Abstract:Finding networks with minimal cost to connect points is a key problem in VLSI design, which can be described as obstacle-avoiding shortest path and minimum Steiner tree problem according to whether the number of points is greater than 2. Connection graphs, such as track based on graphs GC and GT and free area based graphs GF and GG, are effective tools for the shortest path problem, which is the foundation of the Steiner tree problem. The contribution of this paper includes three points: The dynamic algorithms for querying the shortest path between two points on each connection graph are designed and analyzed for the first time; secondly, all algorithms for Steiner problem on each connection graph are analyzed; The number of candidate Steiner points is reduced from O((e+p)2) to O((t+p)2) in the 3-Steiner algorithm on GC, where e, t, p presents the number of edges, extreme edges of obstacles and terminals; An average Q(t) algorithm for 3-Steiner problem are designed on GG.

Abstract:HLA (high level architecture) is the standard for modeling and simulation put forward by the American Department of Defense. Time management is an important component of HLA while GALT (greatest available logical time) algorithm is RTI (runtime infrastructure)抯 critical technology in implementing time management. An improper GALT algorithm can easily result in a deadlock so that the whole federation can not advance any more. On the basis of the GALT algorithm introduced by Frederick Kuhl, the principles of deadlock are discussed and some important results are revealed in this paper. If deadlock occurs in a simulation, all federates must have the same GALT and the same output time respectively, and GALT is also equal to output time, and a federate whoselookahead is greater than zero must suspend because of a NMR or NMRA request, other than a TAR, TARA or FQR request. Finally, a GALT algorithm without deadlock is also brought forward in this paper that is calledStature-Measuring, and this algorithm can provide reliable technology support to develop time management services of RTI.

Abstract:Despite the steady progress made in the area of speech recognition and a high number of practicalapplications, it is widely acknowledged that recognition technology today is not at the desired level. One mainobstacle is what said "robustness". This paper focus on one popular idea in antagonizing speech systemvulnerability-channel normalization, and presents a new normalization algorithm MLCN (multi-layer channelnormalization), which exploits the recursive compensation progress in two domains (spectral domain and cepstraldomain) to depress the noise and channel distortion, so that the more robust speech representation for the followingprocessing is achieved. A new frequency-dynamic feature extraction algorithm is also proposed due to theintroduction of MLCN, which allows dynamic information integrated in the final feature vectors. Experimentalresults of the gallina system demonstrate the validity of the new algorithm.

Abstract:An approach is introduced to derive specificity in default theories. Compared with other methods, themethod handles priority quite well and has lower complexity. Then the prioritized stationary semantic for defaultlogic is defined. The method can strengthen the cautious stationary default reasoning without increasing thecomputational complexity very much.

Abstract:A method is presented to identify some pieces of specific information in multi-carrier data streams byfeature words and based on PinYin matching. An effective knowledge approximation method is used to judge therelation between feature words and context by statistics theory. The part of speech transfer-value as systemknowledge can be obtained by inductive learning of training corpus. When data streams are evaluated, theevaluation value can be gained according to the system knowledge by matching all feature words and based on theirPin Yin, which examines the comparability with context regular of part of speech between all feature words in datastreams and themselves in training corpus. Further more, if the evaluation value exceeds the threshold, the datastreams will be shielded. Experimental results show that the effect of the experiment system based on this method isefficient for identifying ill information and monitoring & controlling their spreading by multi-carrier data streams.

Abstract:Feature selection has an important application in the field of pattern recognition and data mining etc. However, in real world domains, if there are spatial data operated in the application, the performance of feature selection will be decreased because of without considering the characteristic of spatial data. In this paper, a feature selection method from the point of the characteristic of spatial data, named MEFS (maximum entropy feature selection), is proposed. Based on the theory of maximum entropy, MEFS uses mutual information and Z-test technologies, and takes two-step method to execute feature selection. The first step is predicate selection, and the second step is to choose relevant dataset corresponding to each predicate. At last, the experiments between feature selection algorithms MEFS and RELIEF, and between ID3 classification algorithm and classification algorithm based on MEFS are carried out. The experimental results show that the MEFS algorithm not only saves feature selection and classification time, but also improves the quality of classification.

Abstract:Multiwavelets is a new development to the body of wavelet theory. Multiwavelets simultaneouslly offers orthogonality, symmetry, and short support, which is not possible with scalar two-channel wavelet systems. A new theory and algorithm for speckle reduction of ultrasound (US) image with multiwavelets multiple resolution analysis (MRA) are presented and investigated in this paper. The decomposition ratio and reconstruction error of two kinds of multiwavelets are evaluated and the optimum pre-filters suitable for the two kinds of multiwavelets are obtained. Fully and clearly analytic expression of multiwavelet transformation is given .The experiment of speckle reduction to US image is implemented by choosing local error as threshold. Results show that multiwavelet transformation is useful for speckle reduction and there have little noise and reserve important image features such as boundary when compared to results obtained from existing the denoising methods alone.

Abstract:The noniterative algorithm of discrete hierarchical MRF (Markov random field) model has much lower computing complexity and better result than its iterative counterpart of noncausal MRF model, since it has causality property between layers. A new model based on the hierarchical MRFhalf tree model is proposed for only one image can be obtained in image segmentation, whose MPM (maximizer of the posterior marginals) algorithm is inferred too. The proposed model not only inherits the advantages of general hierarchical MRF model but also does better: it makes large image more tractable within much less time, prevents data underflow appeared in computing, and alleviates the block artifacts occurred in hierarchical models. It is especially fit for large scale images.

Abstract:XML is extended with functional dependencies, which are fundamental to semantic specification. Based on DTD, tree model for XML and path expressions, definitions for node value equality and node set are given. The definition for functional dependency, logical implication and path closure are studied; the satisfiability of functional dependency in a given DTD is proved. A sound and complete set of inference rules is presented, and an algorithm to calculate closure for XML paths is proved.

Abstract:Data exceptions often reflect potential problems or dangers in the management of corporation. Analysts often need to identify these exceptions from large amount of data. A recent proposed approach automatically detects and marks the exceptions for the user and reduces the reliance on manual discovery. However, the efficiency and scalability of this method are not so satisfying. According to these disadvantages, the optimizations are investigated to improve it. A new method that pushes several constraints into the mining process is proposed in this paper. By enforcing several user-defined constraints, this method first restricts the multidimensional space to a small constrained-cube and then mines exceptions on it. Experimental results show that this method is efficient and scalable.

Abstract:In this paper, the authors attempt to revisit the behaviour of HITS from a different point of view. Namely, a similarity-based analysis model is proposed to observe the distillation procedure. By defining a generalized similarity, an algorithm is presented, which can improve the quality of distillation using only hyperlinks. A topic exploration function is also integrated into the algorithm framework, which enables end-users to search less popular topics when multi-topics are involved in queries. The experimental results reveal two benefits from the new algorithm: the improvement of distillation quality without utilizing any content information of pages, and an additional ability to explore the topics emerging in the query results.

Abstract:Mining maximum frequent itemsets is a key problem in many data mining application. Most of the previous studies adopt an Apriori-like candidate set generation-and-test approach. However, candidate set generation is still costly, especially when there exist prolific patterns and/or long patterns. In this paper, a fast algorithm DMFIA (discover maximum frequent itemsets algorithm) and its updating algorithm UMFIA (update maximum frequent itemsets algorithm) based on frequent pattern tree (FP-tree) for mining maximum frequent itemsets is proposed. The algorithm UMFIA makes use of previous mining result to cut down the cost of finding new maximum frequent itemsets in an updated database.

Abstract:Ambiguity of query terms has been a long-standing problem in information retrieval field, which becomes more serious in Web searching. A method for automatic query expansion based on query logs obtained from users?daily usage is suggested. This model establishes probabilistic relationship between terms in documents and in user queries through statistical learning from the log, and selects high-related expansion terms based on Bayesian theory. These expansion terms are added into the original query to formulate a new one in order to improve the effectiveness of retrieval. Experimental results show that this technique is more adaptive to Web searching, and can improve the precision of document retrieval markedly compared with conventional ones.

Abstract:An important feature that differentiates moving object database (MOD) from other databases is that MOD can support queries for locations recorded not only at the sample time, but also at time between samples and even at time in the future. The most important issue in MOD research is to build models of moving objects?mobility and their locations update. The advance in mobile Internet and location-based services requires appropriate models be built for popular dumb terminals, such as cell phones, which can only be located by auxiliary facilities like GSM/CDMA networks. Considering that users of such kind dumb devices often move on the road network, a novel model for these moving objects to calculate their velocities in the past and future and to update their locations with varied sample time based on road network is presented in this paper. Compared with traditional velocity models, the proposed model can reduce the location prediction errors effectively. And when comparing with the model of updating locations with uniform sample time, the proposed model can reduce the communication traffic between moving objects and locating facilities with the same average prediction error.

Abstract:Mobile IPv6 offers two methods of multicast services for mobile host called remote subscription and home subscription, which have complementary advantages and shortcomings. An efficient mobile multicast protocol provided in this paper integrates remote subscription and home subscription. It uses a hierarchical route architecture, which reduces the frequency of rebuilding multicast tree because of hosts?mobility. The results of analysis and simulation are also given.

Abstract:Path expression is one of the core components of most XML query languages, and many evaluation methods for path expression queries are proposed recently. However, there are few researches on the issue of path expression optimization. In this paper, two kinds of path expression optimizing principles are proposed, named path shorten and path complementing, respectively. The path shorten principle reduces the querying cost by shortening the path expressions with the knowledge of XML schema. While the path complementing principle tends to substitute the user queries with the equivalent lower-cost path expressions. The experimental results show that these two techniques can work on most path expression queries and largely improve the efficiency of path expression query processing.

Abstract:Recommendation system is one of the most important technologies in E-commerce. With the development of E-commerce, the magnitudes of users and commodities grow rapidly, resulted in the extreme sparsity of user rating data. Traditional similarity measure methods work poor in this situation, make the quality of recommendation system decreased dramatically. To address this issue a novel collaborative filtering algorithm based on item rating prediction is proposed. This method predicts item ratings that users have not rated by the similarity of items, then uses a new similarity measure to find the target users?neighbors. The experimental results show that this method can efficiently improve the extreme sparsity of user rating data, and provid better recommendation results than traditional collaborative filtering algorithms.

Abstract:In order to efficiently exploit high performance network, researchers have developed many user-level protocols, which can achieve high bandwidth and low latency the bottom hardware supplies. But user-level protocols supply a completely different API, which makes them only support science computing, and traditional Socket-based network application program with kernel-level protocol cannot run on them. To solve this problem, a IP supporting module has been implemented on user-level protocol BCL-3, which makes it support both science computing and existing TCP/IP-based network application programs efficiently. And based on software overhead analysis of TCP/IP, BCL-3 adopts some optimization in IP supporting module, which makes TCP/IP over BCL-3 achieve high performance. The improved BCL-3 has been run on Dawning3000L super server. On Dawning3000L, TCP/IP over BCL-3 has achieved performance with maximum bandwidth of 938Mbps and minimum one-way latency of 48.1μs.

Abstract:Computer forensics is the technology field that attempts to prove thorough, efficient, and secure means to investigate computer crime. Computer evidence must be authentic, accurate, complete and convincing to juries. In this paper, the stages of computer forensics are presented, and the theories and the realization of the forensics software are described. An example about forensic practice is also given. The deficiency of computer forensics technique and anti-forensics are also discussed. The result comes out that it is as the improvement of computer science technology, the forensics technique will become more integrated and thorough.

Abstract:Watermark technology has been developed to tackle the problem of unauthorized copying and distribution of digital data. Several different schemes have been proposed in the last few years, but most of them are symmetric, i.e., the key used for watermark embedding is just the same used for watermark detection. However, in many applications, an asymmetric scheme is needed, where the secrete information used to detect the watermark is not enough to modify, counterfeit or remove the watermark. In this paper, some watermark verification protocols based on bit commitment and zero knowledge proof are proposed. The ownership prover insert the watermark into the host signal using symmetric watermark technology based on spread spectrum. The watermark detect key is sent to the verifier hiding in bit commitment. By the interactive protocol between the prover and the verifier, the verifier can extract the embedded watermark, but he can not modify, counterfeit or remove it. Protocols are proposed to verify one watermark bit and several watermark bits respectively. Those protocols can be used to verify the watermark information inserted into image, audio and video using spread spectrum watermark technology.

Abstract:Most previous DWT-based watermarking algorithms belong to either private watermarking algorithms or fragile watermarking algorithms, but there are few DWT-based robust public watermarking techniques for copyright protection. By taking full advantage of the masking characteristics of the human visual system, first a JND (just noticed difference) threshold matrix based on block is given in this paper, and then a robust image-adaptive public watermarking technique operating in DWT domain is presented. Firstly, the 8? blocks of the original image are rearranged into a 1-D Hilbert sequence in Hilbert scanning order. Then two neighboring blocks are selected from the Hilbert sequence of the host image blocks in turn, and 1-level DWT is applied to the two chosen blocks. Finally, a corresponding detail subband is chose from three detail subbands of the two neighboring blocks at a time, respectively. A binary watermark with visually recognizable patterns is embedded into the host image by modifying the polarity of the average value of the two corresponding subbands. The embedded watermark is invisible to human eyes and adapted to the original image by exploiting the HVS masking characteristics. The experimental results show that the proposed algorithm is effective and robust to common image processing operations and some geometric distortions such as cropping, pinching, pixel-shift, and so on, especially, it receives better robustness under signal enhancement operations. So a conclusion can be made that the proposed technique is practical.