design system is introduced and a form of Bulletin Board that supports dynamic task assignation and a kind of conciliatory approach when conflict occurs in design are presented. According to autonomous and cooperative features, the form of Bulletin Board combines the advantage of black board and contract-net and introduces a style of distributed and centralized combination, weakens some failings of black board and contract-net during dynamic assignation in distributed environment. Then, it puts forward a conflict pre-checking and conflict processing approach. This approach aims at finding and solving conflict during the early design phase and decreasing resource waste.

Abstract:The software crisis hasn't been handled completely after its appearing 40 years ago. Many scientists and enterpriseries consider that the lack of the formal, effective technology and method for software quality management is the primary reason. The total quality management has become an integrated theory with the developing of the socialization production. Many international standard and industry standard has been made. In this paper, the state of Chinese software industry is analyzed and the modeling of software quality management and its realize framework matched national requirment are built.

Abstract:In this paper, a sufficient condition and a necessary condition on liveness for Weighted Asymmetric Choice (WAC) net system which is an important subclass of Petri net system are presented. Moreover, a judgement condition for live and bound WAC net system is given, and the sufficient and necessary condition for WAC net liveness monotonicity is shown.

Abstract:Object-Oriented program slicing technique is widely used in program analysis, program comprehension, software testing, and software maintenance etc. In this paper, the basic concepts of abstract data slice and intra-class slice are created at first. Then, the cohesion existing in JAVA language is discussed based on these slices. Finally, the data cohesion, method cohesion and class cohesion are presented and solved by analyzing the relationship between these slice and data, method or class.

Abstract:In this paper, the spatial data models and data structures in geographic information system (GIS) are first discussed. Then the topological relations about point-point, point-line and point-area of spatial objects are qualitatively analyzed. The reasoning algorithms of topological relations of spatial objects are also given. Finally, the fuzzy techniques are applied in reasoning algorithms of topological relations of spatial objects in order to solve fuzziness and uncertainty in the GIS easily.

Abstract:As a standard object-oriented modeling language, UKL (unified modeling language) is lack of formal dynamic semantics. In this paper, an operational semantics for UML statechart diagrams is proposed according to UML1.1 specification documents. The formalization is able to deal with most of the features of UML statechart diagrams, which sets the basis for code generation, simulation and test cases generation. It has been used to generate test cases and to simulate the test process on platform of Rose98.

Abstract:In this paper, a quickly collision detection and amend arithmetic is proposed. In detection phase, bounding box and hierarchical structure model are combined to cull out the majority of detection. And many collision types are reduced to vertex/plane collision, so solving linear equation can be simplified to algebra operation. In amend phase, amend are combined into numerical integration by using preconditioned gradient method. This speeds up the whole calculation.