摘要:Torus networks win lots of industrial and academic attention by virtue of the superior architecture proprieties.The design of efficient deadlock-free routing algorithms is an important aspect of interconnection networks research.Against the problem that torus networks need numbers of virtual channels to support adaptive routing, we propose an adaptive routing algorithm:Gear,which needs only 2 virtual channels to support deadlock-free adaptive routing in Virtual Cut-Through switched Torus.Gear implements fully adaptive routing by constraining the use of some special virtual channels on the concept of Center Distance.We verify the efficiency of the algorithm with simulation. The results show that,in the same circumstances,the advantage of proposed Gear over classic Dimension-Order Routing and Duato's Protocol is very apparent.
摘要:This paper proposes BuOA (Business unit Oriented Architecture),a novel architecture style for enterprise Web applications.Compared with the traditional layer architecture style,BuOA vertically decomposes a Web application into a group of BUs (Business Units),each of which implements a complete and cohesive business function.To establish loosely coupled relationship between BUs,interactions between them are categorized into four patterns:observing,injecting,weaving and binding.We also provide development toolkits and runtime environment to support development and deployment of BuOA-based applications.Based on a concrete example,we show that BuOA is able to manage system complexity well on design,implementation and deployment stages in software lifecycle,with favorable support on parallel development and dynamic evolvement for enterprise Web applications.
关键词:business unit;connector;business unit oriented architecture;software architecture style;modularization
摘要:Lock-based synchronization may become a performance bottleneck which limits the concurrency in shared-memory machines.In order to solve this problem,a lock-free synchronization algorithm based on hardware CAS(Compare And Swap)primitive is proposed in this paper.In the proposed method,compare and swap instruction provided by underlying processor is used to implement non-blocking synchronization for shared variables in multi-core or multi-thread environment.Global mark value is introduced to avoid performance overhead caused by bits reservation of memory word in the traditional design,and guarantee the consistency.Theoretical analysis and experimental results show that the proposed method can efficiently support arbitary multi-word CAS synchronization,improve the concurrent access performance and provide good scalability.
摘要:Aiming at the problem of global communication management and reliability design in application-specific Network on Chip (NoC),the concept of network monitor is introduced,which monitors overall network real-time conditions and implements path allocation algorithm.Then a novel dynamic routing scheme based on Network Monitor called DyRS-NM is presented.The proposed scheme has the ability to discover and deal with both congestion and permanent faults and distinguish them from transient faults.DyRS-NM can avoid transient faults by using retransmission scheme,and also can detour congested and permanently faulted links by recalculating routing paths.The RTL-level circuits design of the network monitor and fault-tolerant router are realized,and experimental results with a MPEG4 decoder application mapped onto network monitor based 4×4 mesh NoC architecture verify the system performance and cost of area and power consumption.Compared to both static XY routing and fault-aware dynamic routing FADR,significant performance improvements can be achieved by using the DyRS-NM scheme with acceptable additional cost.
关键词:network on chip;dynamic routing;fault tolerance;congestion
摘要:In the self-organizing network (SON) of 3GPP LTE,the algorithms dynamically adjusting many parameters together are better than those adjusting only one parameter.EWPHPO algorithm can not dynamically adjust the handover parameters and the load of the individual cell timely,which causes the dropping probability not to be controlled effectively.In this paper,the WPLBHO algorithm was suggested to improve the EWPHPO algorithm in order to solve these problems.The numerical results were given and compared with those from the EWPHPO algorithm under the same simulation conditions.It is shown that the dropping probability and the ping-pang ratio decrease about 20% and 5%,respectively,but the handoff failure rate hardly vary.Then the total performance of the system increases about 25%.On the other hand,the average load time decreases about 50%.These show that the total performance of the system is evidently improved through our algorithm.
摘要:For the uncertainty natural language processing research in the field of artificial intelligence,this paper,introducing the lattice-valued logic into the concept lattice,proposes the method of transforming the natural language into linguistic truth values and establishes the linguistic truth-valued concept lattice based on linguisitc truth-valued lattice implication algebra.In order to reduce the complexity of constructing the concept lattice,we demonstrate the specific process of constructing a complete linguistic truth-valued concept lattice through combining the multiple sub-concept lattices after analyzing the relationship among the linguistic truth-valued formal contexts and formal concepts,and then give the union theorem and algorithm based on logical conjunction.Experiments show that comparied with Bordat algorithm,the appropriate union algorithm can exponentially ruduce the time complexity when increasing the number of attributes.
摘要:Through the analysis of IAMS algorithm,this paper proposed a zero idle timeslot anti-collision algorithm which is based on continuous collision bit detection (CCBD).CCBD algorithm introduced continuous collision bit detection mechanism,and divided tag collision into individual bit collision and continuous bits collision to process.CCBD algorithm can obtain the actual existence of the continuous collision bits by using continuous collision bit detection mechanism,thereby avoiding the generation of idle timeslots and unnecessary collision timeslots.The theoretical analysis and experimental simulation show that the CCBD algorithm has overcome the shortage of IAMS algorithm and has a good performance during RFID process.
关键词:radio frequency identification(RFID);anti-collision algorithms;continuous collision bit detection mechanism
摘要:When optimizing high-dimensional function,the quantum intelligence algorithms is easy to fall into local optimum.Aiming at the problems,this paper proposes a quantum tabu search algorithm.A strategy which can drop as the index level and adjust dynamic circularly is proposed in the aspect of qubit phase incremental space;a partitioning method which is relevant to the optimum solution in the tabu list and can adjust dynamically is proposed in the aspect of neighborhood space of the candidate solution phase,and a local optimization processing method of candidate solution is added.In order to verify the validation of the algorithm,simulations are made in extreme value of high dimensional function and multidimensional knapsack problem.Experimental results show that the proposed algorithm owns fast convergence speed and high solution precision.
摘要:A novel Elite immune clonal selection co-evolutionary particle swarm optimization algorithm (named,EICS-CPSO) is proposed based on the elite strategy and co-evolutionary mechanism.The algorithm is consisting of one elite subpopulation and several normal subpopulations based on collaborative computing frame.The elite individuals having high fitness from each normal subpopulation will be selected into the elite subpopulation,during the evolution process.The elite subpopulation will be promoted by the immune clonal selection operator with adaptive wavelet mutation.Furthermore,a simple Cauchy learning operator is utilized for accelerating the convergence speed of the pbest particles while the migration scheme is employed for the information exchange between elite subpopulation and normal subpopulations.The performance of the proposed algorithm is verified through a suite of standard benchmark functions,which shows a faster convergence and global search ability and also has a good dynamic optimization performance.Moreover,the parameters of the EICS-CPSO are analyzed in experiments and the results show that EICS-CPSO is insensitive to parameters and easy to use.
关键词:elitist strategy;coevolution;particle swarm optimization(PSO);artificial immune system (AIS);wavelet
摘要:DNA computing employs molecule manipulation to solve NP complete problems that can not be solved using traditional Turing machine.With the deep studying of DNA computing, we found that DNA computation suffers from relatively high error rates.How to decrease error rates has become an important part of DNA computing.This paper present DNA self-assembly model which decreases error rates through decreasing experiment operations.A new N Queen problem algorithm based on the self assembly model is proposed.The proposed algorithm needs O(n2) types of tiles and the complexity of experiment operations is O(1).Obviously,this algorithm significantly reduces the complexity of the experiment, thus improving the accuracy of experimental results.
关键词:DNA-based computing;DNA self-assembly model;N queen problem;tile model
摘要:If the traditional packet matching algorithm is not able to be used in IPV6 environment,it must be of poor performance.This paper combines the real number coding differential evolution algorithm with the traditional packet matching algorithm,imports coefficient of variation in fitness value design and introduces distribution feature to adaptive adjust the acuteness degree of variation.Numerical experiments show that it has higher performance compared with the traditional ones.Package matching time performance has a weak correlation to the number of rule is another significant feature in the algorithm,so it is suitable for processing high peacekeeping large-scale package matching problem.Applying the algorithm to IPV6 network,the packet can linear speed forward.The proposed method is applicable to the firewall,router and other network equipment.
关键词:differential evolutionary algorithm;coefficient of variation;IPV6
摘要:This paper proposes a privacy preservation method based on random projection to overcome the curse of dimensionality in privacy preserving data mining.To prevent leaks of random matrix which can lead to the reconstruction attack,it first proposes the concepts of secure subspace and secure subspace mapping.Then,it constructs a secure subspace mapping using hash technique,which is implemented by a random projection matrix,and it achieves a low distortion embedding while preserving the data privacy.Finally,it proves that the secure subspace can preserve the Euclidean distance and inner product between any two original points.The experimental results show that the proposed technique can ensure the data quality in different data mining applications effectively under the precondition of preserving data privacy.
关键词:privacy preservation;high dimensional data mining;hash technique;random projection;secure subspace
摘要:In the field of blind image separation based on sparse component analysis,the separation efficiency and accuracy is directly affected by the valid number of clustering samples.For this problem,a new algorithm for detection of points in the Haar wavelet domain where only single source contributions occur was proposed.The algorithm identified the single source points(SSPs)by comparing the absolute direction between diagonal component and horizontal component of Haar wavelet coefficients of mixed images.After screening SSPs,the signal features are sparser.The experiment results showed that the algorithm could estimate the mixing matrix faster and more accurately,and it could inspire to identify the latent variables by statistical histogram.
摘要:The rate distortion optimization (RDO) which uses mode-selection based prediction coding can achieve the highest coding efficiency in H.264/AVC video coding standard.However,exhaustive checking of all the prediction modes increases computational complexity significantly.In this paper,we propose a hierarchical decision structure for fast inter-frame mode selection to reduce the complexity of the exhaustive RDO mode decision process.Different modes of macroblock are first classified into 4 layers in terms of the characteristics of the structure of inter-frame prediction coding in H.264/AVC.Using the coupled relationship between the spatial and temporal complexity of the motion-estimation prediction coding,the adaptive threshold for each layer can be calculated.The obtained threshold can then be applied to each layer to terminate the mode selection process while maintaining almost the same R-D performance in H.264.Experimental results show a 45.79% reduction on computational time and the sane RD performance as the baseline coder by using the proposed fast selection method.The results also demonstrate that the proposed algorithm is superior to two recent fast mode decision algorithms,with the peak signal-to-noise ratio (PSNR) being improved by 0.05dB on average when compared to Zeng's algorithm and the entire encoding time being reduced by about 20% on average when compared to Choi's mode decision algorithm.
摘要:Learning from group probabilities helps to protect the privacy of users and has become a hot topic in the community of machine learning.The traditional group probabilities based learning methods have gained certain success,however,they still fall short when the prior information are not fully provided.In order to solve this problem,a novel transfer learning method called transfer group probabilities based learning machine (TGPLM in abbreviation) is proposed by integrating group probabilities into the principle of structure risk minimization.In TGPLM,a novel learning criteria is proposed based on reusing the related domain knowledge by minimizing domain similarity distance,which makes the proposed TGPLM not only make full use of the group probabilities in the current scene,but also learn the existing useful knowledge in the history scene effectively.Experimental results on the artificial,UCI and PIE face datasets show the effectiveness of the proposed method.
摘要:A space mapping optimization method based on the group delay of the reflection coefficient is introduced.A high temperature superconducting (HTS) 8-pole filter and a quadruplexer are designed using the proposed method.The resultant filter has a center frequency of 3000MHz and a band width of 40 MHz.And the four channels of the quadruplexer are 3640MHz~3740MHz,3780MHz~3820MHz,3880MHz~3930MHz and 3970MHz~4030MHz,respectively.The simulation results of the filter and the quadruplexer show that the return loss is higher than 22dB in the passband and the ripple coefficient is less than 0.1dB.In addition,the isolation of the four channels of the quadruplexer is more than 40dB.It is shown that the simulation results is consistent with the theoretical results,which verifies that the group delay of the reflection coefficient based on space mapping method is an efficient and accurate approach to design HTS filters and multiplexers.
摘要:A novel human-computer interaction(HCI)is developed based on multimodal visual features aiming at some limits at present.Two-dimensional Gabor wavelet is adopted to extract some visual features of global face orientation,which overcomes some difficulties including extraction of some facial distinct features,discrimination among some different facial orientations.An efficient and fast approach to locating center of eyes is proposed based on facial geometric distributions without considering facial resolution,eyes closing or opening and user's wearing.Some prominent multimodal visual features for classification are selected to machine learning and training to determine the pointing target after evaluating performance of some extracted visual features.Non-wearable and natural HCI modal can be realized in which user can move freely without wearing any markers when he points at some targets.Their daily skills can be exerted fully during HCI.Experiment results indicate that the developed approach is efficient and can be used to natural non-wearable HCI.
摘要:There are some unresolved issues left behind for many traditional dynamic gesture recognition methods,such as Hidden Markov Model(HMM),Neural Network(NN),and statistical classifiers.For example,they require a large number of training examples and the involvement of expert users in the training process.Moreover,they are used for some specific gesture sets which are difficult to be extended.In this paper,we first build a task model and a state transition model for vision-based dynamic gestures.Then we propose a method for 3D dynamic gesture recognition based on small sample learning.Next we design a toolkit for development of user-defined gestures.Finally,we develop a gesture-based interactive television prototype.Experimental results verify the validity of our method.
摘要:A 2-D potential analytical solution in the oxide layer and space charge region is derived.The unknowns are able to be solved by a group of linear equation,which is orthogonal expanded from the connected condition identity of oxide layer and space charge region.The solution of potential extreme point is also presented.There are many advantages of small calculating amount,without any compromise in accuracy and adapter parameters.Moreover,this model can be directly used in circuit simulation.The potential distributions and threshold voltage of NMOSFET have been derived in subthreshold.Excellent agreement with MEDICI has been observed.
关键词:metal-oxide-semiconductor field effect transistor(MOSFET);potential analytical model;gate oxide layer;space charge region
摘要:Originally from the noise and errors in measurement,the fluctuations make the extraction of MOSFET's threshold voltage unstable,which will become worse in the case of the derivation operation.We proposed that the Savitzky-Golay filtering algorithm could be introduced into the derivation process in order to alleviate the influence of fluctuations effectively.A criterion for assessing the goodness of filtering was presented,and then the extraction of threshold voltage of MOSFET could be achieved stably and automatically,which will be in favor of the characterizing MOS devices and designing integrated circuits.
摘要:To solve the problems that the conventional brain-computer interfaces (BCIs) transfer the mental intents into the control commands with a low judgment accuracy and unsatisfied information transfer rate (ITR),a novel hybrid BCI called the P300-SSVEP-Speller was developed on the basis of the P300-Speller to combine SSVEP blocking (SSVEPB) features with P300 potentials.The offline usage of twelve subjects was investigated in the study.Results show that the elicited SSVEPB has little influence on the emergence of the P300 potentials in the degree of temporal signals.In addition,from the view of character recognition,the novel P300-SSVEP-Speller system has a better performance than the conventional P300-Speller on the judgment accuracy and ITR.
摘要:The TORA system model is transformed to a feedback cascade model with the partial feedback linearization and a global coordinate change.A two-steps recursive backstepping algorithm is proposed for the TORA system to obtain a nonlinear controller.During the design process,the system stability is guaranteed with the Lyapunov theory.The simulation results show the TORA system under any initial states is globally asymptotically stable to its origin.
摘要:This paper presents a programming language for designing signed multiplier circuit.The key idea is using instruction to express the encoding units,addition tree units and fast adder units of multiplier,and using the connection of instruction description to obtain a multiplier.The multiplier of program through Lex and Yacc translate source code containing connection into Verilog code.Seven typical structures of 32 bits signed multipliers are obtained by the instruction description.Under 200MHz synthesis condition and in GRACE 0.18μm process,these multipliers are run for logic synthesis,placed and routed,static timing analysis,and power analysis.The experiment results suggest that the speeds of all the seven multipliers show advantage over that produced by Synopsys design ware,and the multiplier performance composed of modified Booth Radix4 encoding,redundant binary addition tree and carry skip adder exceeds that produced by Synopsys design ware by 35%.Therefore,this language can be used to the application of high performance multiplier design.
摘要:A multi-granular computing architecture,namely a top-down decomposition and bottom-up synthesis is proposed.The structure plays a very important role in quotient space model(QSM),and the QSMs with different structures generally vary.With respect to composition,a composition of topologies of different quotient spaces is generally not a quotient topology,while a operator composition of operators of different quotient spaces is a quotient operator.As for decomposition,it acquires the two important conclusions when defining the concepts of problem equivalence and reversible decomposition.One is that the granulation for two structures by means of either chained or directed are equivalent,the other is that an orthogonal decomposition in QSM with a algebraic structure is reversible,but a similar conclusions is not true in QSM with a topological structure.
摘要:CPS is a kind of networked embedded system.Its trustworthiness and cost are impacted by energy consumption.Because software is the main component of CPS,it is influenced by the energy consumption of CPS.So it has become a major challenge for CPS development to design low-power,high trustworthiness CPS software.Existing analysis and evaluation methods have problems such as the experimental high cost and big price.If the trustworthiness problem of software can be found in the design phase,it can effectively reduce the cost and improve development efficiency.So modeling,analyzing and evaluating the software architecture energy consumption of the CPS software architecture design stage can help find the energy consumption design defects.It can effectively improve the trustworthiness of the CPS software and reduce development costs.Against this,first build the CPS software architecture energy consumption based on the energy consumption time Petri nets;then by analyzing the trustworthiness problems of the CPS software influenced by the energy consumption,propose the evaluation method for CPS software based on the energy consumption model;finally,indicate the effectiveness of the proposed methods through an instance.
摘要:Conformance test is performed to verify and validate the correctness of a system implementa- tion.In view of easily causing the problem of state space explosion,and due to time constraint in data processing for real-time system,a symbolic test generation method is proposed.Firstly,symbolic transition systems and timed automata are extended to establish a semantic model TSIOSTS,based on which an extension of timed conformance relation tioco is defined.Then,with tioco and the symbolic execution strategy,a symbolic timed behavior tree of the system model under test is yielded and transformed into test cases.Finally,the proposed method is applied to conformance test of temporary speed restriction server of the CTCS-3 train control system,and the results present feasibility and validity of the method.
关键词:real-time system;conformance test;time safety input-output symbolic transition system;symbolic test generation;test case generation
摘要:Marching-on-in time (MOT) method is used to solve the time domain surface-wire integral equation (TDSWIE),so that we can solve complex problems,such as wire antennas mounted on the perfect electrically conducting (PEC) surfaces,etc.In order to improve the late-time stability of the method,an efficient scheme based on accurately calculating the elements of the impedance matrix is adopted.In the numerical simulations of a dipole antenna mounted on a PEC sphere and a spiral antenna mounted on a PEC antenna pedestal,the transient surface current,S11 parameter,the input admittance and the directivities of the antennas are obtained.Numerical results demonstrate the efficiency and accuracy of this method.
关键词:marching-on-in time (MOT) method;time domain surface-wire integral equation (TDSWIE);antennas
摘要:Transverse flux motor has been initially applied to various fields with higher torque density.The purpose of this survey is to provide a review of transverse flux motor research.The topologies and their advantage of transverse flux motor are given,basis on a brief analysis of the mechanism of the transverse flux motor in this paper.The research of transverse flux motor is summarized.The main research is pointed out,including structure study,three-dimensional magnetic field analysis,cogging torque,leakage flux,power factor,SMC research,control systems and applied research.The development trend of the transverse flux motor in high torque and low speed is explored.
摘要:A new method of linearity based on predistortion for memory nonlinear system is introduced in this paper.Different from the traditional predistortion models (direct study model and indirect study model),the transfer function instead of the inverse transfer function is obtained for memory nonlinear system.The transfer function is presented as piecewise linear function first,and then be used in predistortion process.The results show that the NMSE(Normalized Mean Square Error) of the memory nonlinear system output is reduced about 80dB,and the ACLR(Adjacent Channel Leakage Ratio) is reduced over 70dB,comparing with no predistortion system.The memory and nonlinear characteristics are eliminated,and the relationship of input and output becomes linear and memoryless.
关键词:predistortion;piecewise linear;adjacent channel leakage ratio;memory nonlinear system
摘要:Combined with the trend of the master-slave synchronization in the nanosecond precision,this paper analyzes the deviation and noise type in the master-slave synchronization,describes how to use the "Loose Coupling" Phase Lock Technique to rectify deviation and restrain noise."Loose Coupling" Phase Lock Technology has achieved better result in the laboratory test and practical project application,master-slave relative frequency deviation in the laboratory test is better than ±5E-14 a day (single slot thermostatic crystal clock);the time interval deviation in project application is better than ±20ns 7 days(rubidium atomic clock),the relative frequency deviation is better than ±3.3E-14 7 days(rubidium atomic clock).
摘要:The tracking of target is a challenging issue in computer vision.In this paper,we propose a visual object tracking algorithm based on ML estimation and L2-norm.Firstly,the model of sparsity constrained ML is established.Abnormal pixels in the samples will be assigned with low weights to reduce their affects on the tracking algorithm.Then,L2-norm minimization is used to solve the sparse coding.Finally,the object tracking results is obtained using Bayesian MAP estimation.Compared with other popular methods,our proposed method reduces the computational complexity and has stronger robustness to abnormal changes(e.g.occlusion,rotation,scale change,illumination,etc.)
关键词:sparsity constraint;maximum likelihood (ML);L2-norm minimization;Bayesian MAP(maximum a posteriori probability) estimation
摘要:This paper introduces FFC(formal function cryptanalysis)of block cipher that expresses each bit in the ciphertext as pure formal functions in terms of the bits of the plaintext and master key by symbolic computation.As an application,we give a meet in the middle attack on 13-round LBlock lightweight block cipher.The meet in the middle attack has a complexity of 276.2 13-round LBlock encryption using only 1 known plaintext,which is better than the algebraic attack given by Nicolas Courtois etc.at FSE 2012 on 8-round LBlock,with data complexity 6 known plaintexs.
关键词:formal function cryptanalysis;method of formal coding;LBlock;symbolic computation;meet in the middle attack;block cipher
摘要:This paper proposes a novel soft-switching voltage source inverter with parallel resonant DC link to improve efficiency of inverter.We added auxiliary resonant unit to DC link of conventional inverter to make DC-bus voltage decreased to zero periodically,which realized zero-voltage operation of all switching devices in inverter.Furthermore,we also operated auxiliary switches under zero voltage switching or zero current switching.In addition,we avoided the bulk capacitor in the circuit and there was no center tap potential variation problem.We analyzed the operation principle and presented the equivalent circuits at different operation modes.We had built a 5kW laboratory prototype.Experimental results show that we can operate main switching devices under zero voltage and enhance the efficiency by 3% under rating power,compared with hard-switching inverter.
关键词:zero voltage switching;DC link;parallel resonant;inverter