Abstract:Model transformations are heavily used in model evolution,refinement and refactorings. Model transformations are carried out against certain constraints to preserve certain properties of the models. During model evolution,model transformations should preserve system interfaces; during model refactoring,model transformations should preserve system behaviors. In order to prove that a software transformation satisfies transformation constraints,constraints should be formalized first. And in order to automate the proof,the process of the proof should be universal to be supported by algorithms. This paper proposes an approach for formalizing transformation constraints with graph productions. With the formalized constraints and software transformation rules,an algorithm is also proposed based on critical pair analyzing technologies to automatically prove whether a transformation rule satisfies a transformation constraint or not. The proposed approach is validated with a motivating example used throughout the paper.

Abstract:In collaborative software development, WinWin state is hard to achieve because stakeholders are concerned about different aspects of software development such as technology, human, and process. Frequently the influences among the win conditions of stakeholders are implicit thus the conflicts are hard to find. This paper uses a tri-dimensional requirements model, named TRISO-RM, to describe the win conditions of stakeholders on different aspects of software development. TRISO-Elements, each of which is formed by interconnected actors, an artifact, and an activity, are used as the medium to find, build and maintain the relationships among stakeholder win conditions. The production mechanism of model clashes is discussed and the process to find and avoid them is presented based on TRISO-RM. In particular, the application in the development of SoftPM is described to illustrate and validate the TRISO-RM method.

Abstract:Dynamic compilation is an effective optimization,but the overhead is too heavy,the accuracy of information is not enough,and the redundant code is growing rapidly in the current information collection and continuous monitoring. In this paper,the dynamic compiler is designed based on online feedback and continuous monitoring in Intel ORP (open runtime platform). To solve these problems in information collection,the program instrumentation technique is improved. The continuous monitoring for the type of virtual method’s receiver object is implemented. Based on the information from online feedback and continuous monitor,the dynamic inlining optimization is invoked by the compiler. The dynamic unloading algorithm that releases the redundant code in the dynamic optimization is introduced. The results of SpecJVM98 and Java Grande Forum Benchmark show that the performance of JVM is improved on average and the system load is reduced by the dynamic unloading algorithm.

Abstract:Recommendation system is one of the most important technologies applied in e-commerce. Similarity measuring method is fundamental to collaborative filtering algorithm,and traditional methods are inefficient especially when the user rating data are extremely sparse. Based on the outstanding characteristics of Cloud Model on the process of transforming a qualitative concept to a set of quantitative numerical values,a novel similarity measuring method,namely the likeness comparing method based on cloud model (LICM) is proposed in this paper. LICM compares the similarity of two users on knowledge level,which can overcome the drawback of attributes’ strictly matching. This work analysis traditional methods throughly and puts forward a novel collaborative filtering algorithm,which is based on the LICM method. Experiments on typical data set show the excellent performance of the present collaborative filtering algorithm based on LICM,even with extremely sparsity of data.

Abstract:Clustering has been traditionally viewed as an unsupervised method for data analysis. In real world application,however,some background prior knowledge can be easily obtained,such as pairwise constraints. It has been demonstrated that constraints can improve clustering performance. In this paper,the drawback of only incorporating pairwise constraints in clustering is firstly analyzed,and then an inherent prior knowledge in data sets,namely space consistency prior knowledge is exploited. The method of utilizing space consistency prior knowledge is also given. Incorporating the two types of prior knowledge into original spectral clustering,a density-sensitive semi-supervised spectral clustering algorithm (DS-SSC) is proposed. Experimental results on UCI (University of California Irvine) benchmark data,USPS (United States Postal Service) handwritten digits and text data from TREC (Text REtrieval Conference) show that the two types of prior knowledge can supplement each other in clustering process,leading to substantial performance enhancement of DS-SSC over other semi-supervised clustering methods which only incorporate pairwise constraints.

Abstract:The tracking of deformable hand gesture is a very important task in vision-based HCI (human-computer interaction) research. A novel real-time tracking approach is proposed to capture the motion of deformable hand gesture with single camera. The proposed approach uses a set of 2D hand models in place of high-dimensional 3D model. It achieves auto-initialization by firstly using Bayesian classifier to do posture recognition,and then locating fingers and fingertips to fit image features to recognized posture. It solves the problem of interference among fingers during tracking successfully by the integration of K-means clustering and particle filter. Moreover,a state checking process is embedded into tracking method,and it realizes resumption from tracking failure and update of hand models automatically. Experimental results show that the proposed method can achieve continuous real-time tracking of deformable hand gesture with high precision,and thus it can meet the requirements from real-time vision-based human-computer interaction.

Abstract:Super-Resolution (SR) reconstruction has been a very hot research topic currently. A kind of generalized MRF (GMRF,generalized Markov random field) models is firstly proposed based on the recently reported bilateral filtering. The GMRF model is not only edge-preserving and robust to noise,inherited directly from the bilateral filtering,but also connects the bilateral filtering with the Bayesian MAP (maximum a posterior) approaches much concisely. Meanwhile,an improved numerical scheme of anisotropic diffusion PDE’s (partial differential equation) is deduced based on the GMRF model. In the MRF-MAP framework,a new SR restoration algorithm is subsequently proposed for both cases of Gaussian noise and impulse noise,utilizing the generalized Huber-MRF model which guarantees strictly global convergence. The half-quadratic regularization approach and steepest descent are exploited to solve the energy functional. Experimental results demonstrate the effectiveness of this approach,both in the visual effect and the PSNR value.

Abstract:This paper proposes a SVC (support vector clustering) based fusion rule according to unsupervised learning strategy. By employing the rule in multifocus image fusion applications,it solves the problems of region overlapping and abrupt transition brought about by the SVM (support vector machine) based fusion rule. The quality of the fused image is further enhanced. The undecimated discrete wavelet transform is applied to source images for multiresolution decomposition. Image feature data is extracted by means of grid,and it is then fed into the SVC algorithm which will generate distinct clusters. These resultant clusters are further processed by the domain discrimination algorithm and eventually distributed to two separate domains defined as complementary domain and redundant domain,in which choose-max method and weighted average method are used respectively to produce multiresolution representation of the fused image. Finally,the fused image is reconstructed by performing the corresponding inverse wavelet transform. The relation between the parameter q of SVC algorithm and the parameter RMSE used to evaluate the fused image is studied in detail. It is indicated by theoretical analysis and experimental results that SVC is appropriate for image fusion. Moreover,comparative studies show that the proposed SVC based fusion rule outperforms the existing SVM based ones.

Abstract:This paper is mainly concerned with the relation-algebraic aspects of the well-known Region Connection Calculus (RCC).It is shown that the complemented closed disk algebra is a representation for the relation algebra determined by the RCC11 table,which was first described by Düntsch.The domain of this algebra contains two classes of regions,the closed disks and closures of their complements in the real plane,and the contact relation is the standard Whiteheadean contact(i.e.aCb iff a∩b≠? .

Abstract:With the successful development of frequent item set and frequent sequence mining,the technology of data mining is natural to extend its way to solve the problem of structural pattern mining—Frequent subgraph mining. Frequent patterns are meaningful in many applications such as chemistry,biology,computer networks,and World-Wide Web. This paper proposes a new algorithm GraphGen for mining frequent subgraphs. GraphGen reduces the mining complexity through the extension of frequent subtree. For the best algorithm available,the complexity is O(n³·2ⁿ),n is the number of frequent edges in a graph dataset. The complexity of GraphGen is O〔2ⁿ·n^2.5／logn〕,which is improved O(√n·logn) times than the best one. Experimental results prove this theoretical analysis.

Abstract:Availability is one of the most important properties of a storage system. However,it is very difficult to guarantee availability in P2P storage system because of peer churn. This paper argues that it is unfeasible to provide the same availability level to all peers,so it presents a novel P2P storage architecture which builds on the basis of hierarchical management and differentiated service. This architecture has two important characters: First it uses a hierarchical organization according to peers’ character instead of organization as a whole; second,it provides different availability according to peer’s contribution instead of an unbiased service. This not only simplifies the organization of large-scale peers,but also provides a good incentive mechanism. This paper firstly presents a more precise peer behavior model,and then proposes three peer organization strategies and examines their efficiency on different hierarchy in order to study their applicable scopes. Finally,it summarizes the strategies of keeping availability in different hierarchies.

Abstract:People can obtain information from heterogeneous data resources by semantic querying based on ontology. The difficult problem is how to reformulate a global query into multiple sub-queries over those local data bases. This paper presents a kind of uniform expression for queries of concept instances and a method in dividing the query expression into the local ones. All local query results can be integrated for the users’ purposes. Using these methods to query integrated data resources,users can get right results they desire.

Abstract:Using ontology and context knowledge in data mining is one of the effective waies to improve data mining accurateness,which can add general knowledge and certain knowledge in decision factors. How to apply ontology and context knowledge in data mining is discussed in this paper. Firstly,the integration model of ontology and context knowledge is presented,which includes context information categories,context information extended on ontology models and context transformation method. Based on those,using the hierarchy structure of the ontology and context knowledge integration model as an example,the induced learning algorithm is presented in terms of the integration ontology and context knowledge. The experiment of the induced learning is presented and its result is more effective and accurate.

Abstract:Distributed real-time main memory databases are usually applied in some time-critical applications. For these applications,rapid and efficient recovery in the event of site crash is very important. Firstly,through analyzing the recovery requirements of failure in distributed real-time main memory databases,the recovery correctness criteria of distributed real-time main memory database are presented. Then,a real-time dynamic crash recovery scheme based on log (RTDCRS) is presented,and the correctness of RTDCRS is proved. RTDCRS adopts the real-time logging scheme integrating the characteristics of partitioned logging,ephemeral logging and uses nonvolatile high-speed store as logging storage area in order to reduce the logging cost during the normal running. After a site crash,a dynamic recovery strategy based on classification recovery idea is adopted to decrease the downtime. Performance tests and evaluations results show that RTDCRS has better performances than the traditional recovery schemes in two aspects: The missing deadlines ratio of transactions and the time of system denying services after crashes.

Abstract:Optimized association rules are permitted to contain uninstantiated attributes.The optimization procedure is to determine the instantiations such that some measures of the roles are maximized.This paper tries to maximize interest to find more interesting rules.On the other hand,the approach permits the optimized association rule to contain uninstantiated numeric attributes in both the antecedence and the consequence.A naive algorithm of finding such optimized rules can be got by a straightforward extension of the algorithm for only one numeric attribute.Unfortunately,that results in a poor performance.A heuristic algorithm that finds the approximate optimal rules is proposed to improve the performance.The experiments with the synthetic data sets show the advantages of interest over confidence on finding interesting rules with two attributes.The experiments with real data set show the approximate linear scalability and good accuracy of the algorithm.

Abstract:In the Internet,the exponential growth of user traffic has been driving routers to run at higher capacity. Traditional routers consist of line cards and centralized switching fabrics. The centralized switching fabric in such a router,however,is becoming the bottleneck for its limited port numbers and complicated scheduling algorithms. In addition,the fabric is the single point of failure (SPF) in the router. Direct networks,such as 3-D Torus topology,have been successfully applied to the design of scalable routers. They show good scalability and fault tolerance. Unfortunately,its scalability is limited in practice. This paper introduces another type of direct network,called cellular router (CR). With a little modification,this network shows excellent topological properties. Based on this network,two classes of minimal routing algorithms are introduced. The load-balanced minimal routing (LBMR) algorithm makes use of path diversity and shows low latency and high throughput on both uniform random (UR) and Tornado traffic. This paper also discusses some other aspects of the routing algorithms,such as effects of queue length and scheduling algorithms. The CR architecture is a promising choice for the design of scalable routers.

Abstract:The existence of malicious users could damage the correctness and availability of the peer-to-peer (P2P) e-commerce systems. Reputation-Based trust mechanisms can recognize these malicious peers by computing the trustworthiness of the peers. The validity of a reputation-based trust mechanism relies on some well-chosen trust factors,which directly influence the computation of trust value,the accuracy of the trust mechanism,and the resistibility of the trust mechanism to various attacks. However,there are some problems in the above three aspects of existing reputation-based trust mechanisms in P2P environments,such as the selection of trust factors. This paper presents a novel reputation-based trust mechanism for P2P e-commerce systems. In this mechanism,a peer has two kinds of reputations,namely local reputations and global reputations. The local reputation of a peer relative to another peer is calculated in terms of the reference peer’s rating of the transaction between the two peers,whereas the global reputation is computed based on all peers’ rating of the transaction between them. To compute the local and global reputations precisely and to obtain stronger resistibility to attacks as well,many comprehensive factors in computing trust value are introduced in the mechanism. To estimate the validity of the rating given by peers,a quality model and a computational method are also employed to evaluate the objectivity and the credibility of the rating,respectively. To compute the trust value of a peer,the concept of belief factor is introduced to integrate the local reputation with the global reputation. Furthermore,a method is put forward in this paper for determining belief factor. Finally,the effectiveness and resistibility of the proposed trust mechanism are analyzed theoretically and evaluated experimentally. The experimental results show that the proposed trust mechanism outperforms existing mechanisms,and can effectively be applied to the P2P e-commerce system.

Abstract:In mobile IPv6,a route optimization (RO) protocol is proposed to solve the triangle routing (TR) problem,which allows packets to be routed along an "optimal" path. However,RO introduces additional control massages which may cause high signaling costs,and thus RO is not always optimal. The main idea of this paper is to make a performance comparison between TR and RO under diverse network conditions,and find a more effective routing scheme. Several key parameters are introduced into the mathematical analysis such as the packet arrival rate,the mobility rate and the distance,and the total cost function is founded to capture the trade off between network resources consumed by signaling and data routing. Based on the analysis,a new scheme (packet-to-mobility route optimization,PMRRO) is proposed to minimize the total cost. In PMRRO,the route between the MN (mobile node) and the CN (correspondent node) is chosen adaptively according to the value of key parameters such as PMR (packet-to-mobility). Simulation results indicate that the PMRRO scheme has better performance than RO or TR,and it is useful to improve the efficiency of mobility management.

Abstract:Although multicast can be implemented in every layer of network,the existing multicast protocols rarely meet the demands both for flexibility and efficiency at one time. In general,hardware multicast and IP multicast are more efficient than overlay multicast,but the situation of flexibility is quite the contrary. This problem is much pressing in hybrid networks,in which the two demands have the same priority. This paper proposes a novel protocol named half overlay multicast routing that integrates IP layer in-group regional broadcast,address-and-port-translation based overlay multicast and extensible support for hardware multicast into one single model,where the three mechanisms are chosen dynamically for data delivery. By this approach,HOMR obtains the flexibility of overlay multicast. The simulation results show that the overhead of HOMR is low and the performance of HOMR is comparable with pure IP multicasting.

Abstract:In window-based Internet service fault management,improper time window size setting will affect the fault diagnosis algorithm. In order to reduce the impact,challenges of Internet service fault management are analyzed in this paper,and a layering model is recommended. Bipartite graph is chosen to be the fault propagation model (FPM) for each layer. A window-based fault diagnosis algorithm MFD (multi-window fault diagnosis) is proposed for the bipartite FPM. MFD takes the correlation of adjacent time windows into account. As a result,it can reduce the impact of improper time window size setting. Simulation results prove the validity and efficiency of MFD.

Abstract:Network anomaly detection has been an active and difficult research topic in the field of intrusion detection for many years. Up to now,high false alarm rate,requirement of high quality data for modeling the normal patterns and the deterioration of detection rate because of some "noisy" data in the training set still make it not perform as well as expected in practice. This paper presents a novel network anomaly detection method based on improved TCM-KNN (transductive confidence machines for K-nearest neighbors) machine learning algorithm,which can effectively detect anomalies using normal data for training. A series of experiments on well known KDD Cup 1999 dataset demonstrate that it has lower false positive rate,especially higher confidence under the condition of ensuring high detection rate than the traditional anomaly detection methods. In addition,even provided with training dataset contaminated by "noisy" data,the proposed method still holds good detection performance. Furthermore,it can be optimized without obvious loss of detection performance by adopting small dataset for training and employing feature selection aiming at avoiding the "curse of dimensionality".

Abstract:EPFTS (Ethernet-oriented physical frame timeslot switching) is designed to carry the most popular data link frames,i.e.,Ethernet MAC (media access control) frames,and the transmission time for an Ethernet-oriented physical frame (EPF) is defined as a timeslot for transmission and switching the physical layer frames. To resolve the data scheduling problem in EPFTS switches,this paper proposes a new type of scheduling mechanism called TRWFS (timeslot-reservation based weighted fair scheduling) based on the characteristics of EPFTS. The principle of TRWFS is to control the request time of inputs to outputs according to the number of reserved timeslots of each traffic flow and utilize a two-phase iteration mechanism to resolve request conflicts problem. This paper also puts forward three algorithms of TRWFS to show that the implementation complexity of TRWFS is about the same as round-robin based scheduling algorithms. Further simulation results show that even under heavy load conditions,TRWFS algorithms can better guarantee reserved timeslots of each I/O pairs and get better delay and throughput performance compared with other typical algorithms.

Abstract:This paper proposes a scheme to defend distributed denial of service attacks (DDos) based on the source and destination IP address database. The scheme establishes the source and destination IP address database (SDIAD) by observing the normal traffic and storages SDIAD in a three dimension Bloom Filter table. Then this paper cumulates and analyses the new pair of source and destination IP address based on the slide non-parametric cumulative sum (CUSUM) algorithm to detect the DDos attacks quickly and accurately. The secheme updates SDIAD by using a delayed update policy to keep SDIAD timely,accurate and robust. This secheme is mainly applied in the edge router and it can detect the DDos attacks efficiently either the edge router or the last-mile router is the first-mile router. The simulation results display that the secheme do a good performance in detecting DDos attacks.

Abstract:The switching fabric in the Avici TSR uses a 3-D torus topology with each line card carrying one node of the torus.It requires no centralized switching fabric,whereas its scalability is limited by the bisection bandwidth.This paper proposes a novel architecture called Cellular Router (CR).There exist some problems with the basic CR architecture.They are solved by introducing Mirror Points (MPs).This paper also gives the design of line cards in this architecture.In the end,the design of routing algorithms is introduced on this architecture.The CR architecture shows excellent scalability and fault tolerance.It is a promising choice for the design of scalable routers.

Abstract:Network forensics is an important extension to present security infrastructure,and is becoming the research focus of forensic investigators and network security researchers.However many challenges still exist in conducting network forensics:The sheer amount of data generated by the network;the comprehensibility of evidences extracted from collected data;the efficiency of evidence analysis methods,etc.Against above challenges,by taking the advantage of both the great learning capability and the comprehensibility of the analyzed results of decision tree technology and fuzzy logic,the researcher develops a fuzzy decision tree based network forensics system to aid an investigator in analyzing computer crime in network environments and automatically extract digital evidence.At the end of the paper,the experimental comparison results between our proposed method and other popular methods are presented.Experimental results show that the system can classify most kinds of events (91.16% correct classification rate on average),provide analyzed and comprehensible information for a forensic expert and automate or semi-automate the process of forensic analysis.

Abstract:Considering a three dimensional TH-UWB wireless sensor network with n static identical randomly located sensor nodes in a sphere of volume normalized by one cubic meter, the formula for upper and lower bounds on the energy dissipation of some sensor node transmitting a data packet containing R bits to the sink node by multihop fashion are derived using Vapnik-Chervonenkis theory and Voroni tessellation covering sensor field.The main finding is that the upper and lower bounds on the energy dissipation are in inverse proportion to the node density n in the network.Thus,large-scale dense TH-UWB wireless sensor network is more preferable.

Abstract:DDoS attack represents a big problem to the Internet community for its high profile,severe damage,and difficult defending.Several countermeasures are proposed for it in the literature,among which,Probabilistic Packet Marking (PPM) is promising.However, all the existing marking schemes are bearing limitations in some aspects.In this paper, a new packet marking scheme is proposed,which is more prompt because of fewer packets needed,more scalable and more efficient in computation compared with other schemes.Furthermore,this scheme limits attackers' ability in spoofing trace message.

Abstract:Chameleon signatures are non-interactive signatures based on a hash-and-sign paradigm,and similar in efficiency to regular signatures.The distinguishing characteristic of chameleon signatures is that they are non-transferable,with only the designated recipient capable of asserting its validity.This paper introduces a new chameleon hash function based on bilinear pairing and builds the ID-based chameleon signature scheme.Compared with the conventional chameleon hashing functions,the owner of a public hash key in the ID-based chameleon hashing scheme does not necessarily need to retrieve the associated secret key.The scheme enjoys all the attributes in the normal chameleon signature and the added characteristics of ID-based cryptography based on bilinear pairing.