最新刊期
卷 39 ， 期 2 ， 2011
- 摘要：Visual media domain applications are a kind of complex applications which are computation intensive and bandwidth exhausting.This paper described a heterogeneous multicore SoC named EVMP-SoC,which is specially optimized for embedded visual media applications.The chip was implemented using SMIC 0.13 μm CMOS standard cell library technology.We evaluated this chip with some micro application kernels.After exploiting the parallelisms on the chip,optimization based on high level transformation was done to improve the performance of execution of nested loops.Finally,through a case study,the chip shows high efficiency and availability.关键词：heterogeneous architecture;multicore SoC;embedded visual media process;domain applications specific SoC2|2|3更新时间：2025-07-16
- 摘要：An efficient method that can estimate PN sequence of direct sequence spread spectrum (DS-SS) signal is presented in a non-cooperative context,on this base,spread-spectrum hidden message can be extracted blindly.At first,the optimization model with the cumulative correlation value between estimate sequence and SS signal as the objective function is established,and then the PN sequence can be estimated by Genetic Algorithm.Finally,by analyzing the correlation between PN sequence and stego-signal,the hidden messmate can be extracted.The method has been carried out on PC,the experimental results show that,PN sequence estimation performance of the low-SNR DS-SS signal is good,and the extracted hidden message can be identificated.关键词：direct sequence spread-spectrum;pseudo-noise sequence;spread-spectrum steganography;steganalysis;generic algorithm3|1231|1更新时间：2025-07-16
- 摘要：When tracking multiple objects,prior models such as path consistency assumption model are generally established in order to handle occlusion problems.However,if the assumption is greatly distinguished from the real trajectories of the objects,the tracker is doomed to fail.To solve this problem,we put forward a definition of interactive particle filter which adaptively selects appearance template for a particle to measure its likelihood by judging the occluding relationship between each two samples of different objects.The experiments illustrate that our method accurately locates the object even if it is completely occluded as well as the trajectory is impossible to be predicted.关键词：multiple objects tracking;interactive particle filter;occlusion;observation template selection2|1589|13更新时间：2025-07-16
- 摘要：In order to make the pen-based user interface easy to develop with personalized requirements and diverse devices,we propose a model driven development framework.For a pen-based user interface,firstly we present a general development framework,then build a platform independent model and a platform specific model based on the model driven architecture.Furthermore,we introduce the transformation way from the former model to the latter one.At last,a toolkit named "Iris" is provided to support the development.The example application built by "Iris" shows that the model driven development framework can benefit the development of a pen-based interface and reduce the complexity of the development efficiently.关键词：model driven;pen-based user interface;software development method;human-computer interaction2|1212|1更新时间：2025-07-16
- 摘要：We present a hybrid-index structure for high-dimensional data which named HKD-tree (Hybrid K-Dimensional Tree).To make use of two-level parallelization of multi-core clusters,we combined with KD-tree and LSH,which uses LSH in the leaf nodes of KD-tree.Compared with the traditional index structure,the hybrid index structure has effective parallel processing ability and good scalability,which is suitable for the multi-core cluster platform and high-dimensional data indexing.The experiment results show that the performance of the hybrid index structure is superior to the traditional index structure on the multi-core cluster systems.关键词：index;HKD-tree;high-dimensional data;multi-core;LSH2|1141|2更新时间：2025-07-16
- 摘要：Flash memory based storage systems are becoming increasingly prevalent.Garbage collection plays an important role in such systems.This paper analyzes popular file system overload,and points out the problems of existing GC algorithms:the performance of Garbage Collection decreases dramatically under high capacity utilization,which has a great impact on write speed.Then the paper proposes a new flash memory garbage collection mechanism based on differential evolution algorithm (DEbGC),which considers the speed and wear leveling in garbage collection.The experimental results show that DEbGC could amortize the overhead of garbage collection,decrease erase count and number of page copies while satisfying overall performance and wear leveling.关键词：differential evolution algorithm;garbage collection;wear leveling;flash file system2|1528|7更新时间：2025-07-16
- 摘要：The removal of Poisson noise is essential in medical and astronomical imaging.In the framework of Bayesian-MAP estimation,a sparsity regularized convex functional model is proposed to denoise Poisson noisy image in terms of the sparse representation of the underlying image in an over-complete dictionary.The negative-log Poisson likelihood functional is used for data fidelity term and non-smooth regularization term constrains the sparse representations of the underlying image over the dictionary.An additional term is also added in the functional to ensure the non-negative of the denoised image.Based on the Split Bergman iteration method,a multi-step fast iterative algorithm is proposed to solve the above model numerically.By introducing an intermediate variable and Bergman distance,the original problem is transformed into solving two simple sub-problems iteratively,thus the computational complexity is decreased rapidly.Experimental results demonstrate the effectiveness of our recovery model and the numerical iteration algorithm.关键词：image denoising;sparse representation;Poisson noise;split Bergman method;proximal operator2|4057|14更新时间：2025-07-16
- 摘要：Nonnegative matrix factorization (NMF) is an increasingly popular technique for data processing and analysis.For an incomplete data matrix,the weighted nonnegative matrix factorization (WNMF) is employed to decompose it.But the searching step size in WNMF is not optimal along the given searching direction.This paper studies the incomplete nonnegative matrix factorization (INMF) and proposes an accelerated algorithm.First,INMF is transformed into solving alternatively two nonnegative least squares (NNLS) problems.For each NNLS problem,the exact step size is chosen along the searching direction.Then,the complexity of NNLS problems is analyzed.Finally,experimental results show that the proposed method outperforms WNMF.关键词：nonnegative matrix factorization;incomplete nonnegative matrix factorization;missing data problem;weighted nonnegative matrix factorization;nonnegative least squares2|2188|11更新时间：2025-07-16
- 摘要：Variable range analysis is essential for compiler optimization,static analysis and software testing.This paper extends the classical interval abstraction,and defines the new numeric interval-set algebra,Boolean interval algebra and reference interval algebra.Then it presents a unified method of variable range analysis based on abstract interpretation (RABAI).This method uses widening operator to compute variable range in loop,undefined value to describe procedure parameters,and applies function summary as a stand-in for the function.RABAI compresses variable value ranges,and can detect infeasible paths in program.Finally RABAI is applied in Java code analysis tool DTS,and DTS can reduce the false positive rate of bug checking result.关键词：software testing;static analysis;abstract interpretation;interval abstraction;range analysis;infeasible path2|2420|13更新时间：2025-07-16
- 摘要：To solve feature model’s chaos and entanglement problems in certain domain and decouple the feature model and requirement model,we design a feature model componentization method.The method introduces the concept of Role,and implements an algorithm based on it called Feature-Role-Component Algorithm which maps domain features generated by requirement elicitation and analysis to different model components.The Role plays the role of intermediary,and decouples the feature and component,which enables convenient selection and composition between feature variants and enhances the componentization level of the system.关键词：feature model;role;model componentization;feature-role-component algorithm2|1105|3更新时间：2025-07-16
- 摘要：Intricate malwares can result in the failure of Virtual System,and enable the system to be in an unsafe state and difficult to restore.The existing policies thwarting this extreme attack are ineffective.In this paper,based on cooperative recovery among multiple Virtual Machines and agent-based lightweight intrusion detection,an efficient recovery mechanism is proposed for Virtualization systems against malware attacks.The basic policy is to deploy an Emergency Response/Recovery (ER) agent on Virtual Machine to identify the state of the system,and cooperative security among multiple nodes is carried out so that the infected nodes can be rapidly recovered.Simulation results also demonstrate the practicality and efficiency of the proposed schemes.关键词：virtualization system;malware;attack;recovery mechanism3|1397|0更新时间：2025-07-16
- 摘要：
In this paper, a distributed multi-view object tracking algorithm is proposed to address the occlusion problem. The Bayesian sequential tracking framework is used to model the multi-view tracking problem and implemented with particle filtering. In our algorithm, the centralized computing unit is no longer needed. Image acquisition, processing and data fusion can be performed by each camera in parallel. Moreover, an adaptive observation model and an adaptive state transition model are also proposed to enable efficient data fusion and robust tracking against various occlusions. Experiments have verified the effectiveness of our algorithm.
关键词：distributed multi-view object tracking;particle filtering;adaptive observation model;adaptive state transition model+S2062|2191|4更新时间：2025-07-16 - 摘要：Planning as satisfiability is a universal planning framework,which translates the classical planning problem to a series of satisfiability problems,and then solves them using different kinds of excellent SAT deciders.Almost all the existing propositional encoding methods based on planning as satisfiability include a mass of overlapped axioms and redundant axioms,which have a serious influence upon the magnitude of the corresponding encoding theories and the performance of corresponding planners.In this paper,by considering overlapped axioms via analyzing the Graphplan-based encoding,which are necessary but duplicate in the final encodings,we propose a deciding strategy to reduce them.By considering redundant axioms,which are useless in the final encodings,it proposes a deleting strategy to reduce them.The soundness and completeness of proposed strategies are justified.By implementing this reduction and comparing with SATPLAN2006 planner,the experimental results demonstrate that the above strategies can effectively reduce the magnitude of objective encodings without degrading the planning efficiency.Furthermore,a deciding strategy of overlapped axioms via analyzing the state-based encoding is proposed and its soundness and completeness are also justified.The above reduction strategies can also make it possible to improve the potential power of handling larger scale problems,easier to remove redundant axioms,rather than reshaping the whole knowledge base.关键词：intelligent planning;planning as satisfiability;propositional logic;axiom reduction;encoding2|1185|0更新时间：2025-07-16
- 摘要：Most regions in real world have uncertain boundaries.The present models based on crisp regions can not deal with the Direction Relations between the Regions with Uncertain Boundaries(DRRUB).Using broad boundary to represent the uncertain boundary,a novel approach based on the model of MBR(minimum bounding rectangle)is proposed for modeling DRRUB.In this paper,DRRUB are described by MBR-based cardinal direction relations and rectangle algebra,then we study the consistency-based composition of DRRUB,and a method is put forward for calculating this consistency-based composition.关键词：uncertain regions;board boundary;direction relations;minimum bounding rectangle(MBR);rectangle algebra;consistency-based composition2|1157|6更新时间：2025-07-16
- 摘要：In order to solve Traveling Salesman Problem(TSP) more efficient using artificial immune algorithm,using for reference of hierarchical and co-evolutionary idea,a two-floor model based on multiple-population immune evolution as well as Hierarchical Co-evolution Immune Algorithm (HCIA) based on competition-cooperation is put forward.Multiple subpopulations are operated by bottom floor immune operators:local optimization immunodominance、clonal expansion and other clonal selection operators、amelioration of antibody diversity based on improved Particle Swarm Optimization(PSO) algorithm.Multiple subpopulations are also operated by top floor genetic operators:selection、antibody migration、mutation.Through those operators,excellent antibody affinity maturation and diversity of antibody subpopulation distribution was enhanced,the balance between in the depth and breadth of the search-optimizing was acquired.Experimental results for TSP indicate that HCIA has a remarkable quality of the global convergence reliability and convergence velocity.关键词：TSP;artificial immune algorithm;hierarchical;co-evolution;PSO2|2746|4更新时间：2025-07-16
- 摘要：As an important task of data mining,outlier detection has attracted much attention.We discuss the issues of outlier definition and detection based on rough set theory.We propose a new definition for outlier-rough sequence outlier,and the corresponding outlier detection algorithm RSOD.The algorithm constructs three kinds of sequences exploiting the notions of knowledge entropy and significance of attribute in rough sets,and detects outliers by analyzing changes of the elements in the sequences.We compare algorithm RSOD with the current outlier detection algorithms on UCI data sets.And experimental results show that our method is effective for outlier detection.关键词：outlier detection;rough sets;data mining;sequence;knowledge entropy;significance of attribute4|1310|10更新时间：2025-07-16
- 摘要：Cycle level full system simulator based performance evaluation has been indispensable to increasingly complex System-on-Chip designs.Although x86 processors have been the most ubiquitous processors in both commercial and science computing world,there is a scarcity of open source simulators that enable academic researchers to experiment with new x86 microprocessor based designs.This paper presents a cycle level full system simulator for PKUnity-86 System-on-Chip platform,which is based on Geode GX x86 processor and AMBA bus architecture.PKUsim-86 can boot up Microsoft DOS,Windows 98,Windows XP and run common x86 applications.On-the-fly switching between the functional simulation and performance simulation mode is supported.PKUsim-86 executes 0.86 million simulated instructions per second on average.Compared with real machines,the relative accuracy of PKUsim-86 is acceptable.关键词：full system simulation;performance evaluation;system-on-chip;x86 processor2|1147|0更新时间：2025-07-16
- 摘要：Traditional QoS management mechanism of memory controller cannot effectively control the latency of processor accesses.To solve this problem,this paper proposes the PRCC mechanism.Based on the prediction of the future processor accesses,PRCC analyzes the impact of device accesses on the latency of processor accesses,and prevents issuing the accesses that will result in that processor accesses change from row hits to row conflicts,so PRCC can reduce the impact of device accesses on the latency of processor accesses.The evaluation results for PKUnity-3(SK) SoC show that after applying PRCC service,the impact of device accesses on program execution time is reduced from 24% to 8%.By parameter adjustment,PRCC achieves better performance trade-offs between processor and devices than traditional methods.关键词：DRAM;memory access scheduling;system-on-chip;Quality of Service2|1222|0更新时间：2025-07-16
- 摘要：A self-adaptive resource location service model (DST-RLS) based on distributed spanning tree (DST) and search radius limited resource location algorithms SRL were proposed to resolve the bottleneck of the resource location index in the cloud computing.The rule of selecting representative was also presented to make the representatives distributed uniformly in the DST-RLS model,and which improves the load balance of the resource location service.The SRL algorithm supports the complex queries and has time complexity of constant level.The mathematical analysis and simulation results show that the DST-RLS model provides good load balance,reliability and self-adaptive;by limiting the search radius,the SRL algorithm can reduce the search time,lower the network bandwidth usage,and whose performance is better than DST resource location algorithm.关键词：resource location;distributed spanning tree (DST);P2P networking;cloud computing;data grid2|1444|6更新时间：2025-07-16
- 摘要：Feature selection for high-dimensional sparse feature space is an open issue for machine learning research,prevalent 1-norm regularization approaches share some theoretical drawbacks,such as lack the ability to select out grouped features,and can not select more features than the sample size.This paper considers the sparse modeling problem from the stochastic complexity theory perspective,and derive an easy computable model from its Minimax bound approximation.The proposed approach is proved to be optimized,and can perform automatic feature selection similar to its 1-norm penalized alternatives,but overcome their drawbacks.Furthermore,it does not rely on any parametric assumptions about the true data-generating mechanism,which makes it broadly applicable.Various simulations performed with both synthetic and real biological data show that the proposed approach performs similarly to the popular 1-norm penalized counterparts in ordinary experimental setups,and outperforms the other methods in robustness and predictive accuracy for extremely sparse problems.关键词：machine learning;bioinformatics;feature selection;regularization;high dimensional2|1670|0更新时间：2025-07-16
- 摘要：Based on fast evolutionary programming,a novel self-adaptive Cauchy evolutionary programming ACEP to solve the combinatorial optimization problem is proposed.It adopts advantages of Cauchy mutation and alters the search steps in time by adjusting the parameter.Compared with classic evolutionary programming and fast evolutionary programming,it only needs a half population size can be achieved the optimal solutions.The empirical experiments on 0/1 knapsack problem are carried out,the results have supported the superiority of Self-adaptive Cauchy evolutionary programming.关键词：self-adaptive Cauchy evolutionary programming;fast evolutionary programming;knapsack problem2|1020|1更新时间：2025-07-16
- 摘要：Cross validation noise-tolerance classification algorithm is an important method which deals with noisy data set classification problem.According to sample complexity theory,cross validation noise-tolerance classification algorithm validity was proved and applied conditions was given.And noise-tolerance data stream ensemble classifiers was proposed in this paper.Theory and experiment indicated,in contrast with tradition cross validation noise-tolerance classification algorithm,this method had more prediction accuracy.关键词：cross validation;noise-tolerance;classification;ensemble classifiers2|1204|1更新时间：2025-07-16
- 摘要：In order to improve the reasoning efficiency of PC (Product Configuration) and reduce the diversity of variant product in manufacture enterprise,the product reconfiguration technology in PC for manufacturing is studied.In view of the similarity and centrality of user’s requirements in some conditions,the reconfiguration study begins from the existing configuration results to resolve the customized configuration problem.On the basis of PC design based on generic product structure,a description method of configuration results with XML format is proposed.Based on this work,a product reconfiguration method and its key algorithms are given,in which the configuration page is used as the minimal processing unit.The experiment result shows that the above reconfiguration method can reduce the reasoning time with the reconfiguration proportion increasing.关键词：product configuration;generic product structure;product reconfiguration;configuration reasoning2|958|1更新时间：2025-07-16
- 摘要：One of the most important kinds of queries in Spatial Network Databases to support Location-Based Services is the k-Nearest Neighbors (k-NN) query.In this paper,we propose a novel approach to efficiently and accurately evaluate k-NN queries in spatial network databases using network space diagram.This approach is based on partitioning a large network to small regions,and then precomputing distances both within and across the regions.Our empirical experiments with several random data sets show that our proposed solution outperforms approaches that are based on on-line distance computation by up to one order of magnitude.关键词：moving object;spatial network databases;distance index;k-nearest neighbors (k-NN)2|986|3更新时间：2025-07-16
- 摘要：The design of an asynchronous pipelined microprocessor based on transport triggered architecture is introduced.The distributed control in the asynchronous TTA microprocessor,the data dependence might give rise to a wrong result of the executed program.A data source select scheme is proposed to guarantee instructions running correctly on the asynchronous TTA microprocessor.The micro-architecture of the proposed asynchronous processor is presented,and an asynchronous microprocessor is implemented in an 180nm technology.The experimental results show that the implemented asynchronous microprocessor can run correctly,and the asynchronous circuit style can offer a low power solution for future microprocessor design.The power dissipation of asynchronous TTA core takes only about 40% of the synchronous equivalent.关键词：transport triggered architecture;asynchronous circuit;low power;VLSI design2|1377|2更新时间：2025-07-16
- 摘要：The architecture-independent hot trace extraction method is presented,with which the processor architecture-independent trace co-processor synthesis is achieved by promoting trace-based HW/SW partitioning from the level of machine instruction to that of control data flow graph of intermediate code.To support trace predication in cyclic structure for better average hit rate,hash-signature based trace predication is proposed.Based on the above work,a processor-independent co-processor synthesis method is implemented,which can be seamlessly integrated with the hardware and software development process as a system-level design optimization tools.The experiment reveals that the result system performance is increased by 22.6% over instruction trace based method.关键词：co-processor synthesis;intermediate representation;trace;speculative execution2|1203|0更新时间：2025-07-16
- 摘要：The fast implementation of bilinear pairing on elliptic curve has heavily depended on the Miller’s algorithm.In this paper ,an improved Miller’s algorithm is presented and the basic idea of this algorithm is that it combines the {2,3}-double-base chains with the Miller’s algorithm,thus,it can reduces the iteration times in Miller’s algorithm efficiently during the computation of bilinear pairing.What is important,this new algorithm can not only be applied to the case of super-singular elliptic curve,but also can be applied to the case of ordinary elliptic curve.As indicated by the experimental results in this paper,the computational efficiency of our new method has been improved by 10.6%~20.3% on average than other existing methods.关键词：double-base chains;divisor;Miller’s algorithm;Tate pairing4|1562|2更新时间：2025-07-16
- 摘要：The purpose is to discuss the linear complexity of the generalized self-shrinking generator which is based on the m-sequences of GF (q),and analyze the linear complexity on the condition when prime q greater than 3.Reach the upper bound of the generalized self-shrinking generator linear complexity on GF(3),and diminish the linear complexity upper bound.The linear complexity upper bound on GF(q) can have better accurate value.关键词：linear complexity;m-sequences;generalized self-shrinking;characteristic polynomial;stream cipher4|1000|1更新时间：2025-07-16
- 摘要：The concept of symmetric Boolean functions treated in cryptology is transplanted into quantitative logic,and the concepts of symmetric logic formulas and pseudo-symmetric logic formulas are introduced.It is pointed out that logic formulas in two valued logic are closely related to Boolean functions while they have crucial differences.It is proved that the ratio of the number of symmetric formulas with n atoms over the number of all formulas with n atoms converges to zero when n tends to infinite.It is proved that the set of truth degrees of symmetric logic formulas is dense in .It is proved from the viewpoint of topology that the set consisting of all symmetric logic formulas is a nowhere dense set in the classical logic metric space.关键词：symmetric logic formula;truth degree;dense;classical logic metric space;nowhere dense4|1239|16更新时间：2025-07-16
- 摘要：In terms of syntax,the concept of the syntactical truth degree of formulas in the system L*is given in this paper.Then a series of the corresponding properties of the syntactical truth degree of formulas are discussed.Moreover,the similarity degree and pseudo-metric between formulas induced by syntactical truth degree are investigated.In addition,the continuity of the operations in the pseudo-metric space is proved.So then we can provide a possible framework for developing approximate reasoning based on the theory of syntax in the system L*.关键词：the system L*;syntactical truth degree;similarity degree;pseudo-metric;the pseudo-metric space2|1256|8更新时间：2025-07-16
- 摘要：In order to solve the problem of multi-sensor fault detection and isolation,a unified fault model for lock-in-place,gain time-varying and bias time-varying is established.We resort the output equation to converted sensor fault into state disturbance of the dynamic system.A residual generator which can well be used to fault detection and isolation is designed.In addition,we also put forward the method of solving unknown parameters in it.The effectiveness of this method is verified through example simulation.关键词：sensor fault;residual generator;fault detection and isolation;unobservability subspace;fault feature2|1750|8更新时间：2025-07-16
- 摘要：A new modeling and the analysis methods were proposed for Buck converter in continuous current mode(CCM)in this paper.The models of the zero-state responses and the zero-input responses for Buck converter were established.The delay characteristic of the responses was found and was expressed by using a time -sequence.An accurate closed-form analytical solution(symbol form)was obtained in this paper.The analytical solution expresses the relation between the duty and output voltage of converter and is a decoupled form between the inductor current and capacitor voltage.The analytical solution can express the transient state,steady state and ripple of the converter and the switching frequency were not limited.A Soft-Start strategy for Buck converter based on digital control was realized by using the method proposed and was more precise and simpler.The simulation experiments showed the validity of the method proposed.关键词：Switching converter;time-sequence;transient state analysis;precision analytical solution2|1155|1更新时间：2025-07-16
- 摘要：Device collaboration technology is the supporting technology of the large-scale area management and "the Internet of things",so it is of great significance to research device collaboration technology and its system software.First,the stage of device collaboration technology is summarized.Large-scale device collaboration technology is considered as the key issue.How to support large-scale device collaboration,rapid control speed and collaboration security needs to be investigated.Then the software architecture of device collaboration system is summed up,and three key technologies of large-scale device collaboration are analyzed.By analyzing a number of device collaboration applications and the technology scheme,the characteristics,advantages and shortcomings of the present researches are pointed out.The impact of network infrastructure is analyzed.Finally,the future research topics are discussed.关键词：device collaboration;large-scale device collaboration;device collaboration process model language;device collaboration execution mechanism2|2107|4更新时间：2025-07-16
- 摘要：The surface of cobalt crust contains plenty of nonlinear manifold features.In order to solve ultrasonic recognition of cobalt crust using its manifold features,a kind of Discriminatively Regularized Least-Squares Classifier(DRLSC) is introduced in this paper.At first,the fact is found that the original DRLSC is not of convexity in general.And then,based on Ho-Kashyap least-squares algorithm,a modified DRLSC which is of convexity and its kernelized edition(Kernel DRLSC,KDRLSC) are proposed.At last,the proposed classifier is used in ultrasonic recognition of cobalt crust.The experimental results show that the echo recognition correct rates of cobalt crust is improved using the proposed KDRLSC classifier.关键词：manifold learning;DRLSC;Ho-Kashyap classifier;cobalt crust echo recognition2|920|0更新时间：2025-07-16
- 摘要：The efficiency of Differential Power Analysis (DPA) depends on the correlation between power consumption and data value,as well as the attacker's understanding of circuit details.To counteract DPA attack,this paper presents a novel logic,Dual-Rail Configurable Logic (DRCL),which combines the characteristics of dynamic differential logic and configurable logic.The DRCL has constant power consumption which is independent of the data value.At the same time,the uniform structure of DRCL prevents attackers from revealing circuit details by layout reverse analysis.The experimental results show that the proposed logic DRCL has better power constant than the typical DPA resistant logic WDDL.关键词：security chip;side channel attack;differential power analysis;dynamic differential logic;configurable logic2|1345|2更新时间：2025-07-16
- 摘要：In-order microprocessors are increasingly adopted in a variety of multi-core chips due to their advantages in low power,low cost and high scalability.To further satisfy the performance requirement of single-thread applications,improving the load latency tolerance of in-order microprocessors is crucial.We propose an energy-efficient executing ahead mechanism which pre-executes the following instructions instead of stalling the processor when a long-latency cache miss occurs.This mechanism dynamically adjusts the executing ahead policy based on the prediction results of the performance benefit predictor to identify and eliminate the useless executing ahead periods.A confidence-based branch predictor is proposed for unresolvable branches during the useful executing ahead periods.Experimental results demonstrate that the performance is increased by 24.14% only with 4.31% energy overhead on average.Compared with two existing methods,the mechanism proposed in this paper decreases the energy consumption by 7.72% and 10.72% while achieving comparable performance enhancement,thus improves the energy-efficiency by 10.3% and 11.39%,respectively.关键词：single-issue in-order microprocessors;executing ahead;load latency tolerance2|1074|2更新时间：2025-07-16
- 摘要：Because of the features of catadioptric imaging,when the object has linear motion in real space,its trajectory on omnidirectional images is not a straight line,but conic.It leads to the big error in prediction of object’s next position when applying trajectory prediction method for common perspective image.The authors adopt unifying catadioptric imaging sphere projection model to project object’s imaging point in omnidirectional images onto the sphere where the author then establishes omnidirectional kalman filter to predict next position of the object.The result of prediction will project back to omnidirectional image to fulfill position prediction of objects.Experimental results in synthetic and real omnidirectional video demonstrate an obvious improvement in accuracy and stabilization of modified omnidirectional kalman filter in object position prediction.关键词：catadioptric omnidirectional image;Kalman filter;object tracking;trajectory prediction2|1479|0更新时间：2025-07-16
- 摘要：GNSS receivers are required to be capable of receiving weak signals to satisfy indoor Applications.When detecting weak signals,the receiver may falsely acquire or track the cross correlation peaks of strong signals due to the cross correlation interference.Coherent and non-coherent integration are combined to improve the reception sensitivity.Meanwhile,the detection threshold is adjusted automatically by estimating the cross correlation interference power from strong signals to mitigating cross correlation interference.Simulation results show that the receiver can detect weak signal of -160dBm for 20ms coherent and 1s non-coherent integration without the interference of cross correlation.As the power of cross correlation interference increases,the resultant false alarm probability is maintained below the present value by automatically increasing the detection threshold.关键词：GNSS receiver;weak signal;high sensitivity;cross correlation interference3|1624|4更新时间：2025-07-16
- 摘要：Rijndael-256 is analyzed in this paper.We trace the propagation of the plaintexts structure at bit-level to obtain the property,and present a new 3-round distinguisher which needs least chosen plaintexts of all the known integral distinguishers.In this distinguisher,32 chosen plaintexts are encrypted by 3-round cipher,each bit of the cipertext is balanced.Based on the new distinguisher,reduced-round Rijndael-256 are attacked.The paper also analyzes the distinguisher from another point of view,which can also be applied to other byte-oriented ciphers with SPN structure.关键词：block cipher;integral attack;rijndael;bit-pattern3|1459|3更新时间：2025-07-16
- 摘要：This paper proposed the design of Dawning 5000 multicast network based on fat-tree topology.Multicast's latency can be lowed by reducing the interference between unicast and multicast,while the completion of multicast can only be guaranteed by freeing the deadlock.To solve deadlock problem,this paper proposed a deadlock avoidance design based on Resource Bulletin Board.To reduce the interference,this paper proposed a multicast routing algorithm which selects light- weighted multicast path by eliminating heavy-loaded switches.Compared with existing load balancing routing algorithm,the algorithm proposed in this paper can fully use light-weighted paths and achieve three times performance improvement.Besides,under many-to-many multicast communication,Dawning5000 multicast network can achieve more than 90% multicast throughput.关键词：high performance computer;multicast;deadlock;routing algorithm;fat-tree2|1398|1更新时间：2025-07-16
- Technical support is provided by Beijing Founder electronics co., LTD 京ICP备09064830号-19
京公网安备11010802024621 - It is recommended to read the content of this site in Chrome&IE9+. Please switch to extreme mode in browser 360.
- Cookies We use cookies to help provide and enhance our service and tailor content. By continuing, you agree to the use of cookies.
0