Abstract:The traffic matrix is one of the crucial inputs in many network planning and traffic engineering tasks,it is agreed that accurate traffic matrices are crucial,but it is usually impossible to directly measure traffic matrices. So,it is an important research topic to infer traffic matrix by reasonably modeling,and incorporating the measurement data of limited links,as well as other prior information.This paper presents the basic concept of traffic matrix and its estimation mechanism,categorizes and analyzes more than twenty different approaches to traffic matrix estimation problem proposed recently,and summarizes current research achievements on traffic matrix estimation problem.This paper also discusses the performance and estimation errors of current approaches. Finally,future research directions and potential applications of these researches are identified.

Abstract:This paper presents two parallel algorithms to compute reversal distance of two signed permutations on different models. These two algorithms are based on Hannenhalli and Pevzner's theory and composed of three key steps: Construct break point graph, compute the number of cycles in break point graph and compute the number of hurdles in break point graph. The first algorithm runs in O(log²n) time using O(n²) processors in SIMD-CREW model. The second one can solve the problem in O(logn) bus cycles by using O(n³) processors on the linear array with a reconfigurable pipelined bus system (LARPBS).

Abstract:It is necessary to bring global optimization in covering algorithm to improve its precision of classification.So a probabilistic model of covering algorithm is put forward in this paper.Firstly,the covering algorithm is ameliorated to kernel covering model(Gaussian function is the kernel function),then a kind of finite mixture probabilistic model for kernel covering model is introduced according to the probabilistic meaning of Gaussian function.Finally,the global optimization calculation is inducted based on maximum likelihood theory and Expectation Maximization Algorithm.Therefore,the algorithm optimizes the covering network broadens the application domain of covering algorithm and improves its robustness.The experimental results show that the optimized probabilistic model of covering algorithm can improve the accuracy of classification.

Abstract:The difficulty of Dynamic Multi-Objective Optimization(DMO)problem lies in either the objective function and constraint or the associated problem parameters variation with time.In this paper,based on the immune clonal theory,a new DMO algorithm termed as Immune Clonal Algorithm for DMO(ICADMO)is proposed.In the algorithm,the entire cloning is adopted and the clonal selection based on the Pareto-dominance is adopted.The individuals in the antibody population are divided into two parts:Dominated ones and non-dominated ones,and the non-dominated ones are selected.Three operators are introduced into ICADMO,which guarantees the diversity,the uniformity and the convergence of the obtained solutions.ICADMO is tested on four DMO test problems and compared with the Direction-Based Method(DBM),and much better performance in both the convergence and diversity of the obtained solutions is observed.

Abstract:In order to put fuzzy reasoning into the framework of logic and lays a solid logical foundation for fuzzy reasoning both syntactically and semantically,this paper transforms FMP(fuzzy modus ponens)into GMP (generalized modus ponens)by formalizing fuzzy reasoning and transplanting it into the classical propositional logic.Base on the concept of truth degrees of formulas,the sustentation degrees between formulas are put forward and a new kind of optimal solving mechanism is established for GMP and CGMP(collective generalized modus ponens).Existence theorems of optimal solutions are proved both for GMP and CGMP,and it is pointed out that there exists a completely similar reasoning mechanism between the classical propositional logic and the fuzzy logic. The graded method presented in this paper makes the algorithmic realization of solution procedure possible and serves as a guideline for the graded reasoning about knowledge.

Abstract:Existing works of reduced support vector set method find the reduced set vectors based on solving an unconstrained optimization problem with multivariables,which may suffer from numerical instability or get trapped in a local minimum.In this paper,a reduced set method relying on kernel-based clustering is presented to simplify SVM(support vector machine)solution.The method firstly organizes support vectors in clusters in feature space, and then,it finds the pre-images of the cluster centroids in feature space to construct a reduced vector set.This approach is conceptually simpler,involves only linear algebra and overcomes the difficulties existing in the former reduced set methods.Experimental results on real data sets indicate that the proposed method is effective in simplifying SVM solution while preserving machine's generalization performance.

Abstract:Presently,a whole logic system is still absent to represent and reason various kinds of preferences.In this paper,a preference logic MPL(logic of many kinds of preference)is introduced to supply the gap.In addition, recurring to the minimal/maximal specificity principle,a non-monotonic semantics of MPL's language L_MPL is defined,and the application and capability of MPL are initially investigated by rewriting preferences expressed by ranked knowledge based into L_MPL.Moreover,the conclusion and the prospective are presented in the end.

Abstract:Conventional algorithms of particle swarm optimization(PSO)are often trapped in local optima in global optimization.In this paper,following an analysis of the main causes of the premature convergence,it proposes a novel PSO algorithm,which is called InformPSO,based on the principles of adaptive diffusion and hybrid mutation.Inspired by the physics of information diffusion,a function is designed to achieve a better particle diversity,by both taking into account their distribution and the number of evolutionary generations and adjusting their"social cognitive"abilities.Based on genetic self-organization and chaos evolution,clonal selection is built into InformPSO to implement the local evolution of the best particle candidate,gBest,and make use of a Logistic sequence to control the random drift of gBest.These techniques greatly contribute to breaking away from local optima.The global convergence of the algorithm is proved using the theorem of Markov chain.Experiments on optimization of unimodal and multimodal benchmark functions show that,comparing with some other PSO variants, InformPSO converges faster,results in better optima,is more robust,and prevents more effectively the premature convergence.

Abstract:Association rule mining is one of the most important and basic technique in data mining,which has been studied extensively and has a wide range of applications.However,as traditional data mining algorithms usually only focus on analyzing data organized in single table,applying these algorithms in multi-relational data environment will result in many problems.This paper summarizes these problems,proposes a framework for the mining of multi-relational association rule,and gives a definition of the mining task.After classifying the existing work into two categories,it describes the main techniques used in several typical algorithms,and it also makes comparison and analysis among them.Finally,it points out some issues unsolved and some future further research work in this area.

Abstract:DNA sequence is one of the basic and important data among biological data.Researching DNA sequence data and then comprehending life essential is a necessary task in post-genomie era.At present,data mining technique is one of the most efficient data analysis means,which finds out information hidden in data.It has also become main data analysis technique adopted in Bioinformatics.It has been applied in DNA sequence analysis, which has got wide attention and rapid development.And considerable research achievements have emerged. Provides an overview of research progress in DNA sequence data mining field.In more detail,it proposes three research phases including statistics-based data mining methods application,general data mining methods application,and specialized DNA sequence-oriented data mining methods design,and then elaborates that sequence similarity is foundation of DNA sequence data mining technique.It also analyzes and comments some key techniques in this field by combining with biological background,such as DNA sequential pattern,association, clustering,classification and outlier mining.Finally,future work and open issues are given,including the research of a novel storage model and index methods,the design of data mining algorithm based on biological domain knowledge.

Abstract:Motivated by the multiple requirements of data sharing,privacy preserving and knowledge discovery, privacy preserving data mining(PPDM)has become the research hotspot in data mining and information security fields.Two main problem are addressed in PPDM:One is the protection of sensitive raw data;the other is the protection of sensitive knowledge contained in the data,which is also called knowledge hiding in database(KHD). This paper gives a survey on the current KHD techniques.It first introduces the background in which KHD appears. Then it mainly presents the techniques on sensitive association rule hiding and classification rule hiding.Evaluation of KHD methods is discussed after that.Finally,it points out three future research directions of KHD:Design of measure function based on target distance in data modification techniques,inverse frequent set mining in data reconstruction techniques and design of general KHD method based on data sampling.

Abstract:This paper analyzes the performance of a buffered crossbar switch under bursty traffic.It derives the saturated throughput for a buffered crossbar switch with multiple queues at each input port by the proposed analytic model.The saturation throughput sharply decreases from 1 and converges to 0.5 with the increasing of average burst length,and it approaches 1 as the number of queues per input increases.The accuracy of the theoretic analysis is also investigated by extensive simulation.Results from this paper can be used as a guidance to design optimal buffered crossbar switches.

Abstract:This paper proposes a quantitative QoS control mechanism for Web cluster based on control theory.It's featured with distributed control,effective and flexible systematical design,good extensibility,high availability and deployment convenience.Implementation and experiments validate the feasibility and effectiveness of the proposal.

Abstract:This paper addresses the issue that what is the optimal topology for the purely distributed peer-to-peer (P2P).It is usually assumed that unstructured network will form a Power-Law topology.However,a Power-Law structure is not always the best for all the applications.In this paper,the influence of the overlay topology on P2P search is firstly investigated and the precise relation among the distribution of the degree,the query rate and the success rate is attained.Then an optimal distribution model of degrees in terms of the popularity of data items is proposed to guide the overlay construction.Finally,the experimental results show that the proposed topology is effective to improve the success rate for the Random Walks.The fundamental result is significant for the optimal overlay structure and offers a new understanding of the replica deployment in unstructured P2P.

Abstract:A new edge sampling approach called the'Router's Vector-Edge-Sampling(RVES)'is presented,which is simple for PPM(probability packet marking)to be implemented and deployed.In the graph model,vertexes are denoted by network interfaces instead of routers,while edges by vector edges instead of traditional ones.With better simplicity of implementation and flexibility of policy deployment,RVES features effectiveness for Distributed Denial-of-Service(DDoS)attack reconstruction.PPM technologies based on traditional edge sampling are still applicable.Prototypes have been deployed in the Internet and experiments prove the effectiveness and feasibility of RVES.

Abstract:This paper presents a control-flow-based program behavior extended model EMPDA(extended model based on push down automaton)by adding invariance constraints to control flow model,which can describe some invariance properties while a program is running safely,and enhance the ability of intrusion detection.By distinguishing the importance of system calls according to practical applications,this paper divides the program behavior model into core model and secondary model to reduce the workload of the model and improve the learning efficiency.Experimental results show that the extended model has better performances in many aspects,such as coverage speed,false positive rate and the capability of intrusion detection.

Abstract:Analyzing the existing P2P(peer to peer)routing algorithms,Flabellate Addressable Network(FAN) routing algorithm,an efficient second-moment-based resource routing algorithm supporting multi-dimensional resource description is proposed.Peers are mapped into a multi-dimensional Cartesian space with FAN routing algorithm that manages the subspaces and searches resources based on the peers'second-moment.The routing efficiency of FAN algorithm is up to O(log(N/k)).When a peer joins and leaves the FAN network,the cost for updating routing messages is O(klog(N/k)).The experimental results show that FAN routing algorithm has advantages of high efficiency of routing and low cost of network maintenance,and is an efficient structured P2P resource routing algorithm supporting multi-dimensional resource description.Some improved routing algorithms based on CAN(content-addressable network)can also be implemented in FAN network,and they can obtain better routing efficiency and lower maintenance cost.

Abstract:The approximate invariance of the middle frequencies energy relationship between video adjacent frames under photometric distortion and spatial desynchronization is discovered by analysis.Based on this approximate invariance,a new robust video watermarking scheme is proposed.It can adjust embedding strength adaptively according to human visual system features.Experimental results show that the proposed scheme is resilient to photometric distortion,spatial desynchronization and combined distortion effectively.

Abstract:A new approach and an idea for exploration are presented to the concurrent deniable authentication based on witness-indistinguishable(WI)within the framework of universally composable(UC)security.A definition of an ideal functionality for deniable authentication is formulated.A new deniable authentication protocol is proposed based on two primitives of the verifiably smooth projective Hashing(VSPH)and non-committing encryptions(NCE).This new approach is practically relevant to VSPH based on the Decisional Diffie-Hellman (DDH)assumption and NCE based on the decisional composite residuosity(DCR)assumption.Compared with a timing constraint and public directory model,simulation of the concurrent protocols is not needed to restrict an adversary capability in a common reference string(CRS)model.The protocols are forward deniable and UC security against adaptive adversaries.Unlike previous proposals with the CCA2 public-key cryptosystem or multi-trapdoor commitments paradigm,the new paradigm leads to more efficient protocols.

Abstract:An algorithm combining source statistics and network congestion control for robust video transmission over wireless network is proposed in this paper. Based on scene modeling and characteristic analyzing, all layers generated by scalable coding are classified into several types and two queuing are made respectively according to their contribution to network congestion and to the quality of reconstructed video. Then the source rate, unequal error protection level and congestion control strategy are dynamically adjusted according to different network status in which the packet loss is caused by network congestion or by unreliable transmission in wireless channel. The simulation results show that the proposed method can achieve better results than MPEG-4 video source coding with fixed rate Turbo coding and than the one which dynamically adjusting source coding and channel coding rates and using network congestion control method of selectively dropping I, B or P packets.

Abstract:In this paper,the strength of AES-256 against the related-key impossible differential attack is examined. Firstly,a carefully chosen relation between the related keys is presented,which can be extended to 8-round(even more rounds)subkey differences.Then,a 5.5-round related-key impossible differential is constructed.Finally,an attack on 7-round AES-256 and four attacks on 8-round AES-256 are presented.

Abstract:3D shape analysis is a key problem for 3D model retrieval.In this paper,a shape signature describing the spatial distributions of 3D model's components is proposed.The primary motivation of the approach is to emphasize the object's structural attributes due to human vision theories.Firstly,a model is decomposed into several meaningful components,each of which is represented by a sub-mesh,then the component's mass center and percentage of the surface area are combined to form a component feature,and finally a set of component features satisfing the specified conditions are used as the shape signature ofa 3D model.Based the signature,an approach of 3D model retrieval is presented.It performs well despite the presence of different resolutions and connectivity,and is fast when measuring shape similarity.Experimental results demonstrate the ability of the proposed signature to retrieval 3D modes.

Abstract:In order to exchange data between PDE(partial differential equation)modeling technique and the conventional CAD(computer aided design)systems,this paper presents a novel algorithm for approximating PDE surface by tensor product Bézier surface based on constrained optimization.The algorithm is also improved by the subdivision property of tensor product Bézier surface.The computational examples and the results of error analysis demonstrate the validity of the algorithm.

Abstract:This paper proposes a method for quickly handling collision between character and garment in 3D garment simulation that aims to use in interactive real-time simulation environment.During the pre-processing stage, geometric shape of the character model is approximated with several simple colliding objects,which involve spheres and simplified convex hulls according to skinning animation technique.During the real-time simulation stage,these colliding objects driven by skeleton replace character meshes for fulfilling collision handle with cloth pieces.Moreover,in order to quickly compute collision response information,the space locality of intersection test is exploited by introducing external map mechanism.Experimental results demonstrate that the method can effectively avoid penetrations between cloth piece and character model,and significantly reduce the computational cost of collision handle.The real-time system,even with complicated garment mesh,still keeps a high simulation frame rate.

Abstract:With the concept of the packing problem,this paper presents a color image representation method based on the non-symmetry and anti-packing pattern representation model(NAM).By describing the NAM and the method of the binary-bit plane decomposition(BPD)for the color image,a novel NAM-based representation algorithm for the color image is proposed.Also,the total data amount of the algorithm is analyzed.The theoretical and experimental results presented in this paper show that the NAM-based representation method can reduce the data storage much more effectively than the hierarchical structural linear quadtree-based representation method and is a better method to represent the color image pattern.The method is a new research area with respect to the color pattern representation and is valuable for the theoretical research and potential practical values such as decreasing the storage room,increasing the transmission speed,quickening the process procedure,matching pattern,and so on.

Abstract:Video-on-Demand,which is an important application in peer-to-peer(P2P)networks,has attracted a lot of research interests recently.As P2P network is easy to deploy and can provide a large-scale substrate networks, many P2P VoD content distribution schemes have emerged to offer the basic function of data transportation.This paper presents a survey on the state-of-the-art P2P VoD content distribution schemes.It first summarizes the key issues in designing P2P VoD schemes and then categorizes these schemes into four groups according to data transportation method.Finally,it discusses the application-level performance and outlines possible future work.

Abstract:Multi-Projector tiled display wall is thought to be an effective technique to tackle the conflict between the increasing demands for super-resolution display and the resolution limitations of a single display equipment,but currently there lack of normal forms and normalized methods to support screen calibration with a high precision and high reliability.This paper tries to solve this problem with a new color calibration scheme.The main problems in screen calibration and the existing solutions are introduced first.Then the screen calibration process and geometry calibration methods are listed.After that,evaluation criterions for screen calibration are discussed,and a generalized color model for projectors color calibration and a visual seamlessness algorithm are presented in detail. Only employing a digital camera,the new scheme can build tiled display walls with high-precision visual seamlessness,which results in higher construction efficiency and lower maintenance cost.The methods have been verified and applied in a variety of tiled display wall,and have shown great theoretical and practical significance for building and maintain tiled display walls.